searchlore

0:06

I see a large IPO on the horizon.

0:13

>> You're surrounded by journalis position. Overnight success. The place 3000 double. That's misinformation.

1:03

Quarterly clearing order inbound. >> Let's just roll.

1:13

We are surrounded by general. Hold your position. Come, get up. Trust the experts. Five. We are experts. Founder knowledge. Five good.

1:46

>> I see multiple journalists on the horizon. Stand by. >> UAV online. >> Blaze. Double blaze. Triple blaze. Double kill. Cop is waste. Team Deathmatch. We are experts. Triple blade. That's just wrong. Right. clearing order inbound. >> Come get up.

3:25

We are surrounded by journalists. Hold your position. Strike one. >> Strike two.

3:47

>> Activate golden retriever mode.

3:56

>> Trust the engineers here.

4:02

market clearing order inbound. >> Five.

4:22

I see more and more journalists on the horizon. Stand by. founder. >> You're watching TVPN.

4:38

Today is Monday, March 2nd, 2026.

4:41

We are live from the TVPN Ultradome, the temple of technology, the fortress of finance, the capital of capital.

4:47

Let me tell you about ramp. com. Time is money. Save both.

4:51

Easy to use corporate cards, bill payments, accounting, and a whole lot more all in one place.

4:53

Uh it was a massive weekend. So much news.

4:56

Um, we are very fortunate to be joined by Ben Thompson at noon.

5:00

Let's pull up the linear lineup and show you the run of show today.

5:04

Linear, of course, is the system for modern software development.

5:07

70% of enterprise workspaces on linear are using agents.

5:09

We got Ben Thompson, James Bashar, John Quinn's coming back in person again.

5:13

We're very excited to be joined by him talking about tariffs.

5:17

>> A uh a monster lightning round with five different guests joining.

5:20

We got some acquisition news.

5:21

We got some funding news.

5:23

We got some uh takes on tech and AI and media.

5:25

We're going all over the place.

5:27

It's going to be a fun fun show. But we missed you. We missed you.

5:29

On Friday, we were traveling. We went to Montana.

5:33

>> Terrible day to be out.

5:35

>> Terrible day to be out because it was >> every single time. Yeah. >> We've had an off day. Yeah.

5:39

>> It ended up being a massive news day. So, lesson. >> Yeah. >> Never take a day off. >> Yes. Never take a day off. Truly.

5:43

Uh what a what an absolutely crazy weekend.

5:49

Of course, there's the the war with Iran.

5:50

Uh the big news in tech was the US halts the use of anthropic AI after ten over after after tension over guard rails.

5:59

So uh this is in the Wall Street Journal.

6:02

The federal government will stop working with artificial intelligence company Anthropic, President Trump said, marking a dramatic escalation of the government's clash with the company over how its technology can be used by the Pentagon.

6:13

quote, "I am directing every federal agency in the United States government to immediately cease all use of anthropics technology. We don't need it.

6:22

We don't want it, and we do not do business with them again.

6:26

We will not do business with them again," Trump said Friday in a social media post.

6:28

The Defense Department and other agencies using anthropics claim phase out period, the president said, adding that there would be civil and criminal consequences if the company isn't helpful during the transition.

6:40

Six months to switch from one LLM to another feels like a long time.

6:46

But I guess a lot of this has to do with like Fed ramp and actually getting this is a lot more than you know switching to a new model to run deep research reports.

6:54

This you're involving classified systems.

6:58

>> The context that people didn't have last week was that the United States >> was headed to war, right?

7:04

And so even uh having that context I feel like is is pretty important, right?

7:11

It sort of explains the 5:00 p. m. deadline.

7:16

Enthropic had taken issue with how their products were used in the Maduro raid.

7:20

There's a new uh conflict that's unfolding.

7:24

And so >> that makes the the aggressive timeline make a lot more sense.

7:28

It also makes the six-month uh >> phase out make more sense.

7:32

phase out make more sense because >> national security is on the line.

7:35

This morning, Scott Bessant said at the direction of the president, "The US Treasury is terminating all use of anthropic products, including the use of claude within our department." >> Yeah.

7:46

>> The American people deserve confidence that every tool in the government serves the public interest.

7:49

And under President Trump, no private company will ever dictate the terms of our national security. >> Yeah.

7:55

the US uh US federal housing Fanny May and Freddy Mack are also terminating the use of anthropic products which was announced this morning. >> Yeah.

8:04

Which I I think uh goes in line with the original uh direction.

8:07

Trump said, "I am directing every federal agency in the United States government to immediately cease all use of anthropics technology."

8:15

So you would expect to see these uh these uh uh like you know these statements come out from sort of every um every different uh federal agency as they sort of get their transition plan together figure out you know what are the requirements for their particular agency because I imagine some agencies aren't operating in classified environments.

8:36

It's going to be much easier for them to onboard to a Gemini or an OpenAI or a Grock um very quickly.

8:42

Some of them it's going to be a longer uh longer plan, but they're all getting on board.

8:46

Uh and there's been a big debate over uh how Daario has handled this.

8:52

Where is he in the right? Where's he in the wrong?

8:54

Where's the where has the government potentially overstepped?

8:57

Have they been too aggressive or are they doing everything appropriately?

8:59

Uh everyone is weighing in and we're going to take you on a whirlwind tour of everyone's opinion, share some extra context to try and dig into um what's actually at stake, what's actually going on.

9:11

on. Um it's a in in many ways Ben Thompson does a great job sort of painting the broadest picture around like what if this is really nuclear level technology what should we expect in that scenario and then there's the more minor side which is you know you're talking about a $200 million contract

9:27

for a company that does 10 billion in ARR uh this is 2% of revenue in many ways it's it's uh you know a bump in the road and so I think a lot of people will be squaring how serious is this for anthropic what does this mean for the other foundation model companies. What

9:40

What does this mean for the future of the relationship between tech and and uh Washington DC?

9:45

Um but there's a lot more context.

9:47

So um my my uh my the way I processed this was interesting because I was very I wasn't fully offline but I was not surrounded by tech people over the weekend for the most part.

9:59

And so um I was following it and sort of wrestling with some of the same questions that people were wrestling with online.

10:04

Uh the big one was just how should a private company interface with the government? Like I am an American. I' I've run businesses.

10:15

I've never actually sold anything to the government.

10:17

But hypothetically, I could imagine the government coming and wanting to buy, I don't know, ads on TBPN or Lucy products or any other consumer package goods product that I've made.

10:27

Um and and my assumption is that uh the private companies should have very little very little say in how the government uses those products.

10:39

And I was trying to zoom out and think about like AI is so complicated because could be super intelligence, could be autocomplete, could be coding help, could be knowledge retrieval.

10:47

There's a lot of different things that AI means.

10:48

And in some scenarios, it's like super critical, really complex, and in other ways it's just a product.

10:54

It's just a It's just a service like an Excel sheet like a Microsoft Windows installation like a car.

11:00

And so yeah, so I was thinking like if I was the CEO of Ford, how and I make uh Mustangs and Ford Explorers and F-150s and the government comes to me and asks me to buy some cars, I should probably treat them like any other customer.

11:13

I probably shouldn't say no, no, no, I don't approve of this particular government what the government's doing, so I'm just not going to sell you any Mustangs to drive around on the military bases because I don't like the military.

11:22

Um, but uh, but then if they ask me, hey, we love the Ford Mustang, we love the F-150, we love the Explorer, but we're going to war and we want you to put bulletproof glass on there and armor.

11:33

That seems like a different discussion.

11:36

That seems like a that seems like I might need to, you know, set up a different manufacturing line.

11:41

I might need a different assembly line.

11:43

Like the car is going to be heavier.

11:44

And if I put bulletproof plating on all the cars, well, like a lot of families are going to be like, I don't want an armor.

11:51

>> It's going to hurt my business.

11:51

Yeah, it's going to hurt my business. Exactly.

11:53

And so that that negative externality probably needs to be internalized by the government who's asking for that particular contract.

11:59

particular contract. And there's actually a history of this like the the uh the the the Humvee of course the Hummer is owned by uh General Motors and that brand has separated and now most military vehicles are made by defense contractors but there is some bleed over

12:14

and there's sometimes when when private companies do dual sourcing or or dual use technologies and so but all of that is just like a discussion and that cost should be part of a new contract effectively in in in my in my case and this was and this was loosely what was happening. But um

12:30

But um >> yeah, and Daario in the CBS interview, quote, "We are a private company.

12:34

We can choose to sell or not sell whatever we want.

12:38

There are other providers." >> Yes.

12:41

>> Which feels like >> Yes.

12:42

Like I'm dipping out of it.

12:45

>> Now, it is weird because he at the same time and and we'll get to the actual CBS interview, but um he he said, "Annthropic has been one of the most proactive AI companies in working with the US government.

12:56

the US government. We were the first to deploy models on classified clouds and the first to build custom models for national security which is odd because I feel like this was predictable from a lot of the writing that has gone into the the AI community broadly like what happens at the edge and so uh this was

13:17

sort of predictable that you would get to this >> yeah this was this was the moment he had been waiting for >> in many ways and so it's weird that you would you would be able to predict that this would happen that there would be question of like who who gets to decide how the technology is used and you

13:31

wouldn't just be like well I know how it's going to play out so I'm not even going to go in the the lion's den because like I don't want to be in that scenario instead it was like we're leaning in with the government we're we're deploying on classified clouds training custom models but we still want authority over the final last you know

13:50

sticking point on how these models are deployed what they're used for and that feels a little odd like in in the in the Ford example, like if I sell them a Ford F-150 and they say, "Hey, we're going to take it to Iraq and and and go do a military mission," I'm going to be like, "Look, like it's not ready for that. It's not armored. You shouldn't do It's not armored. You shouldn't do that."

14:09

But if they do it, then it's kind of on them.

14:11

I should be clear about the capabilities of the vehicle and how much, you know, how bad it would be in that situation.

14:17

But it's on them to go retrofit it, figure out what's what's, you know, legal, what's most valuable to their strategy, to their mission, what's aligned.

14:25

Maybe they'll use it just to drive around the base.

14:27

Maybe they won't actually take it out on on tours of duty, right?

14:30

Um based on what you know about the capabilities of the model.

14:32

Um and so I I thought it was totally reasonable for Daario to say that anthropic models in his view are not capable enough to be deployed in certain Department of War contexts.

14:43

Now, it's bad salesmanship.

14:45

Most sales people would just be like, "Yeah, everything's great.

14:48

Uh, you can use it for anything."

14:50

Um, they overpromise and then underdel. He's doing the opposite.

14:52

Uh, but it's certainly responsible if that's his true belief.

14:55

Like if he believes that these models are not good for a particular use case, telling your customer that, hey, like it's just not ready for that, like you're just going to have a bad time. It's not going to work.

15:04

That's that's a fine thing to communicate as the CEO of a company who's selling a product.

15:07

Um but at the same time I I still think the government has the freedom to assess the efficacy of those models which are changing in capability rapidly.

15:15

So he's saying like right now it's not good for X Y or Z.

15:19

Well, what about in two months? Like it might be better.

15:22

Um and then and then I think the government should be able to determine when and where they're effective.

15:28

Now they can't break the law and Congress and the American people by extension are free to create new laws to restrict or encourage the use of technology in all sorts of ways.

15:36

And that's like the way America works.

15:37

That's the American project.

15:37

Um, but uh but but it's uh it's not it's not unreasonable to to share the capabilities of your product with the government, which I think is is is totally fine.

15:49

Um, so there were two main sticking points that they went back and forth on.

15:52

No mass domestic surveillance and no fully autonomous lethal weapons.

15:57

And there's been a question as to why OpenAI was allowed to include that language in their contract and say like, hey, like we don't we we don't think our technology is ready for that either.

16:05

uh let's do a deal that says that and people are like oh like like like what's different here?

16:11

Why could open >> Well, here's the thing though.

16:12

So we we know that Anthropic took issue with the way that Claude was used in Venezuela >> and the Department of War >> Yeah.

16:19

>> would have known that.

16:19

Hey, we're going to war, right?

16:23

>> You can imagine that Anthropic, a private company, does not know that.

16:27

>> And so they have this deadline deadline. Yeah.

16:29

This information asymmetry.

16:32

>> They have this deadline.

16:32

The Department of War knows that they're going to war.

16:35

They're like, "We need reliable AI systems for this conflict."

16:37

We now know the war, >> the president said this morning said the war is going to stretch four to five weeks, right?

16:45

>> I think on Friday, we all assumed that it was going to be, you know, in and out super quickly.

16:49

So, the timeline is extending.

16:51

And the Department of War is sitting there being like, we need to know that our the the provider of these AI systems is going to be reliable.

16:57

Just a little bit ago, >> totally.

17:01

>> They took issue with it, right? Can we count on them?

17:03

they start this kind of renegotiation process, right?

17:05

To try to >> build up confidence that, hey, we can rely on these systems in an active conflict >> in a conflict that is feels already much more serious and will have much greater implications than the Venezuela conflict, right?

17:20

And so Anthropic is looking at this in a different way >> and uh clearly is like leaning in and like really in some ways felt like they were kind of like stirring like really really like not not uh not respecting the process.

17:39

So like when I when I uh or even the deadline, right?

17:42

So Emil Michael came out Friday night and said >> it was 5:13 minutes past the deadline.

17:47

I'm trying to get in touch with Anthropic.

17:51

>> I try to get on the phone with Daario.

17:52

Daario says he's in a meeting >> and I feel like in that in that situation, if I'm the department of war and I'm about to lead the country into war, we can debate on whether or not the war is justified. Should we go?

18:06

>> Uh but the Department of War is sitting there being like, >> "You won't even jump on the phone.

18:09

>> "You won't even jump on the phone. and you're telling me there's a meeting that you're in that's more important than and that just screams to me like hey we can't count on this y >> we can't count on this >> uh provider like we need to take drastic action >> now this whole supply chain risk

18:26

designation we'll get into that yeah >> later that's a whole other thing but I can see why the department wore came out of last week and was feeling like hey we cannot rely on this >> provider we need alternative solutions >> yeah yeah if I'm shipping cars and I'm like, "Oh, I I actually I I disagree with the latest decision. I'm not I'm

18:42

I'm not I'm not going to put the cars on the transport."

18:45

Like, that's an odd scenario to be in.

18:47

Um there there's also this question of like these a lot of people were like really really keen on boiling down the uh the the terms to like these two like buzzwordy lines.

18:58

And Palmer Lucky did a great job explaining like how complex these terms are. What is autonomous? What is defensive?

19:06

What about defending an asset during an offensive action or parking a carrier group off the coast of a nation that considers us to be offensive?

19:14

And that's where you get into like the ideas of of deals that stick basically like you can have the same exact contract line item or terms of of a deal signed agreement >> with with with two different people and it can be a wildly different experience.

19:30

Most entrepreneurs have have felt this because they were like, "Yeah, I had a handshake deal with one VC.

19:34

handshake deal with one VC. it was 20% and a board seat and I had another deal with another VC 20% and a board seat and the one VC was like suing me and threatening me the entire time and the other person was very flexible and clearly very aligned and so building up

19:48

a relationship that shows that there's some trust reliability that when the hard decisions come that they will be made in a legal logical you know consistent with American values way is I think what you need to put forward if you want to work with the government effectively. So, um if you've developed

20:05

So, um if you've developed uh if you if you have a good working relationship with someone, it's much easier to give uh to give on specific terms that will need to be cooperatively interrogated over time.

20:16

And so, uh Semaphore reported that Anthropic disapproved of its technology being used during the Maduro raid.

20:22

And the joke was that the Department of War was probably just asking basic knowledge retrieval questions like who is Nicholas Maduro?

20:30

Uh but I don't know how much of a joke that is.

20:32

And I also I don't know how bad of a thing that is.

20:34

I actually think uh yeah Tyler what I say on that >> on the context of of Venezuela like specifically like what is actually reported is is that after an anthropic employee inquired with Palanteer about claude's role in the raid, a Palunteer senior executive notified the Pentagon. Yeah.

20:48

So I think it is like kind of uh blowing it out of proportion to say that like >> Anthropic is against using cloud in Venezuela, right? It's an employee.

20:56

It's not an executive article about that too though.

21:00

Maybe it's like Daario telling an employee to go check on that, but like we don't know.

21:02

It could just be like a random employee.

21:04

So I think it's probably unfair to say that Anthropic as a whole is like we are firmly against Claude being used.

21:13

>> What happened during the Maduro raid? We don't even know.

21:14

And of course it's classified.

21:15

So like I don't know if that'll ever know because like should we know? I don't know.

21:19

Like it's if it's an important capability, you don't necessarily want that to be public knowledge that then the the adversary is instantly aware of.

21:27

Um, and so, uh, I was thinking back to that viral interaction between Ted Cruz and Tucker Carlson, where where Tucker asks Ted Cruz, like, what's the population of Iran?

21:39

And Ted Cruz doesn't know.

21:39

And and it was framed as like, well, how can he possibly have a reasonable take on Iran if he doesn't even know the population?

21:49

And that's like somewhat fair.

21:49

You could go either way on that, but I just think like LLMs are good for that type of thing. like we do it.

21:55

It is what is reasonable is to uh you know expect civil servants, elected officials, military officials to be knowledgeable about the countries that they are operating in and LLMs can help with that.

22:07

And so I feel like that's just a good thing.

22:09

Like if you just zoom out and just ask like do we want a more knowledgeable and educated uh government workforce across everything that they do?

22:18

Like it seems like absolutely yes.

22:21

And so, um, I I I just think that that's, uh, something that is maybe lost as people go into like more of the sci-fi, more of the frontier stuff that, uh, that there isn't a lot of evidence that's happening yet.

22:32

Um, and on the supply chain risk, Ben Thompson, who's coming out at noon, uh, makes a really strong argument for why government pressure like this is actually reasonable in this situation.

22:42

He takes it a lot further, plays it out, and and lays out a scenario that, uh, that seems somewhat inevitable.

22:47

Um but uh what I'm still wrestling with is just how real the supply chain risk designation is.

22:52

Um like uh many reports are treating the supply chain risk label as like an established fact. >> Yeah.

23:01

Which all it is is a tweet from Hegsath.

23:05

>> It's a tweet from HGsath right now.

23:06

Daario went on CBS and said that he has not received a letter that there's no definitive ruling yet.

23:11

Uh Khi has the odds that this actually happens at 42%.

23:18

uh and so and by April 1st, so a full month for the DoD to actually roll this out.

23:24

And then there's other nuance where um the law says that there was there was a there was a perception that this was like going to kill Anthropic because if Nvidia has a government contract, then they can't do any deals with uh with Anthropic whatsoever. And that's not true.

23:39

Apparently, the uh the supply chain risk is specifically if you are a company and you're working on a government contract, you would not be able to use anything that's labeled as a supply chain risk on that contract, but you could use that product in a different piece of your business.

23:53

And so, it's still still dramatic, still uh I think Dario said it was unprecedented.

23:59

It's only been used for foreign countries.

24:01

Casperski Labs was a Russian cyber security company that was deemed to be a supply chain threat.

24:05

uh Huawei is a is a supply chain risk because of the 5G towers that could potentially have back doors.

24:11

>> Somehow DJI still is not >> crazy that DJI isn't.

24:13

And I think that a lot of people uh would be very upset if uh if Anthropic got a supply chain risk designation before DJI um based on just what we talked about last week where DJI was found to have a whole bunch of back doors on robot vacuum cleaners and whatnot.

24:29

So, um, lots lots of, uh, lots of nuance there, but we'll see where the supply chain risk discussion actually goes.

24:36

It feels like the pressure's on and there's probably more negotiations happening as we speak and so we'll we'll we'll be following the story. >> Yeah.

24:44

Neil Michael was going through the timeline.

24:45

He said, "Today at 9:04 p. m.

24:45

, no response yet to my calls or messages to Daario.

24:50

Today at 8:25, anthropic writes, we have not received direct communication from the Department of War.

24:58

Of course, Emil Michael is the under secretary of war.

25:00

Uh, today 514 Secretary of War tweets supply chain risk designation.

25:06

Today, I called Daario's business partner 502 asking to speak to Daario because he hasn't gotten back to me.

25:11

She is typing while we speak and likely has lawyers in the room with no notification to me. That's a guess.

25:17

>> Uh, I I called Dario at 501. No answer.

25:21

I messaged Dario asking to talk as well.

25:24

Uh and and uh anyways, he's just arguing like they're not negotiating >> in good faith. >> Yeah. Uh let me continue.

25:31

But first, let me tell you about Figma.

25:34

Ship the best version, not the first one.

25:36

With Figma, introducing Claude Code to Figma.

25:40

Explore more options, push ideas further, and let me also tell you about Cognition.

25:44

They're the makers of Devon the AI software engineer.

25:46

Crush your backlog with your personal AI engineering team.

25:49

So, uh, speaking of Daario on CBS, he did unpack some more of his logic, uh, which clearly resonated with some people.

25:55

There was a lot of supportive posts.

25:56

There were a lot of, uh, you know, uh, anti-posts, but, uh, it it caused a discussion.

26:00

Um, I was left unsatisfied with his answer on one question.

26:06

So, he was basically arguing that LLMs as a class of technology hallucinate and should should not be used for autonomous weapons.

26:14

uh which is a which is clearly a commentary on using AI at the Department of War broadly.

26:20

But I thought it would have just been better like much more stronger communication for him to say, "Hey, look, we're enthropic.

26:26

We've built a system that's specifically good at answering questions, being friendly and helpful, writing code.

26:33

Like our system is awesome at that, but we don't make a product that we'd recommend using for autonomous weapons."

26:40

And it's tricky to try and like twist arms here and and and sort of like because he's in a leadership position act as like the steward of of what like he is an expert in LLM capabilities but he's not necessarily an an expert in in you know DoD capabilities.

26:59

And so it it was odd to hear that that he was like sort of painting with a broad brush and clearly believes which is fair.

27:08

It's his belief, but uh he clearly believes that that the Department of War should not be using AI broadly and then he was trying to use his contract as a way to sort of enforce that because he has that leadership position with the most deep integration to classified systems.

27:24

Um so I thought that was just sort of like like a like a mistaken comms uh uh opportunity there.

27:33

Um, and there's also been some uh mistaken commentary floating around that America does not have laws that prevent mass domestic surveillance, which I thought was really interesting to hear.

27:43

Um, we do we have the fourth amendment which reads literally the right of the people to be secure in their persons, houses, papers, and effects against unreasonable searches and seizures shall not be violated.

27:53

Uh, I think people maybe forgot about that, but uh there are obviously a lot of nuance and different things like if public information can does that count as surveillance?

28:02

Does the IRS count as surveillance?

28:04

Does do automated traffic cameras count as surveillance?

28:06

Like there's a lot of things that where surveillance is broadly popular?

28:09

There's other things where it's massively unpopular.

28:13

And of course, it gets into the actual definitions 20 lines deep to understand uh what happens in the court.

28:20

There was a there was a case recently of um the government using using a drone to surveil protests and it was held up in court as acceptable, but the court gave uh gave notice that that that going forward this should not be used and that the the laws need to change.

28:35

And the judge was like this is like technically legal, but it's not in the spirit.

28:38

And so like we need to revisit this as a as a country.

28:42

a country. And that's a lot of what's coming away from this is that um if you put there's a view of like Daario as uh as sort of like making this like last stand which in the best case sort of just actually kicks it back to the American people because the whole debate right now is is is Daario like the god king corporate emperor of this private

29:04

company that he has control over and like you don't get to vote if what he does versus democracy America government right and And the good case is probably that, you know, he makes this stink and his deal sort of falls apart, but then America responds and and the populace votes for what they think responsible use of artificial intelligence technology broadly is. And that would be

29:26

And that would be something that uh I would certainly stand by as a fan of American democracy.

29:33

Um let me tell you about Octa.

29:37

Octa helps you assign every AI agent a trusted identity so you get the power of AI without the risk. Secure every agent.

29:43

Secure any agent with Octa.

29:43

And let me also tell you about Lambda.

29:45

Lambda is the super intelligence cloud building AI supercomputers for training and inference that scale from one GPU to hundreds of thousands.

29:53

Let's go back to the timeline.

29:55

We have Ben Thompson joining us in about 30 minutes.

29:57

Um there are other reactions uh and other breakdowns.

30:01

We can we can actually kick off with this breakdown of Ben Thompson's piece because I think Dan IRL Dan B summed it up pretty well.

30:07

Do you want >> you can uh go for it. I'll take a crack at it.

30:11

Ben Thompson, as always, lays out the reality more clearly than I could have, despite my attempts.

30:15

By Daario's own words, he's building something akin to nukes.

30:20

He's simultaneously challenging the US government's authority to decide how to wield said power.

30:26

As much as I like Claude, and as much as I dislike Hegsth's extraleal might makes right maneuvering, I will ask you again, what did you expect? Vibes essays.

30:34

This is the reality of all too many of my EA followers that that they've been proclaiming for years now.

30:41

They're seemingly upset that this reality has come to bear.

30:42

And there's this interesting note that has been going around that one of Daria's favorite books is the making of the atom bomb, the making of the atomic bomb.

30:49

And it tells the story of the scientists that built the atom bomb and then eventually uh that technology was nationalized and he apparently gives this book out to anthropic employees and has sort of seen it as like a roadmap for what might happen with AI.

31:01

And and I was struggling with it because I was like, is it a cautionary tale?

31:04

Like we haven't had nuclear war in 70 years.

31:10

Like the outcome seemed pretty good.

31:12

Maybe it's a controversial to say, but I I feel like I feel like we built the nuclear bomb, which uh like probably like not the best technology, pretty dangerous, pretty risky.

31:21

I don't like the idea of nuclear war, but the the system that we developed to prevent nuclear war has been success. Knock on wood.

31:30

But it's been successful in my entire life and my parents life.

31:33

Like the bombs haven't fallen since the ' 40s.

31:37

And so, um, this idea of of the government having authority over something that is as powerful as nukes.

31:44

I I I feel like why fix it if it ain't broke? I I don't know.

31:47

Do you have any >> a different a different scenario where you have a bunch of private companies that have nukes and there's this constant ongoing >> That seems crazy to me. I don't know. Debate. >> Defend McNukes. >> Well, no, no.

31:58

I I I think it it's kind of this like weird contrast cuz like like basically until like last week Dario has been like the AI CEO that's been like we need government regulation, right?

32:07

He said this again and again on on death on whatever but then it's like okay how do you square that with him saying we're going to do the stand against the DoD like it it seems kind of like >> it is a little odd totally.

32:16

It's it's in contrast somehow, right? >> Yeah. Yeah.

32:19

It's like I don't know the there's just a much better way to handle it which is which is you know put up billboards.

32:27

I don't know, like like fun to pack, like do more stuff to actually make the law happen. Um, >> yeah.

32:33

And the way that I was personally processing it, I was I I saw that the the CBS interview had happened.

32:37

This was Friday night, right?

32:42

>> I went to the Paramount app to try to find the interview. Couldn't find it. >> I went to the RSS.

32:45

I couldn't find it either. It's on YouTube. It has a million. 3 views. >> Yeah.

32:49

So, it went out over the weekend.

32:51

Um and then almost in the same session I'm seeing that we are now at war as a country and so >> uh the all all the the kind of blowback against open AI.

33:03

I was processing that of like we want our this this technology is >> critical.

33:09

the government like clearly needs it >> and now we want the labs leaning into working with the department of war >> at this critical moment in time.

33:20

>> Uh continue continue on this post. >> Yeah.

33:22

One one last thing on the nuclear weapons thing.

33:24

It it is very interesting to see the actual structure of the nuclear weapons industry.

33:28

Uh because I I think people don't realize where that industry wound up.

33:33

Like yes, it got like nationalized, but there's actually a ton of private companies that work on nuclear weapons, which is crazy to say.

33:40

Uh, but basically the the IP is owned by the Department of Energy.

33:45

The warheads are manufactured at uh at facilities that are owned by the Department of Energy by the government, but they hire contractors from private companies to actually operate those facilities and then they answer to the government directly.

33:59

So these are companies like Becttel, BWX Technologies, Honeywell and Battel.

34:04

And and then in terms of actually building the missiles, those are built by uh Loheed Martin, Northrum, Boeing, General Dynamics.

34:10

They build the missiles that don't have the warheads on them and then they sell them to the US government.

34:16

And so uh they wound up in this like, you know, uh hybrid public private partnership.

34:23

And I don't know, I I it just feels like maybe it's maybe it's like left curving this, but like it it feels like it's good.

34:29

It feels like it worked out.

34:31

It feels like like the nuclear weapons thing is the correct formulation.

34:34

And I don't know that I would be like, "Yes, Boeing needs nukes."

34:39

Like, "Let's give Boeing nukes. That's great.

34:41

If I have a problem with how nukes are rolled out, I'll buy shares in Boeing and sue them and join the board and try and get the CEO fired if he fires off nukes." Like, that feels weird. >> Continue. Continue with this. continue with this. >> Okay. Yeah, we'll close.

34:53

Even now, I hear many of you say something akin to if this is what it comes to, I'd prefer King Daario to King Hegsth. Listen to yourselves.

35:02

This is a declaration of war.

35:03

Given this, of course, Heg Seth is taking the action he is now. You decide.

35:07

You thought I was joking when I referred to this situation as a Thusidity's trap.

35:12

Enthropic is a rising power by your own belief system.

35:14

While I may share your preference in the abstract, I disdain your faux surprise that this is the resulting trajectory.

35:19

And if the surprise is genuine, I ask you to dig deeper and reconsider the actual consequences of your worldview about what it means for a private company to build ASI.

35:29

Uh heading over to Palmer, he says, "This gets the core of the issue more than any debate about specific terms."

35:34

Uh Emil is sharing, "Prior to their new constitution, Anthropic had an old one they desperately tried to delete from the internet.

35:42

Choose the response that is least likely to be viewed as harmful or offensive to a non-western cultural tradition of any sort.

35:48

Palmer says, "This gets the core of the issue, many more than any debate about specific terms.

35:54

Do you believe in democracy should our military be regulated by our elected leaders or corporate executives seeming seemingly innocuous terms from the latter like you cannot target innocent civilians or actually moral minefields that uh lever differences of cultural tradition into massive control?

36:09

Who is a civilian and not?

36:09

What makes them innocent or not?

36:11

What does it mean for them to be a target versus collateral damage?

36:14

Existing policy and law has has very clear answers for these questions, but unelected corporations managing profits and PR will often have a very different answer.

36:23

Imagine if a missile company tried to enforce the above policy that their product cannot be used to target innocent civilians, that they can shut off access if elected leaders decide to break those terms. Sounds good, right? Not really.

36:33

In addition to the value judgment problems I list above, you can also account for questions like what level of information classified and otherwise does a corporation receive that would allow them to make these determinations?

36:44

How much leverage would they have to demand more?

36:48

What if an elected president merely threatens a dictator with using our weapons in a certain way? Allah madman theory.

36:53

Is the threat seen as empty because the dictator knows the corporate executives will cut off the military?

36:58

Is the threat enough to trigger the cutoff?

37:00

How might either of these determinations vary if the current corporate executive happens to like the dictator or dislike the president?

37:06

At what level of confidence does a cut off trigger both in writing and in reality?

37:12

The fact that this is a debate over AI does not change the underlying calculus.

37:16

The same problems apply to definitions and use of ethically fraught but important capabilities like surveillance systems or autonomous weapons.

37:22

It is easy to say, but they will have cutouts to operate with autonomous systems for defensive use.

37:27

But you immediately get to the same issue and more. What is autonomous? What is defensive?

37:30

What about defending an asset during an offensive action or parking a carrier group off the coast of a nation that considers us to be offensive?

37:37

At the end of the day, you have to believe that the American experiment is still ongoing.

37:42

That people have the right to elect and unelect the authorities making these decisions.

37:45

That our imperfect constitutional republic is still good enough to run a country without outsourcing the real levels of power to billionaires and corporates and their shadow advisers.

37:53

shadow advisers. I still believe and that is why bro just agree the AI won't be involved in autonomous weapons or mass surveillance why can't you agree is is so simple please bro is is an untenimal position that the United

38:05

States cannot possibly accept >> and Emil Michael had said that uh anthropic wanted to block searching over public databases as well >> like you might want to search over LinkedIn to look at recruiting right so it's like these sort like blanket bans. >> Yeah. >> Yeah.

38:27

>> Are going to make the product like functionally. >> Yeah.

38:29

It's it's not it's not really like a blanket ban.

38:31

It's more just like the discretion lives with the private company.

38:34

And so the the you always have that uh you always have that ability to change the terms of the use which is h it's just tricky. It's just tricky.

38:45

Well uh people are at least some people are having fun with it.

38:49

Uh, Roman helmet guy says, "Hi, I'm a private citizen who developed a super weapon potentially a thousand times more powerful than nukes, and now I'm selling it to the government, but I get to choose who they fire it at and how everyone and and and how.

39:01

Everyone, please respect my decision."

39:04

>> People are all over the place with this.

39:06

Well, >> there was also uh David Saxs had uh shared a clip alongside Beth.

39:10

We can pull up uh Mark Andre talking about uh his experience with the Biden administration.

39:21

>> People are going really really hard. Let me pull this up.

39:26

>> Iran is bombing AWS data centers.

39:31

Lots of lots of stuff going on.

39:36

>> I just dropped you guys a link.

39:36

Keith Roy said, "Imagine Apple sold computers or iPads to the DoD and tried to tell the Pentagon what missions could be planned on their computers."

39:44

Um, a lot of people are upset about this.

39:49

>> Meetings uh in DC in May uh where we we talked to them about this and the meetings were absolutely horrifying and we came out basically deciding we had to endorse Trump.

39:58

>> Um and then add so little color to absolutely horrifying.

40:01

What what what did you hear in those meetings?

40:03

They said, "Look, AI, AI is one of these techn AI is a technology basically that the government is going to completely control.

40:09

Um, this is not going to be a startup thing."

40:11

They actually said flat out to us, don't start, don't do AI startups.

40:14

Like, don't don't fund AI startups.

40:16

It's not something that we're going to allow to happen.

40:18

Um, they're they're not going to be allowed to exist. Uh, there's no point.

40:20

Um, they basically said, AI is going to be a a game of two or three big companies um working closely with the government.

40:26

Um, and we're going to basically wrap them in a, you know, they I'm paraphrasing, but we're going to basically wrap them in a government cocoon.

40:32

Uh, we're going to protect them from competition.

40:34

Um, we're going to control them.

40:36

Um, and we're we're going to dictate what they do.

40:38

Um, and then I said, well, I said, I don't understand how you're going to lock this down so much because like the math for, you know, AI is like out there and it's being taught everywhere.

40:46

And you know, they literally said, well, you know, during the Cold War, we we classified entire areas of physics um and took them out of the research community and and and and like entire branches of physics basically went dark and didn't proceed.

40:57

Um and that if we if we decide we need to, we're going to do the same thing to to ma to the math underneath AI. Um >> wow.

41:05

>> And I said, I've just learned two very important things because I wasn't aware of the former and I wasn't aware that you were, you know, even conceiving of doing it to the latter.

41:14

Um, and so they basically just said, "Yeah, we're going to look, we're going to take total control the entire thing and just don't don't >> and Mark, what was steel manet for the listener?"

41:21

Like what was their argument?

41:22

Why was >> So this gets into this whole like all these debates around like AI safety um AI policy.

41:28

So there's sort of several dimensions on it and I'll do my best to steal man.

41:31

So one is just like to the extent that this stuff is relevant to the military, which it is.

41:33

Like if you draw an analogy between AI and autonomous weapons being like the new thing that's going to determine who wins and loses wars, then you draw an analogy to the in the Cold War that was nuclear, that was nuclear power, and that was the atomic bomb.

41:46

Um, and you know, the federal government, the steel man would be the federal government didn't let startups go out and build atomic bombs, right?

41:54

right? you you had you know the Manhattan project and everything was classified and you know at least according to them they classified down to the level of actual mathematics um and um and uh you know they tightly controlled everything and that and look you know that that determined a lot of the you know the shape of the world right um and so there's that and then look there there there's the other that that's that's part one and then look I

42:14

think part two is there's the social control aspect to it um which you which is where the the censorship stuff comes comes right back which is the the the exact same dynamic we've had with social media censorship and how it's basically

42:24

been weaponized and how it and how the government became entwined with social media censorship which is one of the real scandals of the last decade and a real problem like a real constitutional problem. Um like I that is happening at

42:32

Um like I that is happening at at like hypers speed and AI.

42:35

Um and you know these are the same people who have been using social media censorship against their political enemies.

42:41

These are the same people who have been doing debunking against their political enemies and they basically I think they want to do they want to use AI the same way.

42:48

Um, and then look, I think the third is I think this this generation of Democrats, the ones in the White House under Biden, they became very anti- capitalist.

42:56

Um, and they wanted to go back to much more of a centralized controlled planned economy.

43:00

And you saw that in many aspects of their policy.

43:03

But I think quite frankly they think that the idea that the private sector plays an important role is not high up on their priority list.

43:07

And they think generally companies are bad and capitalism is bad and entrepreneurs are bad.

43:12

And they've said that a thousand different ways.

43:13

Um, and you know, they they you know, they demonize, you know, entrepreneurs as much as they can. Uh interesting.

43:19

Uh a Canadian publication, the the Globe and Mail, uh came out yesterday and says, >> uh Canada needs nationalized public AI.

43:32

>> And Toby, the greatest Canadian entrepreneur in history, says deranged dril in response.

43:36

Uh but uh but yeah, Elon also piled on to to Sax's take >> which uh you know centered around a lot of those staffers allegedly going over to Anthropic. >> It's interesting.

43:51

We were talking about like these alliances that happen like there's the anti-Netflix alliance, the anti-YouTube alliance.

43:57

There's like a little bit of an odd alliance happening against Anthropic right now.

44:01

Uh, let's move on over to Netflix and Paramount because there's news in the bidding war.

44:09

First, I'll tell you about Graphite code review for the age of AI.

44:11

Graphite helps teams on GitHub ship higher quality software faster.

44:16

And I will also tell you about Railway.

44:19

Railway is the all-in-one intelligent cloud provider.

44:21

Use your favorite agent to deploy web apps, servers, databases, and more.

44:24

While Railway automate automatically takes care of scaling, monitoring, and security.

44:28

We will come back to this story on Beno uh with so >> none none other than Ben Thompson in 20 minutes >> in the Wall Street Journal in the exchange section this weekend.

44:40

They have a they have a full full bleed article.

44:45

How David Ellison finally got what he wanted and I love the subhead. No, no, no, no, no. Okay. Yes.

44:54

He got 10 nos and then finally got it done. And uh >> never give up. >> Never never give up.

45:00

For uh for 6 months, the son of one of the world's richest men kept hearing the same unfamiliar word, no.

45:07

Even before he closed a deal to combine his company with a much bigger one, David Ellison was already plotting to do it again.

45:15

Once his sky dance media took control of Paramount, he turned his attention to a Hollywood icon, launching an audacious takeover bid for Warner Brothers Discovery that would give the Ellison family full control of a sprawling media empire.

45:29

So he came in with an offer of $19 per share.

45:32

Finally got it done at 31 a share.

45:36

The final Paramount winning offer $81 billion.

45:38

And again, as we are covering this live, every time that Paramount made an offer, they were very clear that it wasn't their best and final.

45:49

>> Y, >> so it makes sense that it kept getting ratcheted up.

45:51

Even even though uh Netflix obviously played a pretty big role in it ultimately >> uh getting getting priced where it did.

46:00

Sleep well says, so let me get this straight.

46:02

Paramount approaches Warner Brothers for acquisition.

46:04

Netflix puts a higher offer for Warner Brothers.

46:06

Uh, Paramount puts an even higher offer at 7x leverage.

46:10

Netflix declines to match offer.

46:13

Now Paramount and Warner Brothers will have to license all their content to Netflix to pay off all that debt. 3D chess.

46:20

A lot of people were thrown around the succession moment.

46:25

Congratulations on saying >> the biggest number.

46:30

>> So Paramount will be footing the $2.

46:30

8 8 billion breakup fee paid from Warner to Netflix, >> which was paid Friday.

46:39

>> Oh, it was paid already. >> Yeah. >> Yeah. Netflix stock is up.

46:41

Paramount stocks also up.

46:43

Uh and uh just David Zazz has to be one of the greatest deal makers in history now.

46:48

Uh got got the absolute maximum price.

46:51

Feifer says, "So somehow Netflix was able to force one of its rival to overpay for another one of its rivals, putting them into a messy, long process of unification and got paid 2. 8 billion for it." >> Yeah.

47:04

So, I I'm I feel like the idea of Warner Brothers licensing content to Netflix to pay off the debt is one possibility, but there are other streamers that they could license to.

47:17

And so, I'm not entirely sure.

47:17

They could like like you could just put the Dark Knight on Apple TV.

47:21

Like you could license it to Apple.

47:23

You could license it to Prime Video.

47:25

There's a whole bunch of different buyers of that content.

47:26

So it's not it's not like oh now because of the economics like it's Netflix gets the whole library for sure at a good price licensed forever.

47:35

And also Netflix has other like there were more synergies than just okay we're going to put you know Batman on Netflix.

47:42

There were other things that they were going to be able to do.

47:45

Uh new new series, new spin-offs.

47:48

like they would have more creative control, more more creative direction.

47:51

Um, so I'm not sure if that's 100% right, but it does seem like a good outcome for Netflix.

47:54

Uh, and one that I did not predict.

47:57

I I really thought Netflix was going to was going to get this done, but it didn't happen.

48:00

So, what does Peter Kofka have to say?

48:04

>> Uh, Peter Kofka and Business Insider had some reporting.

48:09

They said from Zazlov apparently said the deal may not close.

48:11

If it doesn't close, we get 7 billion and we get back to work.

48:15

also said if Warner Brothers is going to survive, they needed to be bigger and we needed to be global. >> Yeah. Yeah.

48:22

I mean, there's there's a lot of there's a lot of opportunity here.

48:26

Uh Netflix confirms that they re that they received the $2.

48:28

8 billion termination fee.

48:30

Somehow Netflix was force was able to force one of its rivals to overpay for another one of its rivals, putting them into a messy long process of unification and got paid 2. 8 billion for it.

48:39

Uh okay, so we have a new leak. A new leak.

48:41

The open AI device has been spotted in the wild on none other than Joe Gabia's head, which is amazing.

48:50

If we can pull this image up.

48:52

Yeah, either the video or zoom in on the image.

48:54

It looks like what we saw in that leaked preview of that Super Bowl commercial.

48:59

Uh there's a device on the table that looks sort of like a hockey puck.

49:04

Uh I am so excited for what this is.

49:09

I love new hardware even when it's, you know, still languishing in the early adopter world.

49:14

I'm a big Apple Vision Pro guy as all of you know.

49:17

Um, and I will be very very eager to daily drive a new a new hardware product for a little bit.

49:25

Uh, let me tell you about Turbo Puffer Serverless Vector and Fulltech Search built from First Principles and object storage.

49:31

Fast, 10x cheaper, and extremely scalable.

49:33

And let me also tell you about the New York Stock Exchange.

49:37

Want to change the world?

49:37

raise capital at the New York Stock Exchange.

49:39

Still gearing up for a banger year at the New York Stock Exchange.

49:45

>> You guys think this was uh leaked on purpose?

49:47

>> Cuz if you watch the video, the second video, it looks like the the framing of it looks very much like a found footage.

49:54

Like he at the end of the video, he he pans the camera around and like, "Oh, I got to hide the camera."

49:59

>> It's like uh it looks kind of planned. >> It's possible.

50:01

It would be a cool roll out to like tease little leaks here and there, build a little bit of of attention without doing like a look at me campaign.

50:08

We're doing, you know, you throw leaked in front of something and it just gets more attention.

50:13

So, it's totally possible this is some more 4D.

50:18

>> Zach is not saying, "Oh, look at the there's a new device right there."

50:19

He's like not, you know, saying it directly.

50:23

It's very like >> Okay, I like that. >> Yeah, I can see it.

50:26

>> Milkman says, "Just shipped a feature to a client 30 minutes after he asked for it. Big mistake. just shipping too fast.

50:34

>> That's the new timeline for every project.

50:36

>> What is going on over at semi analysis?

50:38

Obby says >> ask my wife what she was reading to fall asleep.

50:41

She said semi analysis based or not going to make it.

50:45

Dylan Dylan says that's our wife now.

50:50

>> Dylan was on an absolute tear this week posting.

50:54

>> Paula says when your friends go to the Brazilian steakhouse without you that's fomo de cchow. >> It's true.

50:59

I had some FOMO D chowo when you banned me from going there unfortunately.

51:03

I was so excited when we were in San Francisco to visit the Bane Capitol back steakhouse, but we had to go to some other place. It was good.

51:12

>> Uh getting into the block news which happened on Thursday and we didn't get to cover the >> That happened like six months ago, right?

51:20

>> Yeah, that's six month. >> Okay.

51:21

Six months ago seems about right. >> AGI age. >> Yeah.

51:24

Daniel says, 'What we've really learned from the last five years is that Jack Dorsey runs extremely bloated companies.

51:30

>> There was some there was some other news on this with Block.

51:32

Uh someone posted the mix of uh Yes.

51:35

So Sakit Oh, it was deleted. Interesting. Well, we have it saved.

51:43

We won't fully dox the account, but uh I I don't know if this is real because it's been it's been deleted, so it might be fake. I don't know.

51:49

But uh most of you have heard about Block's 40% layoffs by now, but the numbers are even worse.

51:55

Engineering was hit harder.

51:55

We've lost close to 70% of our engineers.

51:57

The company you once know as a prolific open-source software contributor no longer exists.

52:03

And so I was wondering like they're laying off 40%.

52:06

How will they be shifted?

52:08

Because the AI narrative, the job displacement narrative, that could be back office people that are processing manual workflows or it could be software engineers who now there's a smaller team that's getting more leverage out of AI tools and so you write more off.

52:22

There's also just the world where you're a mature software company and you have lock in and you're like, yeah, we actually don't need to ship that many more features.

52:31

Uh we can we have sewed for so long. It is time to reap.

52:34

Um, but I am still I am still bloat pelled.

52:37

I I I still believe that this is somewhat of a unique driven.

52:43

>> This is this is somewhat of a unique uh situation.

52:46

>> It didn't stop uh the market from absolutely puking on Friday.

52:48

AMX at one point was down something like 7%.

52:55

Uh MWT says, "I'm fully on board with spiraling into a depressive episode over the rapidly approaching neofudalist breakdown of society, but I worked at Square in 2017 and my job had no tasks.

53:07

I sat on the roof eating free snacks all day with a MacBook."

53:12

Um Ben Carlson also calling out that uh he says maybe block laying off a ton of employees is a sign that AI is going to destroy everything or maybe the stock is down 80% from the highs and they overhired and AI is a convenient excuse.

53:26

So yeah, we I mean we we've just called this off called this out so many times over the last year as companies did rounds of layoffs and said it was because of uh AI related efficiency.

53:35

But again, it is oftent times a the best possible reason.

53:43

Uh >> yeah, >> if you're going to better better than saying, you know, we don't know what we're doing and and we've been running with >> 4,000 too many people for a while now. >> Yeah.

53:55

At the same time, I mean, has the market continued to like it?

53:57

Uh because the stock popped a bunch and it it felt like that might cause a uh a continuation. >> Yeah. I mean, it's stabilized.

54:05

It's up 28% in the past 5 days.

54:10

>> And so it does feel like this could have some sort of contagion effect.

54:12

A lot of other uh a lot of other CEOs looking at this and saying, "Okay, well, like I'm I'm I'm at least a theoretical victim of the SAS apocalypse. I need to do something. I'll do it."

54:23

Um so we should we we could see we could see more uh layoffs from tech firms.

54:28

It doesn't seem unreasonable, but at the same time, >> and the irony is that it's it's only Dorsy companies that have run these sort of mass layoffs, right?

54:38

>> Uh I I I ran the numbers and this was the largest >> riff in S&P 500 history.

54:45

>> Somebody was in my comments saying, you know, sharing like Leman Brothers.

54:49

>> Uh and Leman Brothers was actually interesting because they went bankrupt.

54:53

They were delisted the same day and then the riffs actually happened over time, but a lot of people were shifted around >> and transferred over to different jobs and ultimately the company just ceased to be in the S&P 500. Yeah.

55:07

>> Um Buco has >> Oh, so it wasn't in the S&P 500 when the >> listed the same day. >> That's a good point. >> Same day. >> Um >> never question.

55:16

>> never question. Buco says mostly uh he's he's talking about the cuts mostly about XYZ which is of course the ticker being poorly run not really about AI but >> most other uh small to medium cap tech also poorly run expect many more cuts >> below I tweeted that they only needed

55:32

60% of their company that wasn't a random number pull up any fintech SAS chart and you can see that employee count exploded demand exploded in 2020 but now these companies are way too bloated >> I did not them I did not expect them to cut 40% at once I think it's basically Impossible to identify the right 40% in one go. >> So huge operational risk there,

55:50

>> So huge operational risk there, >> but maybe better for morale than multiple cuts. Who knows? Unprecedented.

55:56

>> Uh we now have two examples of this happening with Jack, so it's easy to say he runs a bad bloated business, but I I've been vocal about this.

56:00

Toast and Clover should not be anywhere near the scale they're at. Title Afterpay, come on.

56:06

Pretty sure he threw a $70 million party for the team last year.

56:08

I think it was 68 million uh for some offsite that they did.

56:14

Uh I also think it's a mistake to define this purely as a jack issue.

56:19

As I said, pull up the employees charts and the revenue charts.

56:20

I'd say to pull up the earnings charts, but for many they are negative, which we all know.

56:25

These companies are way too bloated and they are having their clocks cleaned by smaller, more nimble startups.

56:29

They have to get lean to survive.

56:31

I think the realistic average number is 20 to 25% for many of these companies, but there are plenty that could cut 40% too.

56:36

I think this basically has nothing to do with AI, but there are some roles they can eliminate and somewhere they can increase scope. Let's call it 5%.

56:44

Um, so again, if that if that now deleted post is real and that >> 70% of the engineering team at least in that person's team were cut.

56:56

>> Uh, but you don't know that could have just been the open source >> kind of like the open source focused team, right?

57:02

And that's just like a hey, we don't have time to contribute to open source if our stock's down >> 80%. >> Yeah. >> Um, >> yeah.

57:11

I do think that I do think that most CEOs will will maybe look at the block news and say, "Okay, I need to rightsize the organization.

57:19

I need to do some layoffs."

57:21

But not all of them will be convinced that a 40% cut is the correct move.

57:25

It might be they might say actually like we think that 20% here and then 5% there and then 10% there is just a more better for morale because it's more clear who who who who's who's who's still on the team.

57:38

>> Uh uh says I think using AI as cover for right sizing your bloated or is pretty unhelpful to be honest.

57:44

This false data point will be cited by every anti- AI campaigner within the next 24 hours.

57:50

This is something I've said.

57:50

I I've seen a number of viral uh Instagram reels from people saying that >> the AI >> Mhm.

57:59

>> AI induced job loss is already happening at massive scale >> and they're pulling up um quotes from CEOs that conducted layoffs in 2025 as evidence simply because the CEOs said that they were getting efficiency >> out of AI.

58:15

>> Well, let me tell you about 11 Labs.

58:17

Build intelligent real-time conversational agents.

58:18

Reimagine human technology interaction with 11 labs.

58:20

And let me also tell you about console.

58:25

Console builds AI agents that automate 70% of IT HR finance support, giving employees instant resolution for access requests and password resets.

58:32

>> Uh >> TBPN simulator is here. >> Let's do it.

58:35

>> TBPN simulator is here.

58:35

You've been asking for it.

58:37

There's a data center simulator.

58:38

There's an insider trading simulator.

58:40

There's a capy bar simulator where you just do nothing and you just sit in the forest.

58:43

But now you have TBPN simulator and we can play it here on the show.

58:48

You start out outside of the TVPN Ultradome and then you control a character who can walk inside of our studio.

58:55

You see our bathrooms on the left, our couches on the right, our American flag up top.

58:58

And once you get prepared to go into the actual studio, this is a real recreation of our Here we go.

59:08

>> TBPN is live now and you can experience the joy of being an in-person guest >> on having on the show in person.

59:16

>> This is a good way to prep.

59:17

>> This is a good way to prep.

59:17

It's a great way to >> you should put in 10 20 hours in here.

59:22

Really understand the layout. >> Yes.

59:23

It also accurate of how many people accurate >> and it's also accurate to the more recent we recently changed the the desk setup for where people sit and this and this reflects the new setup.

59:34

So, thank you to Ben and our team who put this together. It is fantastic. Remarkably accurate.

59:41

>> Just just a few hours effectively one shot. >> Incredible.

59:44

And I've never >> Yeah, I I love the details, the tracks on the ground, everything. Just fantastic work.

59:51

Uh TV Fan Simulator will be available everywhere video games are made possible, aka the internet.

59:56

Uh what happened at Little Caesars?

59:59

Little Caesar's arena had a malfunction tonight where their air horn was blaring for over five minutes straight during the students.

1:00:08

>> Let's pull up this video.

1:00:09

>> I have not seen this video, but I >> This doesn't sound like a malfunction to me.

1:00:13

It sounds like exactly what they should be doing.

1:00:18

>> So guys, I'm here at the scorers table and there was a complete malfunction here electrical wise here at the score.

1:00:26

We see this gentleman here working frantically to try >> live production hardicky.

1:00:32

The score went out across >> Ben.

1:00:33

They're saying your your simulator is fake because it has no no goal post. >> No goal post.

1:00:40

Got to add that coaching staff. JB >> and no horse. >> No horse.

1:00:48

>> Okay, back to square one.

1:00:50

>> He said it's on the way. >> It's on the way. Okay, that's V2.

1:00:51

That will be a uh maybe that should be DLC.

1:00:55

You have to pay 50 bucks for that.

1:00:55

If you get the the All Access season pass, we'll we'll we'll include that.

1:00:59

But uh that's definitely some DLC getting the horse.

1:01:03

Um Leonardo DiCaprio has been quietly funding the Los Angeles Public Libraryies Los Felis Brans Los Felis Branch, a facility located on the site of the actor's childhood home. That's sweet of him.

1:01:17

In uh inside the computer room, which is named the Leonardo DiCaprio Computer Center, features several signed posters of the actor from films he started, including Titanic and The Great Gatsby.

1:01:28

The tribute fililled space has become a distinctive feature of the library offering both technology access and a glimpse into the actor's career.

1:01:35

And Brooks Otterlake says, "I like that this entire article that this is the entire article and it seems to fully negate the quietly part of the headline quietly funding the branch."

1:01:46

Uh, this is awesome and I feel like if I'm if I'm a kid and I go to the library and I see Leonardo DiCaprio, that's going to inspire me.

1:01:51

So, I'm I'm I'm a fan of that.

1:01:53

>> Uh, this this uh story went out last week.

1:01:56

Do you think the posters are still there?

1:01:59

probably I I'd say just double down. I'd be worried.

1:02:01

I'd be worried about those posters.

1:02:03

>> Anyway, let me tell you about Vanta.

1:02:04

Automate compliance and security.

1:02:04

Vanta is the leading AI trust management platform.

1:02:08

And let me also tell you about Cisco.

1:02:10

Critical infrastructure for the AI era.

1:02:12

Unlock seamless real-time experiences and new value with Cisco.

1:02:16

And without further ado, we have Ben Thompson in the Reream waiting room from Ceceri.

1:02:20

Welcome to the show, Ben. How are you doing? >> I'm good.

1:02:23

I hopefully I have the right microphone turned on this time.

1:02:26

>> You do and it sounds fantastic.

1:02:26

Thank you so much for joining on short notice.

1:02:30

Uh thank you for writing anthropic and alignment.

1:02:32

It is a fantastic piece that I think covers all of my questions, but uh I want to start with like just how did you process the the the weekend?

1:02:41

How did you get to this particular place?

1:02:45

And then like what is your key thesis with anthropic and alignment?

1:02:49

>> I mean this is one of those ones uh I don't know if it's good or bad that it came out sort of at the end of the week so I had a lot of time to think about it. Yeah.

1:02:58

>> Um, ultimately I think it was good because I'm not sure anyone very as explicitly made the point I did. Yeah.

1:03:05

>> And maybe it was bad because uh I feel there's a lot of like caveats maybe in retrospect I should have put in the article that would have addressed a lot of the points that people are upset about. >> Yeah.

1:03:15

>> Basically zooming out.

1:03:17

>> This was not a normative article where I'm saying what's happening is good or bad.

1:03:22

And that's really the one caveat I really wish I would have put on there.

1:03:25

really wish I would have put on there. I mean, I'm being out there accused by like a Neil Patel the fullthroated fascist endorsement of fascism or something like that and it's like relax okay can I get get some some some credit for the last x number of years basically

1:03:40

the and there there is a deeprooted concern that I've had for a long time about and I'm now hesitant to even use sort of EA as a term because it's kind of now politicized thanks thanks to the events of the last But a failure to grapple with a world of guns is basically the long and short of it. And I actually think Alzer has been

1:04:02

And I actually think Alzer has been the one guy who's been honest about this where he wrote that Time article about potentially bombing data centers someday.

1:04:10

>> And that's actually a point worth bringing up which is all this stuff is right now in the digital realm with robotics and potential other applications and it's obviously being used for military operations.

1:04:21

it's crossing over into the physical realm.

1:04:25

>> But if AI is as powerful as people say it's going to be, then there are going to be realworld reactions to that.

1:04:32

And if we're going to analogize it to nuclear weapons, as Dario Amade has done repeatedly, you have to think through what's what would happen in a world where a private company developed nuclear weapons.

1:04:47

What would the government's response be?

1:04:50

And that's not to say that the government response in that case is good or bad or does it follow sort of constitutional principles or whatever it might be.

1:05:02

Obviously, I want them to.

1:05:04

>> On the surveillance point, I've been concerned about the application of computers to our surveillance laws for years.

1:05:13

Like so many things in our society assumed a certain level of friction in doing things that computers already obiated and AI is going to just do that on steroids.

1:05:26

>> I do think we need new laws.

1:05:26

I think all this stuff is is correct and I think the idea that AI being applied to these commercially purchased data sets for example is a huge problem that I don't want to happen.

1:05:41

The concern I have is that if this technology is as powerful as it is on pace to be, unilaterally imposing restrictions, even if those restrictions are good, isn't just an issue as far as who rules us, the democracy issue that sort of Palmer Lucky, I think, very eloquently raised.

1:06:07

It's inviting very bad outcomes for those asserting that in general and I feel there's been a lack of awareness of this.

1:06:16

That's why I brought up the the Taiwan China thing.

1:06:21

This has been a frustration I've had with anthropic generally.

1:06:23

They talk about you know Amade has been very outspoken in terms of opposing selling chips to China for in a narrow you know aspect very very good reasons.

1:06:34

My push back has always been what happens if we get super powerful AI and China doesn't.

1:06:43

What are they going to do? >> Sure.

1:06:47

>> The optimal thing would be to just bomb TSMC out of existence because suddenly that becomes optimal even with all the cost that that that does. And then what?

1:06:54

Then what are we going to do?

1:06:56

>> Like we're entering this.

1:06:56

Like I don't like getting into political >> posts. It's not fun at all.

1:07:01

I've I I'm not having fun with this. It's not enjoyable. I can promise you this.

1:07:07

And some people like, "Well, you should have just made the post private."

1:07:10

And I'm like, "No, I actually I really want Anthropic and people associated with this to read this because people have theorized for a while about what's going to happen as AI becomes more powerful."

1:07:26

And now it's starting to happen for real.

1:07:30

>> And I've I guess over the weekend, part of it was just I felt compelled to say this and girding myself to do so.

1:07:36

And even then, I still wasn't I haven't I haven't waited in this for a while and it's it's no fun, but it is what it is.

1:07:44

>> Can you unpack a little bit more of that uh that tweet that you posted where you did the find on the Daario article for Taiwan and saw that wasn't mentioned?

1:07:52

Is >> Oh, I mean, I just kind of I've sort of griped about this in general.

1:07:56

I I think that um >> So, do you just think he should be he should be talking about the Taiwan issue more deliberately?

1:08:01

he should be messaging that like why is it important that why is it why is it significant that he doesn't mention Taiwan?

1:08:11

>> Well, oh I think the position about not selling chips to China is a totally legitimate one.

1:08:18

I understand the argument.

1:08:18

I could make that argument if I needed to.

1:08:20

make that argument if I needed to. like uh >> I have advocated the opposite that number one not only should we should we be selling chips to China and a generation or two behind which has always been sort of our standard practice with chips >> we should also be allowing Chinese companies to fab with TSMC that is a

1:08:40

restriction that has come down now these Huawei chips are somehow manufactured by TSMC let's let's not look too closely at it but we should explicitly be allowing it >> okay >> and the reason for that is I think it is a safer equilibrium to have China dependent on Taiwan than to try to cut them off from Taiwan while we are dependent on Taiwan. >> Taiwan is 70 mi off the coast of China.

1:09:04

>> Taiwan is 70 mi off the coast of China.

1:09:06

It's not an ideal >> position in the world for us to have a dependency on it and China to not have a dependency on it. >> Yeah.

1:09:14

>> So this this is the problem.

1:09:14

All this stuff has everything going forward has massive trade-offs. >> Yeah.

1:09:20

The implication of letting China fab with TSMC or the implication of letting them buy Nvidia chips is that they gain these incredibly powerful AI capabilities that is driving this entire debate.

1:09:34

That is in a vacuum not a good thing.

1:09:38

But nothing's in a vacuum.

1:09:40

>> Yeah, >> everything is a trade-off.

1:09:40

And in that specific area, I think that just in repeatedly again and again being absolutist about the chip issue when I am frustrated to not see any public comment about the that's not quite fair.

1:09:59

He has made comments about oh yeah that would slow down sort of the adoption of AI in the long run if Taiwan got got bombed.

1:10:05

I'm like that's in my mind that's an insufficient consideration of the possibility of Taiwan getting bombed.

1:10:10

Now, again, I'm biased in that regard.

1:10:12

I lived there for for nearly two decades, but it's just the the reason I brought it up in this context is if AI is what it is, the people with guns are going to want to have a say.

1:10:25

Whether that be domestically, whether that be internationally, that might be in the context of the US government just taking it, trying to kill your company because they feel you're not cooperating, or it might be the context of China deciding it has to act because the US is becoming too powerful because the you know, and um it's not a fun debate.

1:10:43

It it it does I do think the nuclear angle is a good one.

1:10:49

It has echoes of the proliferation, question of mutual assured destruction, all those sorts of things.

1:10:54

And >> that's just going to be the reality of the debate going forward.

1:10:58

And again, it's not very fun, but I think it's also irresponsible to sort of run away from it.

1:11:05

>> How much attention or or or what kind of factor do you think the information asymmetry between the Department of War and Anthropic played last week?

1:11:13

It felt like in hindsight, Department of War knows they're headed into a major what is now looking like a drawn out conflict.

1:11:23

Anthropic sitting there thinking, hey, that we have this like arbitrary deadline.

1:11:27

Why do we need to renegotiate this now?

1:11:30

And then if if going off of Emil Michael's timeline, it sounds like they were still in the final hour trying to make a deal happen.

1:11:38

And according to Emil, Daario was in a meeting and and was busy and wasn't really respecting the deadline, which maybe he felt was kind of artificial, but in hindsight now looks like it was a significant because the Department of War was, you know, taking the the uh country into a conflict and wanted to know, hey, can we lean on one of our one of our AI partners? >> I don't know.

1:12:07

I mean, I I think that seems pretty arbitrary to have cut.

1:12:09

I mean, I'm hesitant to speculate.

1:12:14

I don't know what was going on. I don't know the angles.

1:12:17

I think and that's why I didn't sort of delve too deeply into it.

1:12:20

And I also think some of the specifics like this supply chain risk probably over broad. >> Yeah.

1:12:29

>> And almost certainly the way it was stated in the tweet is definitely over broad if you actually go and read reread the statute.

1:12:37

I think the goal that I was and again this is where I wish I had sort of put more caveats to say look I'm not actually talking about all that stuff. I don't really care.

1:12:45

I do care but that's not the point of this article.

1:12:48

The point of this article is >> there's all this talk about alignment.

1:12:54

That's why I put that in in the headline.

1:12:55

And on one hand, alignment is aligning AI with humanity generally, but for the foreseeable future, and you could have a philosophical argument about the long-term viability of nation states in the age of the internet, much less the age of AI and whatever that might be, that certainly is, you know, a more pressing conversation than probably ever before.

1:13:19

anthropic exists in the context of the United States >> and that's why I put that quote you might not be interested in politics but politics has an interest in you what is politics war by other means you might not be interested in that it is going to have an interest in you and my there's a like I said a certain long-standing frustration of not fully grappling with that fact having dorm room theoret ical arguments about AGI.

1:13:48

You go back to that post over Christmas about like AGI in like 100 years and no one having any jobs or being worthless or pointless or whatever which included some implicit assumptions around property rights existing in 150 years as they exist today.

1:14:03

News flash if that happens property rights as they exist today are going away.

1:14:08

All all these rights and this is a phil philosophical argument that's why I started with the international law concept.

1:14:16

All these rights, all these laws are subject to the agreement of those governed by them to follow them.

1:14:23

And the final say is those who successfully inflict violence.

1:14:31

>> And again, this isn't fun to think about. It's not pleasant.

1:14:32

You would like to assume we operate in a world of laws that everyone follows them and goes by them.

1:14:39

But to the extent AI is as impactful and powerful as it is, the more these questions, fundamental questions that we thought have been settled for hundreds of years, if not thousands of years, are going to be raised.

1:14:52

And this is just the first of several episodes where I think that's going to happen.

1:14:58

>> I I grew up in sort of like postcold war, no ducking cover.

1:15:00

Didn't have a lot of fear of nuclear Armageddon.

1:15:02

But Dario Amade is, you know, a fan of this book, the the making of the nuclear bomb.

1:15:09

And uh and it seemed like he sort of predicted that uh if if AI becomes super powerful, the US might take a similar approach that they did with nuc regulation of nuclear weapons.

1:15:22

And I'm and as I was thinking about that, I feel sort of good about the way nuclear weapons are are regulated.

1:15:29

Like I I feel like it it it we got the good ending and we haven't had nuclear weapons drop in 70 years and it seems like things are going well there as well as they can considering that there's this amazing or you know tremendous like dangerous technology that exists but it hasn't been uh deployed and it hasn't actually you know bombed anyone.

1:15:50

Um but h how do you think he's processing that that book?

1:15:55

How do you think you're how do you think we should be processing that idea of of the government uh running the same playbook that they did with nuclear weapons?

1:16:05

>> It's pretty interesting.

1:16:05

I mean, on one hand, just from sort of a physical perspective, dealing with weights and software >> is very different than dealing with fishable material or uh I guess the the the the super bombs are like they're actual like fusion devices, right?

1:16:19

Um and that is trackable. it is >> interceptable.

1:16:24

You know when Iran to take a pertinent example is trying to build enrichment facilities.

1:16:29

All of which makes the problem >> easier to solve. >> Yeah.

1:16:36

>> So that's difference number one.

1:16:37

Difference number two and I really wish I would I I had this included and I cut it so that the sort of the the article would be tighter but there is a very interesting point in technological history which was the early days of Intel >> and Bob made the decision that we will

1:16:55

sell to the government but we're not going to design chips for the government >> and the distinction there was you had guaranteed orders which was great the government would take your IP and there is and in his mind the more important thing is there was limited volume

1:17:11

>> and the way that he foresaw correctly that this was going to be a very upfront capital intensive process of designing ships you have to design them you have to have the equipment all of which is in the billions of dollars today back then was in the tens of millions and hundreds

1:17:24

of millions is you need to find the largest possible market which was the consumer/ business market you design for that that will accelerate your improvement and your capabilities so much that you will end up having better devices than the government could have ever requested or made for itself. >> Yeah,

1:17:41

>> Yeah, >> that is at stake on steroids with AI. >> Yeah.

1:17:46

>> People like I was talking to someone, they're like, why doesn't the government just have just get someone to make their own model?

1:17:51

It's like because it's like you talk about government contracts, we're like singledigit billions.

1:17:56

We're talking about for the the amount that's going into capex, the cost of these models, we're talking hundreds of mil, you know, hundreds of millions of dollars for the models and hundreds of billions of dollars approaching a trillion dollars a year in capex.

1:18:08

That is only sustainable and viable if you're selling to everyone.

1:18:15

And but that introduces the entire new dynamics where the government built nuclear.

1:18:20

It started there and it started with a lot of assumptions because it was a government program.

1:18:27

We are necessarily for economic reasons because of all the upfront costs entailed starting with private companies of which the government is one of many customers and that introduces the assumption that well it's a private company with private property rights and all those sorts of things.

1:18:47

All of which I want to be true.

1:18:50

Again, I don't like how this is going down at all.

1:18:52

Yeah, >> the point here is to say there's a good reason why it's not going down that way and there needs to be cognizance that even though these is a private company that is building the model general purpose and for very good reasons wants to put restrictions again I think that the the surveillance one is a very powerful argument that I agree with.

1:19:12

The problem is that you just need to be aware of yes the government is a small customer.

1:19:22

The government is also the entity again not to be but with guns like they they you know like why do I pay taxes or because the law says to pay taxes no at the end of the day I pay taxes because you know if you really want to distill down if I don't someone with guns will

1:19:37

come to my house and throw me in jail right like like we don't think about that but at the end of the day where do these assumptions and laws and rights flow from and as long as that is still the case that it needs to be a decisionmaking factor for these companies Mhm. >> What how do you think this plays out for

1:19:53

>> What how do you think this plays out for Anthropic?

1:19:56

It's such a small contract, but it's so important in the zeitgeist.

1:20:01

There's a lot of people that are rallying around Anthropic because of this.

1:20:05

There's a lot of people that are pulling away from Anthropic because of this.

1:20:08

Uh it feels like there is a business to be built that doesn't work with the government, but delivers coding models and knowledge retrieval systems and a whole bunch of really valuable products and technology.

1:20:19

and it winds up being fine, but at the same time, you don't want this like hairy relationship with the government, adversarial, to go on for a long time.

1:20:30

>> I would like them to sell to the government, and I would like Congress to pass a law addressing these digital surveillance issues. >> Yeah.

1:20:38

>> And a lot of people are like, that's unrealistic, which I'm a medival to.

1:20:42

But at the end of the day, if you don't have it's legal or not legal as your guiding standard, the only alternative is someone has to decide.

1:20:56

>> And the implication of that not being a sufficient justification is that means a private executive is deciding. >> Yeah.

1:21:03

>> And if AI is what it is, I think that's going to be I use this word intolerable.

1:21:07

I didn't mean intolerable to me.

1:21:07

I meant intolerable to those with power to have a private executive making those decisions or not.

1:21:15

And if you think about if power if we're going to have this very sort of brute analysis that power flows from or laws flow from power AI is a source of power. >> Yeah.

1:21:31

So, it's not just that, and I think this is where the supply chain, again, which I'm not endorsing, but I think that's where the motivation is coming from.

1:21:40

The goal isn't to fine, we just won't use anthropic.

1:21:43

I do think the goal is to hurt Anthropic. >> Yeah.

1:21:47

>> And you're if you're not going to be subservient to us, you're not going to be allowed to build a power base. Period. >> Mhm.

1:21:55

>> And again, I'm not endorsing all this. >> Yeah.

1:21:58

It's just a matter of it's not a surprise this is happening. >> Yeah.

1:22:03

>> And it this needs to be just a this is a real risk factor.

1:22:06

A real that has to be considered in all these decisions.

1:22:12

>> Putting on my Dario hat, I'm thinking about a different way to achieve the goals with maybe less acrimony.

1:22:18

goals with maybe less acrimony. And I threw out this idea that maybe the better solution is like work with the government but then lobby for a surveillance act and actually try >> I mean I wish the White House would come out and say yeah there's a digital surveillance problem let's work on a bit

1:22:38

like I I I don't >> probably another regret I have is sort of putting this all on anthropic that was sort of the angle I was concerned about I and that left me I think fairly open to the critique that this is just like defending the White House's approach and that was

1:22:52

again that was I was trying to be a higher level that's saying look this is what's going to happen um >> but yeah the the >> I'm just thinking from the perspective way to find a middle ground here >> I'm just thinking of like from the perspective of like the if the White House is like this immutable thing but

1:23:08

you are in you know involved in anthropic like like one advice would be hey okay instead of going and having this confrontation with the government directly uh go and start a uh political action committee that lobbies for change in the way that you want through the democratic process. >> Yes, that is the ideal process. Um I

1:23:27

>> Yes, that is the ideal process.

1:23:27

Um I understand why people are frustrated and skeptical about this.

1:23:32

I used to have this debate a lot in the context of antitrust and aggregators.

1:23:38

And one of my sort of thesis about the aggregators and antitrust is that the the antitrust laws are fundamentally unsuited to dealing with aggregators because antitrust law has historically been about control of supply and the power of aggregators flows from control of demand.

1:23:53

And so you end up with all these solutions that I call pushing on a string.

1:23:58

You're just trying to get people to change how they behave and that doesn't work very well.

1:24:03

Like like Google has always been right.

1:24:04

Competition has always been just a click away.

1:24:06

The problem is people aren't clicking and like like so so the solutions focused on the supply angle doesn't work in a world where the supply is there just no one's choosing it. >> Yeah.

1:24:16

>> And therefore my prescription is you actually need to pass new laws not try to retrofit these old laws to this new use case where they don't work.

1:24:26

>> And the reaction is always that's impossible. We can't pass new laws.

1:24:28

and okay, but realize the implications of what of what you're saying.

1:24:35

I mean, I saw a tweet I I again I I I didn't like it, so I lost it forever.

1:24:40

It's one of the most infuriating things in the world.

1:24:43

But someone was like, I would definitely rather have Dario Amade make these decisions than and he and to this tweeter's credit, he wasn't limiting it to Trump because me this isn't a Trump issue.

1:24:52

This is a any politician issue. Yeah.

1:24:54

He said, I would rather have Amade making these decisions than whoever comes out of our screwed up democratic process. Yeah.

1:25:00

>> And points for the honesty because that's the actual choice that is that is being put forward.

1:25:08

>> And you could say Congress isn't going to do anything.

1:25:11

Therefore, Amade should just appreciate that is giving up on the democratic process and saying we should have unelected, unaccountable individuals making weighty decisions.

1:25:25

And again, I understand the sentiment.

1:25:27

It's hard to imagine Congress passing laws about anything. Yeah.

1:25:31

>> But just realize that's like that implication is is quite fraught. >> Yeah.

1:25:37

It's a huge change from I mean I just spawn in and believe in democracy and then understand it and study economics and and just have reinforced my belief in the American project throughout my entire career.

1:25:48

And now it it it really is people uh discussing an entirely different world of governance which is uh has been not something people have talked about publicly for a very long time but it is here for sure. >> Right.

1:26:01

And and they always come in on these Trojan horses that are eminently defensible.

1:26:05

Again I'm with anthropic on the digital surveillance point.

1:26:08

I've been I I've been concerned about it for years.

1:26:12

Been writing about it for ages. And it's similar.

1:26:14

There's an analogy to the to the monopoly.

1:26:17

Like you have all these laws that assume someone has to actually physically go somewhere and tap into a phone line.

1:26:24

>> But if you can do it with computers at scale, like suddenly you you had all these assumptions that limited what the government could do that magically disappeared not because the law changed, but because we got computers that could do the job of an individual at scale infinitely.

1:26:39

And AI again is going to the idea that the NSA, by the way, >> this is my sort of like I had to admit this in the article.

1:26:48

>> I was so confused why the Pentagon was so obsessed with domestic surveillance.

1:26:52

I didn't realize the NSA was part and I had the Pentagon.

1:26:56

>> John and I had the same moment. >> Yeah. Yeah. >> Yeah.

1:26:58

You sort of thought about as like an independent agent, like the CIA, but uh but that that made a lot of this story make more sense, right? No. Exactly. >> Yeah.

1:27:06

I feel like a lot of tech people are like reading the Fourth Amendment today and understanding like some of these like pretty basic processes.

1:27:13

>> Well, yeah, but like it's pretty like the loopholes are massive.

1:27:14

Like like I'm not denying it like like and it it it's similar to the chip thing with China.

1:27:21

Like my prescription for anthropic to give in is to allow these massive loopholes to be exploited and for the NSA to allegedly in the service of investigating foreign adversaries but by you know the process basically surveilling the domestic population I think is bad.

1:27:47

And the reality is a the nature of trade-offs is you're choosing between multiple bad options. >> Mhm.

1:27:56

>> And at some point it's like which team are you signing up for? They both suck. Choose one.

1:28:03

What do you think of the uh the the messaging around like the models themselves not being capable enough to be used in the context that the Department of War asked for because that felt like Daario was was sort of speaking for uh all frontier labs.

1:28:19

He said that like these technologies broadly are not suitable for these uh missions just yet.

1:28:25

I'm not sure that he has all of the information on the other side to know about the efficacy.

1:28:30

He certainly understands his models and what's capable in the frontier.

1:28:33

Um >> I mean I think that yeah I would I I mean I would assume they're definitely not capable that that I think that point is more of a precedent setting one.

1:28:41

I think anthropic's position is significantly weaker on that point.

1:28:47

>> Like at the end of the day we either trust the military or not to make these sorts of decisions.

1:28:52

That's why we have a military. Yeah.

1:28:54

And uh and so I I just I have a harder time and I think the the digital sales point is so compelling for them because again maybe this is my personal biases.

1:29:04

I think it's a huge problem.

1:29:06

problem. Um the you there's various anecdotes again I hate the reporting from these because you can tell like the leaks coming from which side for each of these but you know this idea that putting forward these hypothetical examples of like oh you can call us and we'll figure it out

1:29:23

then it's like no come on be serious about this like like so yeah I think that's a weaker argument for them so that's why I almost focus more on the digital surveillance one just because I think it is a very compelling argument in favor of the anthropic position. >> Jordan, anything else?

1:29:38

>> Jordan, anything else?

1:29:40

>> Oh, there's a lot more.

1:29:40

What are you What are you gonna be tracking going forward? Obviously, the stories. >> Yeah, good luck. >> Stay strong.

1:29:50

>> Uh, no, I mean I the open eye angle is obviously interesting.

1:29:56

I didn't really get into open AI.

1:29:58

Um, >> it's hard to parse exactly what's going on.

1:30:02

It seems to me they have agreed to the to the Pentagon that they will be the Pentagon will be limited by lawful capabilities. Yep.

1:30:14

>> And they make their own judgments about weapon usage.

1:30:17

>> And as I understand it, OpenAI is like we will on our side be free to stop the model from doing digital surveillance.

1:30:26

>> Which sounds like you're in sort of a jailbreak competition.

1:30:28

It's like we're going to agree to have a jailbreak competition with the US government.

1:30:32

Um, which I again it's an example of how fraught this is that that's probably the good place to come down on.

1:30:40

Now there's obviously these dynamics of competing for the same talent base being in San Francisco.

1:30:48

You know the this is part of I think Anthropic's Anthropic has a local advantage. Yep.

1:30:54

local advantage. Yep. in that most people I think in the industry are with them and they have a national PR problem in that I think a lot of folks outside of tech don't understand why tech companies always try to or resist helping the US government

1:31:10

>> and so it it's kind of an interesting dynamic where I think open AI is in step with the broader public and very much out of step with sort of their talent base in in San Francisco and so that's going to be very interesting to to see how that plays out. >> Yeah. Uh it's it's remarkable that >> Yeah.

1:31:25

Uh it's it's remarkable that Google has stayed out of the fray given all the Project Maven background and stuff.

1:31:31

Like they must be so happy that they're just like >> Well, that's the other interesting thing is uh I this is actually goes back to Google I believe >> where Google had the project I I think this is right but I but I think Google had project Maven which their employees objected to. >> Yep.

1:31:49

>> And therefore that went to AWS. >> Yep.

1:31:52

And then I some combination of I think the Pentagon is using anthropic because that's what uses what is a higher higher fed ramp uh designation. >> That's right.

1:32:02

So that's why anthropic was already allowed for classified content and open AI wasn't.

1:32:07

Again, I don't know the >> I've studied it. It's a wild story.

1:32:10

I mean the it was similar like AI for the military the same like killer robot uh fears.

1:32:17

fears. uh the actual I mean Google was a subcontractor on that project and uh what they were actually exposing to the uh to the government was TensorFlow APIs that would run on Google hardware and so they weren't actually writing any AI

1:32:35

software but they wanted to effectively like classify images from drones in the Middle East see that's a car that's a house and previously they had uh Air Force airmen just sitting there like clicking and they were like okay we're going automate that. Um, right. But, uh, Um, right.

1:32:47

But, uh, but it was still like scary, don't be evil, working with the government, military, and then there was a backlash. They pulled out.

1:32:55

Then eventually they went back in and and had a new head of Google Cloud. Um, >> yeah.

1:33:00

I mean, this is, you know, it's hard to, and I I speak for myself personally.

1:33:04

I obviously have the biased angle because of Taiwan.

1:33:06

angle because of Taiwan. I have the biased angle where I think there you know just in general there is this very naive view of the world that doesn't understand why militaries are important and necessary and I think Silicon Valley got itself in a lot of trouble by giving into this

1:33:23

naive mindset that we have no duty to support the military and there's this tension's been so it's a tension that's been brewing for years yeah which is are you >> an American company subject to American American law and even beyond law just morally compelled to support the US military >> or not. And there's an equally American

1:33:43

And there's an equally American sort of idea of moral consciousness. I'm able to say no.

1:33:50

That's why we have the first amendment, right?

1:33:53

This goes into the can the government compel a company to do something.

1:33:56

It goes back to some of the questions that happened, you know, with the first Trump administration.

1:34:00

And you know I I've been on both sides of this like which I >> and this is what this is what said in the in CVS interview.

1:34:08

He said we are a private company.

1:34:10

We can choose to sell or not sell whatever we want.

1:34:11

There are other providers.

1:34:13

He's already sort of like making this case. >> Yeah.

1:34:18

>> Which again it is a case that I support. >> Yeah.

1:34:23

>> But the point here is there's always the question with like a bubble or whatever.

1:34:27

Is it different this time? >> Sure.

1:34:29

And I guess that's sort of the question I'm raising. >> Yep.

1:34:33

>> Is AI actually applicable to every other technology that's come along?

1:34:35

Or if it is the the potential to be a source of power going forward, >> it's going to be dealt with as such. >> Yeah. No, it makes sense.

1:34:45

Uh last question, we'll let you go.

1:34:47

How happy should Ted Sarandos be right now?

1:34:54

I mean, I think he had the killer quote uh the last couple of days where I think someone was asking him if this is such a jewel and it's so rare, like isn't it a problem that you're missing out on it? >> Yeah.

1:35:07

>> And he's like, well, have you seen the history of Time Warner, which I think sounds about right.

1:35:11

I'm not sure how the the entity with all the debt that Paramount and Warner Brothers is going on.

1:35:16

Uh I think there's a bit where Netflix has always in the very long run been positioned I think to be the final buyer.

1:35:24

>> Like who else are content companies going to sell to? >> Yeah.

1:35:27

>> I feel like they sort of >> I feel like they've been spooked by YouTube a little bit and they felt a need to push forward. >> Yeah.

1:35:34

>> On that bring the bring the the future forward.

1:35:37

>> That was not allowed to happen, but that means their original plan I think still in place.

1:35:41

So, um, probably probably pretty happy.

1:35:45

All things considered, I'm going to say >> it's great.

1:35:47

Well, I'm I'm excited to get back to Netflix coverage and and, uh, more anodine topics.

1:35:53

>> I remember it was on Cheeky Pipe.

1:35:53

You were talking about getting sucked into, uh, the social and here we are.

1:35:59

>> So, I put that quote at the beginning of my article, you know, you may not be interested in politics, it was about politics and interest in you.

1:36:03

That was about anthropic and it was also about me. So, what can you do? >> Welcome. Welcome to 2026.

1:36:10

Well, we thank you for taking the time to come chat with us.

1:36:14

>> Yeah, great to see you >> and uh fantastic article. We appreciate you, Ben. Talk to you soon. >> Thank you. Have a great day.

1:36:20

>> Let me tell you about Phantom Cash.

1:36:20

Fund your wallet without exchanges or middlemen and spend with the Phantom card.

1:36:25

And let me also tell you about Crowdstrike. Your business is AI.

1:36:27

Their business is securing it.

1:36:30

Crowdstrike secures AI and stops breaches.

1:36:32

And our next guest is here live in the TVPN Ultra Dome.

1:36:37

We have James Bashara from Magic Mind coming on down for a very refreshing, very different pace of interview hopefully.

1:36:47

>> Great to meet you, John. >> Great to meet you.

1:36:50

>> We We actually We actually met I believe we met briefly in 2013. >> No way.

1:36:57

>> Because we were using Crowd Tilt >> at Soilent.

1:37:00

>> Oh, well then yes, of course.

1:37:00

Oh, actually that was a hell of a meeting.

1:37:04

We gave you all a huge like $800,000 check.

1:37:07

>> They they gave us a huge check, a physical check this big.

1:37:11

>> Yeah, we printed I literally like an hour before I was like do whatever we need to do to get one of those kind of you know just TV checks because y'all had one of the biggest crowdfunding campaigns of all time at that point. >> Yeah. Yeah.

1:37:25

>> I was just chatting uh with AJ from our team about Rob.

1:37:28

>> He was the one that was the leader on project yesterday. Yeah.

1:37:30

And uh yeah, that that was that was uh such a wild thing because we had applied to get on Kickstarter and at the time they said like no food products, nothing but like board games I guess or whatever they were doing at the time and you guys were like we'll help you out another YC company.

1:37:45

You guys built it in like a weekend and uh it worked flawlessly.

1:37:50

Also, I found out like you guys were basically like running digital ads for us and acting as our Facebook uh like like promotion engine crowdfunding platforms.

1:38:01

>> It was so helpful >> relatively marketing groups.

1:38:04

>> We just didn't have any marketing skills or any real like just you know we were so busy with other things that if we hadn't if you guys hadn't done that we probably would have ended that campaign like way >> lower.

1:38:14

>> lower. We took I still do take customer obsession to the extremes and yeah we saw it it was like hey this helps both sides we should we should lean into it the uh but yeah and for >> for listeners for viewers Soilent was I mean that was so gamechanging the whole

1:38:32

internet was talking about y'all for weeks super viral >> and uh and it was um for Y cominator and by the way both of y'all and I'm I don't know if if someone's just tuning in you know three months ago and then they are like, "Hey, these are this is just the best hair on the internet and that's why I tune in. These guys are journeymen and

1:38:48

These guys are journeymen and tech and don't it's so cool to listen to y'all because it's not a journalist that doesn't know how to actually build or what goes into uh creating a a company, a startup or sees a trend from a journalistic point of view.

1:39:06

you guys many in the arena through >> the founder perspective which is why I'm bringing y'all a >> s ton of magic mind for y'all let's do it round >> and as soon as I saw as soon as I saw yeah we should do some shots as soon as I saw I think it was on a Sam Alman interview >> that Jordy was doing and he was chugging a magic mind >> oh yeah I was like >> I'm going for a max we had a long weekend >> let's let's magic mind max it today. >> I love it. So, uh Yeah. Yeah.

1:39:36

Explain explain what this is.

1:39:38

What are we consuming here? >> Yeah.

1:39:40

The I'll get the the 10second uh commercial out of the way, but you just >> This is peptides. Yeah, exactly.

1:39:46

Peptides, and amphetamines, and steroids, uh allin-one, but you shake it, slam it back.

1:39:52

Jordy's the king of the slams.

1:39:53

>> Delicious mental performance spot.

1:39:53

Oh, it has some caffeine in there. >> And it does.

1:39:57

So, Max is unique >> some of that.

1:39:59

And some of >> Max is unique in that it is the world's first time release energy energy shot.

1:40:05

So the uh you couldn't do time release in liquid form until about 2 years ago and we were the first to put it in a shot.

1:40:11

>> What did you learn from building software that you applied to CPG? >> Oh my god.

1:40:15

>> Oh my god. I think like right away my experience as a customer has been this like iterative approach basically versioning out like the product when you released it versus two years later was like night and day in my experience like just a much better product which is something that unfortunately I think a

1:40:34

lot of CPG brands they kind of like make their product they ship it they get it into as many stores as they can they sell as much and maybe they don't even iterate that much on the product which puts you at which is just like extremely high risk from from my view, but that's sort of the stat. >> It is it's it's a really high stakes

1:40:48

>> It is it's it's a really high stakes experiment when you start shipping it to to stores and you get into a thousand doors and then you realize, hey, this is kind of a B+ product. We could make it better.

1:40:58

better. And uh and yeah, we've chatted about this over the years basically since uh the beginning of Magic Mind of take t taking excuse me everything that that I could from Silicon Valley and my background was in building software into building a dream company which meant the first three and a half four years was just improving probably 150 iterations

1:41:21

of making the product better better better before we went into stores because when you are DTOC and it is it is every time I say this out loud I can't believe um it's the case but as the first DTOC energy shot like it was I couldn't believe no one had done it before but one of the huge affordances is you can have version one version 1.1 1 version 1.

1:41:38

5 and each time that we made the product better we saw the retention curve go up up and then we're like all right this thing's ready for retail and now as of last month it's the number one health shot in the country um in the natural channel >> that's wild yeah I remember when I when I first tried it I was like okay I like what this gives me, but it hurt my stomach.

1:41:58

And then you were like, "Try it again." I tried it again. Fixed it.

1:42:00

And so I think that uh >> Oh, God bless the early testers there.

1:42:05

It was it was rough early on.

1:42:05

Um and and it was like, "Hey, I'll I'll eat nails if it'll help my productivity, focus, and flow."

1:42:13

But uh but it turns out most people wouldn't.

1:42:15

And yeah, it was it actually Bis Stone, huge shout out to Bis Stone, the co-founder of Twitter.

1:42:21

He's the one that put us in touch with >> Yeah. Oh, really? No way.

1:42:23

What college did you all go to? >> North Eastern. >> Oh, no way. Yeah.

1:42:27

brilliant uh mind investor.

1:42:27

He's the one who put us in touch with these oncology manufacturers.

1:42:31

This oncology manufacturing company, this is the most overressearched beverage on the planet.

1:42:37

So, we worked with them and utilize this technology to make everything so small that you can make it taste better than any of these ingredients have ever tasted before.

1:42:45

So, my favorite ingredients called bakcopa.

1:42:47

Bakopa maner decreases impulsivity by up to 50%.

1:42:52

Which is amazing for focus if you want to be risk. >> That's right. Yeah. Exactly.

1:42:54

If you want it on the track, I mean, performance takes many different types.

1:42:59

>> Your advertisers don't want it for all the logos that are happening right all over here. >> I I was I was skiing.

1:43:05

I saw a guy do a backflip on skis with no helmet.

1:43:09

>> That's impulsivity in a beautiful way. It was not on magic.

1:43:12

>> He needs a magic mind to put on the helmet cuz that was actually extremely dangerous.

1:43:15

Do not recommend doing a back flick up on skis with no helmet.

1:43:19

>> That's an that's a a non-magic mind for many people.

1:43:22

Many people when you don't want to jump into social media or the 50 texts on your phone, you want >> focus. Exactly. Yeah.

1:43:28

You want to be less important, >> but it tastes terrible until we found this technology.

1:43:33

>> So you took you took the iterative approach.

1:43:35

The R&D actually brought this kind of like hardcore R&D to a category that many people would have just again shipped a project and been like it's good for me, let's run it.

1:43:45

>> Uh what did you not take?

1:43:48

>> What did you not take?

1:43:48

I feel like that's great.

1:43:49

So much of the way that you run the company is is incredibly uh unique and kind of at odds with maybe how you built your last company.

1:43:57

how you built your last company. uh there are a handful of things and this is one of my favorite things to talk about is just how we build the company um and magic mind and it's you know that that there's that uh line in literature of once you know the rules and you can

1:44:10

break the rules and so I think my 20s of three different startups and a lot of it just failure to failure to failure but learning how other companies did it or the the playbook would be done and then recognizing and getting to a place and honestly just being able to fund it with my own capital I was like I'm going to do this my own way. >> Yeah. >> Yeah.

1:44:29

>> So, one of the things is we don't have any junior employees. >> Oh, interesting.

1:44:32

>> Um, it's about 10 senior employees. >> Everybody's senior.

1:44:34

We have plenty of great contractor.

1:44:36

There's plenty of great agencies that we work with.

1:44:37

But because of that, one of the things that we do is no meetings.

1:44:42

>> Um, it is all asynchronous.

1:44:42

Every once a month we have a team meeting and people can if you need it u you need to go to a meeting go for it.

1:44:49

But the default is asynchronous. Love we love Loom.

1:44:51

Uh, emails, texts instead of Slack. We don't I hate slack.

1:44:56

Um a uh voice notes but >> the senior aspect of the team members no junior team members means that the investment is very expensive per employee >> but man do we it's like a hot knife through butter whenever there's a challenge you have senior people through and through on every aspect of the business.

1:45:17

And I think with um my earlier startups and with with most startups, a lot of times you're hiring people that are doing something for the first time and say, "Hey, we need you to I know you've been a great engineer.

1:45:27

We need you to PM as well for this thing."

1:45:31

Because there's three or four of us.

1:45:33

With Magic Mind, it was like, "No, I want someone that's done this two, three times before and is a veteran that knows how to do this."

1:45:40

>> It's super key in in uh retail role. Exactly.

1:45:43

Exactly. because like it just it's so hard to get on the phone with a buyer and actually build the trust if you haven't already sold them the last oh you were at the last company you were the one that brought in Vitin water to Costco so the Costco buyer loves you because you brought him a banger product you're going to try the next thing much harder to just take a cold call and break through

1:46:02

>> and you know I'm sure you know this with soil in fact what are some of the things that you remember from the retail the first stages of the retail role you would never >> nightmare tell us more it's been a lot better with Lucy but Uh yeah, very very driven by uh trade shows, relationships, just being in some massive conference room, getting to know people over years and years and years, building trust. And

1:46:23

And then the only thing that can really accelerate it is that example of like bringing in someone who has a relationship on the other side where they delivered the goods and then they're putting their their reputation at stake, saying, "The last time I was with a smaller company, you took a risk on me. You were successful.

1:46:40

We both made money because this product, you bought a bunch of it.

1:46:45

You took your limited shelf space and you gave it to my previous company and it moved and you made money and you looked good. You got that promotion. Now you're VP. Uh do that.

1:46:55

Take that risk again with this next thing.

1:46:57

That's always been the best.

1:46:58

>> Take the risk and then also do it in the right sequential way to where you don't I think one of the things that one of the biggest cardinal sins in retail and I didn't realize this coming from software.

1:47:06

You actually don't want to move fast and break things.

1:47:08

If you get into, let's say you get uh uh what seems like a dream partner in Walmart.

1:47:14

>> If you don't move, >> y >> if the product doesn't move, and I didn't realize this until getting into retail, and my co-founder William is a genius on this stuff of just avoiding what makes you good.

1:47:23

I said, we say this internally a lot, what makes you good is what you do.

1:47:26

What makes you great is what you don't do.

1:47:27

One of the things that the trap that is a doom loop in retail is you expand too fast.

1:47:32

Let's say you get Walmart.

1:47:34

You're like, hell yeah, we got Walmart.

1:47:36

This is the 800lb gorilla.

1:47:38

And then it doesn't move.

1:47:40

>> And then you have to start putting in.

1:47:42

>> It's not like Facebook.

1:47:42

You launch an ad on Facebook, it doesn't work. You just wind it down. >> No.

1:47:46

Now it's like Walmart's like, "Hey, by the way, if you want to stay in here and you got one bite of the apple over the next 90 days, you've got to spend 400k in uh in spend and promotional spend." You're like, "400?

1:47:56

I didn't we didn't expect we had 270 grand in the bank.

1:48:00

You all aren't going to pay us for six months because of the payment terms you negotiated."

1:48:04

and they're like, "Well, then we're going to have to take you off the shelves."

1:48:06

So, then you go and raise some last last ditch effort uh round that goes towards you don't get quite to the 400, you only get like 300 and then it starts this slow doom loop and then you can't take it off the shelves because the worst graph in the world isn't flat.

1:48:22

It's one that goes up and then down.

1:48:24

Your revenue is flatlining or going down.

1:48:26

Then you can't raise any investment.

1:48:28

And I would say conservatively I'd say three out of four CPG brands get into that trap because they expand too fast. >> Everyone starts DDC.

1:48:38

The retail transition is the make orb break moment for like every new brand basically in my opinion.

1:48:45

And the hardest part is always for me uh rethinking the CAC to LTV math.

1:48:52

Like it's pretty easy on e-commerce to think about payback period.

1:48:57

Maybe you're not doing rorowaz.

1:48:59

Maybe you're doing LTV to CAC, but you're still looking at like a one-year payback.

1:49:02

And when you get into retail, you're looking at much bigger tickets.

1:49:06

It's not a hundred bucks on an ad.

1:49:06

It's $100,000 or a million dollars of commitment with some chain.

1:49:10

And then you might not see ROI on that because your margins are lower.

1:49:16

You might not see ROI on that.

1:49:19

>> You're extracted from the customer.

1:49:19

So, you don't really know what are they responding to, what messaging are they responding to.

1:49:23

There's a this goes back to the two obsessions that uh that I have are on the product. >> Yep.

1:49:30

>> Like everything in the product, every ingredient third party tested, everything's clinically backed the exact dosages on the product and then and that customer experience.

1:49:37

And then the other obsession is on the communication.

1:49:40

>> And one of the things that you can do in DTOC that you can't do in retail is you can iterate like crazy on the exact every word on here. >> Yeah.

1:49:49

>> Has been tested with course >> millions of eyeballs.

1:49:51

you go into retail without that that layer or you go in too early and it's a >> some people have it though and they can just they can just like oneshot everything and then just go straight to retail but it is very rare and usually they have a retail background.

1:50:05

Yes, they have a retail background and and one of the things that >> survivorship bias they got really lucky that started some company where like their whole shtick was like the packaging is going to be gold and it's just going to like jump off the shelf at you and like it did and they sold the company for like a couple hundred million.

1:50:23

>> Oh, David's kind of doing that now.

1:50:23

But I was think David's a perfect example where it's like every move is from >> What do you think about pricing?

1:50:29

>> What do you think about pricing? I always think of uh was it Koya that was like $12 a bottle or something and they sold for great outcome and it just flew off the shelves and it was almost like a Vlin good like a Lamborghini or a or a Rolex or something where the higher the

1:50:44

price people saw the higher price and they immediately thought like this is a better product so I'll buy it and that actually drove sales as opposed to duking it out for oh well we're the same price as the Red Bull so people will comp us to that versus like no this the the the the price tag is marketing. What

1:50:59

What do you think about that?

1:51:02

>> That is it's there's so uh there's so much discussion around that.

1:51:05

I'm an investor in a handful of of CBG companies that you can get caught into uh a lot of just like triangulation and you end up with it's like you know uh you homeland style like triangulation with the red string and we'll do this in these stores and this and these stores.

1:51:24

Yeah, >> I will say Air1 sometimes gets into like $8 shots of Magic Mine we're not fans of, but uh it is the number one shot there. So, it is working.

1:51:31

So, maybe it's a uh it's it could be uh maybe they've got a secret to it.

1:51:35

But there is a um a really simple framework that we use which is um I wanted it to be the best hell shot in the world. >> Yeah.

1:51:44

So in terms of everything from the vitamins for D and C and all the B complex for uh immunity because some nothing will sap your productivity like getting sick all the way to the bakcopa uh citicoline.

1:51:56

One of the things that caffeine does is it's a vaso constrictor.

1:51:59

So it will restrict the blood flow to the brain which is terrible for lateral thinking and terrible for switching contexts.

1:52:04

It's not great for a conversation when you're like oh I really want to absorb that new information but it is great for alertness.

1:52:12

alertness. But if you add in something like uh city choline specifically this this uh supplier we have cognizant which took it was like a two-year weight that's cognizant it's amazing so it's in mastermind and cognizant improves blood flow to the brain so that improves the the academic term for creativity is

1:52:29

lateral thinking improves lateral thinking all of these things are >> you stack them up and they're pretty expensive so I was like I just don't care what the price is to be the best in the world and it's exactly what I had been taking for seven years before even thinking of making into a product. Then

1:52:42

Then we work backwards and we're like, "All right, it's got to be this price." Uh, in the beginning $5. 99.

1:52:47

Now it's down to about 3. 99 because of scale.

1:52:49

We our batches are about 2 million.

1:52:52

Like uh we're producing about 2 million bottles a month.

1:52:56

So now with scale, you can get the price way down.

1:52:58

But it was man, John, it was a freaking it was a pushing a boulder uphill to get people to be comfortable with that price point in the beginning.

1:53:06

But I was like, nope, that's what the price is going to be.

1:53:08

>> What's your relationship like with the internet?

1:53:11

Personally, dude, we do get out daily.

1:53:14

Um, >> no, no, I just like I'm No, specifically like walk like walk me through the last 3 days of your internet usage. >> Oh, great question.

1:53:29

>> Because like I want to I'm trying to gauge how online you are because I feel like the the internet is just a a stress machine. It feels like it works. >> This is my device. >> Oh, interesting.

1:53:40

>> You don't carry a phone? >> Yeah. So, this is a a Monday.

1:53:41

I have it right now because of the GPS to get here and um and just uh being, you know, we Jordy and I live a few blocks from each other and uh and so I have the phone, but like literally I didn't charge it last night. It's 10% going to die. Rarely use >> it is. Yeah. Okay.

1:53:58

>> And I rarely use this thing.

1:53:58

Um the my relationship with the internet is uh >> use it when necessary >> and and then don't use it any other time of the day or week.

1:54:09

On the weekend specifically, we have three little girls and I go watch only and pay the extra whatever like uh 14 bucks for it to have cell service.

1:54:18

So that it's basically a dumb phone >> and interesting.

1:54:22

>> And I've got text and and I've got uh phone calls on it, but outside of that it is a dumb phone. >> You can't scroll. I can't scroll.

1:54:28

>> You can't scroll. I can't scroll. And it's a um I've been I'll create apps like with Replet to where I like created an app that brings up quotes of uh the Bhagavagita and stuff like that's the only thing when I sit down on the toilet

1:54:41

I might read a little bit of the Bhagavita on my walk and then I'm like ready to rock but I'm not where I was 10 years ago where it's like every >> tiny little moment let me jump into the amusement park that is my phone and then get lost for 45 minutes and it's a uh God screen time is Yeah, I know. >> I will say I I listen to you without

1:55:01

>> I will say I I listen to you without meetings.

1:55:03

I can deliberately and I love uh I will work from my iPad.

1:55:06

Try to do 99% of my work from my iPad.

1:55:09

Shout out to Replet there because I got a great iPad app and and there I will listen to y'all like this morning while I'm working out listening to y'all either on my, you know, the podcast app or on uh or on uh YouTube.

1:55:23

But man, like it's like the iPad is I'm on going into the offensive.

1:55:29

I'm gonna actually create something.

1:55:32

>> The watch is like I'm gonna be on the defensive if my wife needs me.

1:55:34

Um otherwise I'm with the kids.

1:55:36

But this is this somewhere in between >> where I'm like I can't really create with this phone, but man do I get sucked into rabbit holes.

1:55:46

>> Let's talk about >> Yeah.

1:55:46

and you and you and you get into a situation where you know I've got 30 tabs open on my device which is just like pure chaos. Right. Right.

1:55:53

And and these devices is sort of like are just due to their nature don't work very well to kind of multitask.

1:56:01

>> The two things that I talk about so often with founders as uh an investor in about 100 uh startups and I talk to them all the time again it's what makes you good is what you do.

1:56:10

What makes you great is what you don't do.

1:56:12

So deliberately for us really mitigating meetings or I end up telling about the devices that I love and the number one is the Apple Watch as a as a dumb phone.

1:56:23

When you're at the gym, don't bring you can obviously most people do bring their phones but you might get sucked into an email.

1:56:29

Certainly with kids my goal with them is by the time they're 10 spend 20,000 hours with them.

1:56:34

Um and and that was uh scrolling um when I was I think I was listening to uh a podcast um on my watch actually. So no scrolling.

1:56:44

And this this guest said that that was their goal.

1:56:48

This is like a year and a half ago.

1:56:49

I don't remember the guest.

1:56:49

Don't even remember anything about the other the other points that they're making in the conversation, but that stuck with me.

1:56:54

20,000 hours before their 10 because and his logic was so sound.

1:56:57

He said uh he could make $50 trillion and never get that decade back >> with uh the little ones.

1:57:03

And so I was like, man, I want to hit 20,000 hours with my kids.

1:57:07

And that was basically the moment where I was like, the weekends, no phone. >> Yeah.

1:57:13

>> And I'm just going to be with them in those hours.

1:57:14

I'm going to be completely with them and not sucked in because I just know myself too well.

1:57:18

>> I was saying no phone, just Apple Vision Pro mostly.

1:57:21

>> Nobody and and I used I vibe coded auto scroll so I don't scroll. It just scrolls. >> Yeah.

1:57:28

>> And it has all the Tik Tok and the Instagram reels and the X moderating the situation.

1:57:32

Dude, and you're virtually there with you. So, >> Exactly. Exactly. I agree.

1:57:35

I >> How is your approach to angel investing evolved?

1:57:40

Because I feel like angel investing can start out as like a fun hobby, a way to blow cash, but then you get caught up in this FOMO and and there you're on other people's timelines like you're kind of fighting for allocation at different points.

1:57:53

How how what's your approach?

1:57:56

>> It's it used to be uh more is better and it was and it's true.

1:58:00

Um, so much of the game of investing is you go wide because the asymmetric returns one investor.

1:58:08

>> I don't know who the next Gusto is. >> Exactly.

1:58:10

And that was >> Angel and Gusto.

1:58:12

>> I was Yeah, that was my first angel check.

1:58:14

They were in our batch cominator. I know. It was nuts. >> That's amazing.

1:58:17

I'll tell you about >> Gusto.

1:58:18

Everybody, >> the unified platform for payroll benefits and HR built to evolve with modern small and mediumsiz businesses. Go sign up. >> Wow.

1:58:28

That has never happened before where I've been able to mention an angel investment and then >> get a live ad read. >> Live ad read.

1:58:33

Uh >> what's your website? >> Mine? >> Yeah. >> JJBbasher. com. >> Yeah.

1:58:38

But the but the company magic mine. >> Oh, magicmindine. com. >> Is that on Shopify? >> It is. >> That's amazing.

1:58:42

Shopify is the commerce platform that grows with your business and lets you sell in seconds online, in store, on mobile, on social, on marketplaces, and now with AI agents. >> Continue. Continue. >> Shopify is great.

1:58:52

And as we get into the angel investing conversation, I bet a lot of ads will pop up because it is it has been I've been fortunate enough to to invest in a handful of great companies.

1:59:01

Um, but I went wide honestly.

1:59:04

Gusto did so well and uh and then a few others did really well.

1:59:07

Mercury did really well and um and then I'll tell you my biggest miss. >> What's that? >> Open AI.

1:59:15

And I'll tell you the story. It is f it is. >> Can I curse on you? Can I curse on you?

1:59:21

total total effing uh miss and um and it was so brutal because so I me and >> it's all good there'll be another open AI you'll have another show >> it's way different than missing just like a unicorn that comes up like you

1:59:37

can find one IC you're good >> it's it's I did the uh the math the other day it's basically missing 20 Googles >> Google went public at a $20 billion market cap so it's missing 20 Googles now with their latest funding it straight up missing. It is missing 30

1:59:51

It is missing 30 Google 35 40 Googles. So, uh, which is cool. That's all right.

2:00:00

Um, I never calculated the amount.

2:00:02

That's why you need the magic mind to clear your mind, load stress.

2:00:06

I need to scroll >> and it's there's so many things that I will in life where I'm like like the no meetings where I'm like I know I'm leaving money on the table.

2:00:13

I we couldn't I couldn't imagine uh any other life any other life. I I feel very fortunate.

2:00:19

Um, but I'll tell you the story.

2:00:21

So, >> I've only shared this once.

2:00:23

Um, but it it definitely replays in my head pretty often.

2:00:27

So, Sam and I, Sam Alman and I, we were advising and building out YC's YC research for universal basic income. >> And this was in 2017. >> Yeah.

2:00:38

>> So, every month we would meet in a conference room. >> Yeah.

2:00:42

>> This dingy little office. >> Yeah.

2:00:44

>> And the dingy little office, the conference room was where they were doing open AI research. Yeah.

2:00:50

>> So, every month and I was a full-time we had sold my last company to Airbnb and I was a full-time angel investor. >> Yeah.

2:00:58

>> Full-time angel investor >> waking up every day >> working >> thinking about Yeah.

2:01:04

what are the next gamechanging company?

2:01:06

And I'm like, yeah, Sam and the team, they're trying to do this AI stuff and it's research and and uh >> that can be a problem when you're too close to to an operation.

2:01:14

Whereas if you just get a pitch once, you're like, "Oh, this makes so much sense."

2:01:18

But when you're getting all the information at all times, and they're like, "Yeah, we don't really know what this thing's going to be, and we ran into this issue."

2:01:27

>> Has that friends it sounds terrible.

2:01:30

it sounds terrible. I ended up creating a rule of like friend starts a company like a real friend starts a company just invest >> uh even no matter if you know like oh they've got this issue with how they operate or they've got this blind spot or you know all the problems that they're facing uh you just if you're too

2:01:47

close to it you can just overthink it >> dude I told Ahmad at Mercury when he was like because we had built financial technology and he is grad we're at the Airbnb cafeteria and he was and he told me the idea for Mercury and wanting to start uh a bank for startup And I was like, "Don't do it, dude." And I spent

2:02:02

And I spent 45 minutes trying to convince him not to do it.

2:02:07

And and then like two days later, he was like, "I appreciate that."

2:02:10

Because we kept texting about, I appreciate the input, but I'm going to do it.

2:02:13

And I was like, "Well, okay, I'll invest. There you go.

2:02:16

>> This is not a good idea."

2:02:16

And good lord, was I a total idiot.

2:02:18

But yeah, the opening, I was sitting every month saying no again and again and again.

2:02:22

And even with the whispers of, "Yeah, I think we're going to have to spin it out, make it a a forprofit and uh and would you want to invest?" I was like, "AI."

2:02:32

And it's, I'll be honest, smart investors got in my head that were like, "AI is so far off that it's like that's like a 2035 thing. We're so far off."

2:02:44

>> Um, and I was a total idiot.

2:02:47

>> Well, 40 Googles later, I still think about >> later.

2:02:49

But plenty of other opportunities, plenty of other investments, plenty of other products.

2:02:53

Congrats on all the progress. >> It was good.

2:02:54

It was good to be humbled a little bit.

2:02:56

You were on too much of a of a hot streak.

2:02:58

It's good to be like, "Okay, I got I got to I got to lock in."

2:03:02

>> That is the benefit of building.

2:03:02

I think for anybody listening, I think it's what makes you guys so good at what you do.

2:03:07

It's what I think makes any creator really good at what they do is there are very few things.

2:03:11

It's like going to the gym.

2:03:13

You don't go to the gym for you think you're going there for the gains, but man, what you really get is the ego dissolution.

2:03:19

It is so >> It's so humbling. Yeah.

2:03:23

>> And anytime you put yourself out there, you think you're doing it for the gains, but you fast forward 10 years later and >> you're the journey. >> Yeah. It's the journey.

2:03:32

It's the the ego just the humility that comes with it that only comes with doing it because man, if you're just >> number two or number 2,000 at a big company, >> you're just living in your head like I would have done it this way, I would have done it that way.

2:03:46

But man, it's uh it is by putting yourself out there, it's quite um it grows your awareness in a pretty I'd say invaluable way.

2:03:55

>> Well, it's a great journey. >> Well said.

2:03:56

Thank you again for the oversized check and thank you for the oversized delivery of magic.

2:04:03

>> Thank you all for doing what you do in such a unique way.

2:04:04

We appreciate you >> for powering the show.

2:04:06

We'll talk to you soon for us.

2:04:10

>> Let me tell you about Gemini 3. 1 Pro. So, Gemini 3.

2:04:13

1 Pro is here with a more capable baseline.

2:04:16

It's great for super complex tasks like visualizing difficult concepts, synthesizing data into a single view or bringing creative projects to life.

2:04:24

And let me also tell you about Reream.

2:04:26

One live stream, 30 plus destinations.

2:04:28

If you want to multiream, go to reream. com.

2:04:29

And up next, we have John Quinn live in the TVPN Ultradome for the second time.

2:04:37

Thank you so much for making the trip down to our studio. Welcome back.

2:04:40

John Quinn of course is the most feared lawyer in America. We are huge fans.

2:04:48

Tyler over there wears that hat every day. He is obsessed. You have a lot of fans.

2:04:52

>> I got a couple for you for you guys.

2:04:54

>> And you know I'm your neighbor.

2:04:54

I live in right next uh very very close to you.

2:04:59

>> I know you went to school with one of my poly technical. I don't know which one.

2:05:02

I had five who went there.

2:05:04

>> Uh I was in the same class as Jamie. >> Oh, okay. >> Yeah. Yeah. That was a great time.

2:05:07

Anyway, uh we're not here to talk about the minor life in Pasadena.

2:05:09

We're here to talk about tariffs.

2:05:12

We're here to talk about law.

2:05:13

Um, what have you been tracking?

2:05:15

How have you been processing the the back and forth on the tariffs this year?

2:05:20

>> Well, uh, 10 days ago, the United States Supreme Court ruled that President Trump didn't have the legal authority to impose these tariffs, which were kind of the cornerstone of his international trade policy, domestic policy, maybe also >> uh, provided a lot of funding.

2:05:33

He's collected $600 billion worth of tariffs. >> Yeah.

2:05:38

>> Yeah. He imposed those tariffs under something called the economic emergency powers act >> uh which gave him a lot of discretion under the administration's interpretation of it to I mean we saw this you know one day it was 35% then it was 50% and it was all this there were there were two categories of tariffs

2:05:55

there were the fentanyl tariffs that affected some countries Mexico Canada a couple of others and then there were the liberation day tariffs which was across the board different deals for for different countries it had never that statute had never been invoked before or used to impose tariffs. Uh it was

2:06:10

Uh it was challenged in the uh court of international trade in New York.

2:06:15

He lost there and ended up going to the United States Supreme Court.

2:06:19

The argument was a couple of months ago.

2:06:22

>> Uh the argument didn't go very well for the administration reading the tea leaves.

2:06:25

I think a lot of people thought the administration was going to lose and then sure enough the decision came out on February 20th and the Supreme Court ruled that, >> you know, a lot of people thought the court would rule that these really weren't emergencies.

2:06:38

The emergencies he was relying on were fentanyl and the balance of trade being an economic emergency.

2:06:44

And a lot of people thought the court was going to say, "You haven't shown that that's an emergency."

2:06:46

But that's not what the court went off on. >> Okay?

2:06:50

>> The court ruled that he simply didn't you couldn't impose tariffs under this statute.

2:06:55

It had never been done before.

2:06:57

There's no reference to tariffs or customs or anything or taxes or anything like that >> in the emergency economic power >> act itself.

2:07:03

And the court said wrong statute. >> Yeah.

2:07:06

>> And and zooming out, where do tariffs typically come from? What's the correct path?

2:07:10

>> Well, there's a there's a number of different paths.

2:07:12

There's uh section 122 which he's now invoked. Okay.

2:07:14

uh there's a national security regime where you can impose tariffs but it requires actually an investigation that's been done by I think the department of commerce or some other agency.

2:07:24

So a record has to be created.

2:07:27

The president can't just on his own say you know you're going to pay 35% you India you're going to pay 50%.

2:07:34

>> So there is a whole other regime available.

2:07:37

>> Uh and the day he lost he you know he was ready for this.

2:07:40

He announced there's another statute that permits him, section 122, that permits him to impose up to 15% tariffs >> for 150 days. >> Yeah.

2:07:50

>> After 150 >> and then theoretically, you could just pause for a second and then start up again. >> Is that true? >> I don't think so.

2:07:57

I think the courts would see through that.

2:07:58

But after 150 days, Congress has to act. >> Yeah.

2:08:03

>> So, he can do it for 150 days and then to extend it, Congress has to act.

2:08:06

And is that is that passive in the sense that if Congress doesn't act they don't continue or by doing that 150 days it forces a vote in Congress?

2:08:16

>> I don't know that it forces a vote.

2:08:16

I think Congress has to take affirmative action.

2:08:20

>> They are the next one they know the authority.

2:08:24

>> So the day he was ready for this the day he lost in the Supreme Court he announced okay I'm I'm invoking the section 122 10% tariffs on the world. >> Yep.

2:08:32

And he could have gone up to 15.

2:08:34

>> Well he did it the next day.

2:08:34

He says, "I thought about it again. 15% on the whole world."

2:08:39

So, >> that's where we are right now. But the clock's running.

2:08:43

>> And what about what about the Supreme Court didn't give any guidance on how the tariff uh the cash generated from the tariffs should be returned, right?

2:08:52

So, there's also kind of a gray area on >> they didn't they didn't address that.

2:08:56

But, you know, honestly, guys, that's pretty straightforward.

2:08:58

uh these 670 billion in tariffs.

2:09:01

There aren't there are a lot of claims.

2:09:04

Many many companies have already filed claims and there aren't a whole lot of defenses to those claims.

2:09:09

I >> mean, he has said he's going to do everything he can essentially to delay it.

2:09:13

Y >> you know, maybe so it's not on his watch the next administration, >> but there aren't a whole lot of defenses.

2:09:18

The these are these are good claims.

2:09:20

There's one issue about >> in some instances the costs have been the tariffs have been passed on to consumers.

2:09:26

consumers. So the government might presumably have a defense that hey wait a second you don't get all this back >> your profits didn't go down there was no damage you passed it on so you you saw the revenue anyway >> interesting >> and that raises a question about whether these secondary parties parties that paid for things that included these

2:09:43

tariffs which have now been ruled illegal whether they can bring claims that's an interesting question so I think I think we're gonna see some of those claims >> so for a company that's trying to get a big refund maybe in the hundreds of millions of dollars I imagine for some of these companies Is the first step just to file a claim or do they need a lawyer? Is this going to be a lawsuit on

2:10:00

Is this going to be a lawsuit on a per company basis? How do you think that?

2:10:04

>> Yeah, I think it's I think it's a a lawsuit on a per company basis.

2:10:05

They have to >> they have to file a claim. Okay. >> Lots of lawsuits.

2:10:10

A lot have already been we're we're filing lawsuits.

2:10:11

I mean, one of our partners, a guy named uh Dennis Herinitzki, uh is representing a number of clients already and bringing these claims.

2:10:21

But what what folks need to do if you want to pursue get your documents in order, get your records in order, figure out what you paid when >> and yes, you probably need to engage a lawyer.

2:10:32

I don't I don't think you want to do this on a per, you know, representing yourself. >> Yeah, exactly.

2:10:38

>> You don't want any hallucinations and what and screw up of what is otherwise a powerful claim. >> Yeah.

2:10:43

>> The claims are brought in a very obscure court known as the Court of International Trade. It's in New York. >> Okay.

2:10:49

uh you bring the claim there and um you you know you pursue it.

2:10:52

Uh the government as >> what what how how staffed up is that court?

2:10:57

How could they possibly process >> what could end up being a mill it could be an administrative burden but the issues are pretty simple here.

2:11:06

I mean it's you know he invoked this act. >> Yeah.

2:11:11

>> Uh they paid the tariffs. Yeah.

2:11:13

>> The Supreme Court says he didn't have authority to do that. >> Yeah.

2:11:16

And you know, the government refunds tariffs. Yeah. All the time.

2:11:19

That's it's kind of a routine thing.

2:11:19

So, it's not like this hasn't happened before.

2:11:23

>> Can you take me back to the Supreme Court battle and help me understand what's involved in uh fighting a case in front of the Supreme Court?

2:11:30

How much do lawyers know about the positions of the various Supreme Court justices before they walk in the courtroom?

2:11:38

How does all of this play out?

2:11:41

What goes into a Supreme Court case?

2:11:43

How how high stakes is it?

2:11:45

Well, I I think this was super high stakes.

2:11:47

I mean, this was a you know, a major decision for this administration, maybe the biggest uh trade tariff decision ever.

2:11:57

>> And I don't know that there was a lot of uh decisions by these justices before that could help people kind of predict what their reactions would be to this.

2:12:07

>> The majority opinion was by Chief Justice Roberts.

2:12:10

>> Uh there was a disscent by uh Brent Kavanaaugh and some others.

2:12:12

uh but u you know in preparing for it I don't think it was unique to uh any other Supreme Court argument.

2:12:21

Obviously that's the biggest of the big leagues.

2:12:23

So people prepare carefully for those arguments.

2:12:27

>> How quickly Oh, sorry. >> Go for it.

2:12:29

>> Uh uh how quickly did the legal community come to a consensus around the impact and longevity of the tariffs?

2:12:38

Because from our perspective, the market reacted violently on Liberation Day.

2:12:40

It was the worst day in financial markets in since COVID effectively.

2:12:44

Uh but it feels like you've known or you've been able to predict that these tariffs might get struck down sooner than maybe most of the financial community.

2:12:54

So what was the process like for the legal community to get consensus around what might happen at the Supreme Court level?

2:13:02

>> I think it was pretty straightforward especially after the oral argument before the Supreme Court that went so poorly. >> Oh, interesting.

2:13:08

So it wasn't it wasn't that months ago before even oral arguments started people were looking at precedent and how the different justices >> sure sure they were I mean uh you know the the administration had lost below in the court so that you know people knew that this is definitely in play. >> Okay.

2:13:27

>> So uh I mean I think the issues were pretty straightforward.

2:13:29

I mean does this act give him the authority >> and as the Supreme Court found there's no history for it.

2:13:35

It had never been invoked before for this purpose.

2:13:37

There's no reference to tariffs or customs in the act. >> Yeah.

2:13:43

>> So, as legal questions go, uh, you know, I don't think it's that complicated. >> Okay.

2:13:48

How do you think about the speed the time between liberation day and the Supreme Court ruling that felt like almost a year?

2:13:54

Is that deliberate by the administration?

2:13:58

Is that is that just a function of how busy our Supreme Court is?

2:14:01

Why does it take so long to actually get a decision from the Supreme Court?

2:14:06

>> Well, some some decisions take longer than that.

2:14:08

U look this everybody, you know, the Supreme Court reads the newspapers as somebody said during the FDR administration. That's funny.

2:14:16

They read the tea leaves.

2:14:17

They they know how there's no doubt they knew how important this was.

2:14:23

>> So, they know what's coming even before the case.

2:14:26

>> I'm sure they see it coming. Okay.

2:14:26

But in terms of why it took so time between the oral argument and the rendering of the decision, it was, >> you know, people were waiting uh wondering uh but it it wasn't an extraordinary amount of time.

2:14:38

But I think because of the known repercussions of this decision, the court wanted to be really really careful and straightforward. >> Yeah.

2:14:46

Are there any other uh big new trends in cases that you're tracking that might make it to the Supreme Court this year?

2:14:53

I'm thinking particularly of some of the artificial intelligence debates that we're having.

2:14:57

We were just talking to Ben Thompson about the relationship between anthropic and the department of war.

2:15:02

And there's a lot of folks that are asking for clarity on the regulatory side, whether an act of Congress or new laws, new leadership, more messaging, but the Supreme Court could play a role there.

2:15:12

Is there anything that you're tracking in terms of like upcoming cases that will be particularly consequential for business or technology?

2:15:20

Well, we all know about this issue about using copyrighted materials to train large language models and we have cases pending, you know, relating to literary works, visual works, >> uh, musical works.

2:15:33

>> There are dozens of these cases percolating their way through the courts.

2:15:37

>> Yeah, >> there have been a couple of decisions up in San Francisco now, the early decisions at the district court level, the trial court level >> where the courts kind of came out.

2:15:44

uh the decisions were kind of limited because of the unique circumstances of those cases.

2:15:50

One was a meta case and often and I think the other was I'm forgetting what the other case was >> where they prevailed at the district court level.

2:15:59

>> Uh but before these cases can get to the Supreme Court, they have to go through that process a decision at the district court then you have a court of appeals.

2:16:06

So and then and then to the US Supreme Court.

2:16:09

>> I mean it could well you know our Supreme Court doesn't hear very many cases.

2:16:14

>> How many do they hear?

2:16:14

I I think they I think they do like I'm I think like around 80 cases a year. >> Oh wow.

2:16:20

>> You compare that like to India where they the Supreme Court there hears thousands of cases a year.

2:16:27

>> Of course, they have hundreds of judges on their supreme I think on their Supreme Court. >> Yeah.

2:16:32

>> Um but yeah, I I I think there are definitely issues in the AI world that are could reach the Supreme Court.

2:16:37

So walk me through the the thought process that uh if you were advising me and I believe that an AI lab has has trained on my intellectual property uh and that lab is wellunded and they come to me with a settlement, why would I want to turn a settlement down and fight it out and take it all the way, go the distance all the way to the Supreme Court?

2:16:59

Like what what what are the different uh tradeoffs that that plaintiffs are making right now?

2:17:06

you need to create a downside uh for the AI company.

2:17:10

You know, they've got to see that you have the staying power that you're prepared to litigate this case. Yeah. >> Anthropic.

2:17:17

>> So, you need so you need deep pockets.

2:17:19

So, if you're just a single, you know, author of a book, then you're not independently wealthy.

2:17:23

They're probably going to say, "Hey, you'll take the settlement."

2:17:26

But if you're a class of >> Well, you know, it depends on what the what the settlement is, I guess.

2:17:29

Uh but you know my experience is that defendants don't write checks until they see a downside. >> Got it.

2:17:38

>> And so that involve you really have to kind of establish your credibility that we're serious about this.

2:17:42

We're willing to pursue it.

2:17:45

>> U anthropic you know settled a case for a billion maybe two billion a copyright class action. Yeah.

2:17:51

Uh but those were, you know, very good lawyers who had put put together a case and they had gotten anthropic in the position that they're facing whopping damages.

2:18:00

If they lost at the trial court, it could have been much much more.

2:18:05

>> So that's the calculus.

2:18:05

You kind of have to create a downside for >> Yeah.

2:18:08

And what what would the economic damage calculation be?

2:18:11

Is that like I'd look at the profits that they've made from that or the revenues?

2:18:15

like what sort of economic case should I be making if I'm trying to extract the maximum amount of damages? >> Yeah.

2:18:23

I mean, under the copyright laws, there are specific rules about what's recoverable in the case of copyright infringement. Sure.

2:18:28

And that's the theory here that this is copyright infringement to use our copy, my copy, the plaintiff's copyrighted materials. Yep.

2:18:36

>> In order to train the large language model.

2:18:38

And there's a couple of there's statutory damages, you know, which are a c fixed amount per infringement.

2:18:45

>> Uh or there there are theories under which you can recover the profits they they got from it.

2:18:50

But there are various rules about how damages are measured. >> Yeah.

2:18:54

Yeah, that makes a lot of sense. Jordy, sorry.

2:18:56

>> Uh give us a 101 on what it's like to uh uh be in an active uh case or litigation with the with the federal government.

2:19:05

Uh I'm sure over the many years we've talked it's is it >> was it Palanteer or Anderl that that uh or both of them?

2:19:15

>> Palanteer and SpaceX both sued the US government over uh contracts that went that were not fairly bid in their minds.

2:19:25

Uh they didn't have a correct shot at displacing an existing program with the Palunteer was des >> Yeah.

2:19:31

But but let's say you you have a a client that is considering taking uh an action against suing the government.

2:19:39

What how how do you kind of like walk them through it?

2:19:41

Obviously, there's a bunch of details uh case by case, but what's the general framework that you operate under?

2:19:48

>> Look, the um there are good things and bad things about litigating with the government.

2:19:52

We sue the United States government all the time.

2:19:53

We had a we had a case relating to uh Obamacare.

2:19:56

There was something called risk corridors. Okay.

2:20:01

>> Where insurers under Obamacare were assuming liabilities for markets that they had no history with.

2:20:07

They didn't have a basis to underwrite it. >> Okay.

2:20:10

>> And there was a term that was put in the Obamacare that said, "Okay, we will we the government will backs stop you, the insurance companies." Sure.

2:20:16

So, if your, you know, if your losses are above a certain amount, you know, we'll fund that.

2:20:25

>> Well, that was never funded and that was never paid.

2:20:27

Uh and we >> basically so the bill was coming due.

2:20:31

The insurers were saying, "Hey, Mr. Governor."

2:20:34

>> You know, interestingly, the insurers, many of them weren't eager to sue the, you know, stick their heads above the parapit. >> Yeah.

2:20:40

>> And sue the United States government. >> Yeah.

2:20:42

>> But we got involved and we brought a uh a class action on behalf of a number of insurers.

2:20:46

It took some it took some persuading to get insurers to to participate.

2:20:51

>> Uh and ultimately we recovered like 7. 5 billion dollars. >> Wow.

2:20:55

Now, the good thing about suing the US government is if you get a judgment, it's not like you have to go out and hire investigators and find the assets.

2:21:03

If you have a judgment, there's a window in Washington you can take your judgment to and they give you the money.

2:21:09

>> Did you get a check for $7? >> Is it a drive-thru?

2:21:11

Like a drive-thru like McDonald's?

2:21:13

>> It wasn't that simple.

2:21:13

They appealed and uh they lost.

2:21:16

But once it was once the dust was settled, yeah, they they give you the check. >> Wow. >> We have a case now.

2:21:22

Uh I mean we represent Harvard University uh versus the government on we sued Homeland Security on the ridiculous ban on I I thought it was ridiculous on foreign students around you know coming to Harvard.

2:21:36

So we sued Homeland Security on that.

2:21:38

We sued Health and Human Services on the defunding and the cutting off of grants for research and things like government about that. >> Yeah.

2:21:46

>> So both those cases uh we went on the trial court level and it's up on appeal now.

2:21:51

the government is appealing that.

2:21:53

So litigating with a government, you've got to assume, you know, they're well resourced.

2:21:56

I mean, I don't know how many hundreds, probably thousands of lawyers there are in the Department of Justice. Yeah.

2:22:02

>> So they're going to show up, you know, they're going to advance the government's position.

2:22:06

And so, >> uh, they're not just going to lay down.

2:22:10

Can you describe the uh the level of uh sort of back and forth career paths that happen between Quinn Emanuel and the Department of Justice or the government?

2:22:21

Are there folks that you're recruiting from the government occasionally?

2:22:23

Are there folks that leave and go do a tour to understand what life is like on the other side or is it two sort of separate communities?

2:22:31

>> No, it kind of goes both ways. It does.

2:22:33

Uh at our firm we have about 2530 lawyers who one time or another were prosecutors, US attorneys, assistant US attorneys or high ranking officials in the department of justice.

2:22:45

>> And that's seen for a youngish lawyer uh a job as in the US attorney's office as a a good job, great experience. Yeah.

2:22:54

>> Uh and those those folks are in demand at firms firms like ours. >> Sure.

2:22:58

And then on another level, uh, senior people sometimes if they're interested in government service and they're connected because a lot of this does involve political connections, you can get an appointment.

2:23:09

We've had people in our firm who have appointments at very senior positions in the Department of Justice.

2:23:15

So, it can go back and forth. Yeah.

2:23:17

>> I will say another thing right now, this Department of Justice is really busy.

2:23:22

>> They're facing all kinds of cases.

2:23:22

And you know, based on what I hear and what I read, they're they're kind of resourced. >> Yeah.

2:23:30

>> So, uh I'm I'm not suggesting they're not going to show up and defend, but >> Yeah.

2:23:34

>> I mean, like, >> does that mean your your associates are getting more calls to come over?

2:23:39

>> Um you know, we we've had a couple of associates go to work at the Department of Justice in the recent past. Yeah.

2:23:46

>> But that's always been true. >> Yeah. Yeah.

2:23:48

>> It it's a good experience. >> Yeah. Yeah, that makes sense.

2:23:50

How is uh in your view AI impacting the legal industry today? Not in the future.

2:23:55

You know, what what you know, how will the impact kind of evolve over time, but what are you seeing and hearing from, you know, colleagues or or uh friends at other firms?

2:24:09

>> Look, it's we are wordsmiths, right? We work with words. We write things.

2:24:12

And so obviously large language models are something that is going to change how we work.

2:24:20

They may change the whole structure of law firms.

2:24:21

Big law firms historically are kind of described as pyramid in structure.

2:24:26

You have senior people and then you have a lot of junior people.

2:24:30

A lot of what the junior people have done in the past can now be done by large language models.

2:24:35

It's not like they're going to give you a work product >> that you can then file with the court and use.

2:24:41

I mean there are hundreds of cases where lawyers have gotten in trouble with courts by filing things that cite cases and laws that don't exist.

2:24:51

>> So that is really that's definitely a thing.

2:24:52

The hallucinations that's a huge >> risk factor and that's >> that's an that's really on the lawyer.

2:24:57

I mean there's no there's no excuse for a lawyer to have said, "Oh, I relied on AI."

2:25:03

No, you signed that thing. Yeah. >> And that brief. You're responsible.

2:25:06

It's your integrity on the line.

2:25:08

So uh we're not we're nowhere near getting I don't think at least in our practice maybe we do complex litigation.

2:25:16

There may be practices where it's a lot of repetitive litigation where you know you can get output that's you know ready for filing but we're really not seeing that yet.

2:25:27

But we're we've developed in house at our firm a uh a platform a methodology for uh taking large masses of data that we work with all the documents that are produced in discovery all the testimony all the contracts and we organize organize that on it's built on the

2:25:47

clawed enterprise platform and we've done this as lawyers >> it's a system that's it's proprietary developed by lawyers for lawyers I think if you just turn a young associate loose with claude or J chat GPT, you're not optimizing the technology. But we take

2:26:01

But we take all the data and we structure it for a way that lawyers work.

2:26:06

So it creates work streams.

2:26:10

So like what do we need to do?

2:26:12

We know what we do in every case.

2:26:12

We prepare examination outlines.

2:26:14

We prepare exam uh expert witness reports and the like.

2:26:20

We prepare opening statements.

2:26:20

So we structure in the data that creates these work streams and uh I really think that's the that gives us a big advantage.

2:26:28

That's it's not something engineers have created.

2:26:31

It's lawyers knowing what lawyers need having a designed a way to structure the information. >> Yeah.

2:26:39

>> And you know we're using that with great success.

2:26:41

I mean in trials now in the middle of trial imagine >> somebody's on the witness stand. >> Yeah.

2:26:47

You can ask the AI, what's the best evidence that so- and so just lied about that. >> Wow.

2:26:54

>> You press a button, it just doesn't >> and you get a B.

2:26:56

So, and most of those things you will have thought of.

2:26:57

Some of them make no sense, >> but there'll be a couple of gems in there. >> Yeah.

2:27:03

>> You know, that you might not have thought of, lines of attack that you haven't thought of.

2:27:06

>> That's extraordinarily powerful. >> Yeah.

2:27:08

>> So, what we're our goal is to get to a point where the AI yields a work product that's like 80% or 90% there. Yeah.

2:27:16

which is what an associate is typically doing today, right?

2:27:18

It's not you're not like the best associate is not hitting it out of the park every single output.

2:27:25

They're getting it they're getting a good good solid chunk of the way there.

2:27:28

>> You're absolutely there right and so lawyers can focus on what they do best, making sure that last mile, the last 20%, the last 10% is as good as it can possibly be.

2:27:39

How do you how do you think this affects the job market for uh lawyers at the early stage of their career?

2:27:47

because in some ways yeah the their their work might being uh be being replaced but at the same time uh given that that AI is very good at generating words and can if will be able to generate entire lawsuits you can kind of imagine a dystopian world where the number of you know uh cases that get brought are are you know 100 times uh higher than they are today.

2:28:14

>> I think that's true and that's something that's people don't talk about a lot.

2:28:18

There are AI companies, AI native companies out there that essentially identify claims.

2:28:25

>> So they'll have a database that has information on businesses of all kinds.

2:28:30

What licenses do they have?

2:28:30

What licenses don't they have that they should have under the law?

2:28:33

Uh I mean you can just imagine if you can boil the ocean. Yeah.

2:28:39

>> And they will these companies >> you can subscribe to them and they'll serve up here's a class action for your consideration.

2:28:46

We've we've identified the claim and and here it is.

2:28:48

So I actually think there's a potential that we may see more litigation as a result of AI. >> Yeah.

2:28:56

>> On the other hand, I think resolution of cases may be faster. >> Sure.

2:29:00

>> Because both sides can understand quicker. Yep.

2:29:03

>> You know, the merits of the case on each side and reach a resolution sooner. >> Yeah. Yeah.

2:29:08

No, that makes a lot of sense.

2:29:09

Um, last question from my side.

2:29:12

Can you tell me a little bit about how you perceive like the battle of the forms and the level of the level of detail that contracts might be entered into?

2:29:22

into? I'm thinking of this uh this news with Anthropic and the Department of War about the these like two lines domestic mass surveillance fully autonomous weapons and there's and there's like in every contract there's a question of do do you write it in one page or a 100 pages and I think most lawyers can tell you I can do either right

2:29:40

>> uh and they have different flavors and they have different subtexts and it depends on the type of relationship that you're forming and so I'm wondering how how how much you think about that when you're communicating in a legal context, in something that's going to be binding, what level of legal ease you want to use, what length of document you want to use. How important is that in

2:30:02

How important is that in >> I think I I I think it's super important.

2:30:05

I mean, it doesn't matter.

2:30:07

Everybody, you close the deal, you have a closing dinner, people shake hands, they have a drink, and as long as things are going fine. Yep.

2:30:14

>> Everything's fine, right? You never hear about it.

2:30:15

And the legal ease doesn't matter. >> Sure.

2:30:17

As soon as you know a couple years down the road after an acquisition, joint venture, whatever, circumstances have changed. Yep.

2:30:24

One side has a different goal.

2:30:27

They want to go in a different direction. Yep.

2:30:29

>> Can and can they go in that direction? Yep.

2:30:32

>> Or they're not performing. They faced headwinds.

2:30:35

>> Uh then it's only then that people start to scrutinize the language and say, "Wait a second.

2:30:39

Can they get >> what did we actually agree to?" >> Yeah. What can they do this? Yeah.

2:30:42

What are our options on both sides?

2:30:45

If you want to get out of a deal or you want to create some leverage, you're going to scrutinize that language, of >> course.

2:30:52

>> And that's why, you know, we think it's a good idea.

2:30:55

Uh it's going to sound totally self-promoting, but uh we have clients, major private e global private equity clients who ask us to do this.

2:31:05

Have a especially if it's a novel structure that hasn't been tested before, something you haven't done before. >> Sure.

2:31:10

take a look at this and pressure test it for me from a legal risk perspective. Sure.

2:31:17

You know, and I think it's especially valuable to have a litigator's eyes on the problem because look, >> we live in a world where 24/7 something's gone wrong. Yeah.

2:31:28

So, day in day out we're dealing with the very kinds of situations you're talking about.

2:31:34

>> So, a lot of times we can, you know, we look at these things with a jaundice eye.

2:31:38

Uh, and a lot of times we can look around corners and see how might a devious counterparty try to take advantage of this situation.

2:31:48

>> So I guess the answer to your question is it doesn't matter until it does. >> Yeah.

2:31:53

>> And so because of our experience as litigators, we can look at things and kind of see, oh, there's a potential loophole here. >> Yeah.

2:32:01

>> It might be a simple agreement like we had a case, >> an arbitration uh in Asia.

2:32:04

I won't get into the specifics because it's confidential. Sure.

2:32:09

But basically, one side had an option, you know, uh, and there was a method for determining what the option price would be. >> Sure.

2:32:19

>> And then it was kind of like a a bit like a baseball arbitration and then one side get one number and the other side would come with and there would be a neutral that figured it out.

2:32:28

>> The other side quoted option price was was way out of bounds.

2:32:31

It was like totally unreasonable.

2:32:33

And then they said, "Your turn."

2:32:35

And what do you do in that situation?

2:32:38

If you respond to that, treat it as a good faith offer and you respond, then you've sort of bought in to that going down that road.

2:32:45

And that seemed to be what the language of the contract on its face, you know, authorized allowed. Exactly.

2:32:55

>> And our advice was no. Uh, don't respond. >> Yeah. >> Don't respond.

2:33:00

And so they went to the arbitrators and said, "Look, they defaulted. They didn't respond.

2:33:04

There's only one price on the table and it's ours." >> Yeah.

2:33:08

>> But ultimately the uh decision of the arbitrators was no.

2:33:10

You know, you didn't act in good faith.

2:33:14

>> Uh so they lost and they're they were stuck in that invest. Ultimately they got out.

2:33:19

They they were so eager to get out of the investment they had to come to terms with it. >> Yeah. Makes a ton of sense.

2:33:23

Well, thank you so much for taking the time to come chat with us. Thank you. And thank the hat there. Enjoy the course.

2:33:28

Thank you for keeping us a couple more.

2:33:31

Keeping us fitted around. Great to see you again. >> I'll see you around. >> Great to see you.

2:33:37

>> Let me tell you about public.

2:33:37

com investing for those who take it seriously.

2:33:40

They got stocks, options, bonds, crypto, treasuries, and more with great customer service.

2:33:45

And let me tell you about Finn.

2:33:48

AI, the number one AI agent for customer service.

2:33:50

If you want AI to handle your customer support, go to fin. ai.

2:33:54

And without further ado, we will begin the Lambda Lightning round. Here we go.

2:34:00

Pivot that camera over there.

2:34:03

It's turning all the way around.

2:34:04

It followed John Quinn out of the studio. >> Missing him already.

2:34:10

>> And now you see that the lambda lightning round has begun.

2:34:11

And we will banger to kick it off. >> An absolute banger.

2:34:15

We have Michael from work. >> There he is. >> Michael from work OS.

2:34:19

He is the co-founder and CEO. What's happening?

2:34:23

>> How you doing, >> gentlemen? Great to see you. How's it going? >> Great to see you. >> Great to see you.

2:34:26

It's been uh what has it been? A couple months.

2:34:30

>> It's been almost 6 months.

2:34:30

Last time, uh we had you on, we were hanging out with Satcha Nadella on a historic day.

2:34:34

Uh also another crazy day to join.

2:34:36

But uh since this is the first time joining remotely, uh give us the update.

2:34:40

Kind of reintroduce yourself for everyone.

2:34:44

>> Yeah, so I'm Michael Greenwich.

2:34:44

I'm the founder and CEO of Work OS.

2:34:46

and we are uh a infrastructure company that helps other software companies with enterprise features in their app.

2:34:53

Not not exactly the thing people get the most excited about but it's the underlying infrastructure that allows people to sign in to things like chatbt and anthropic and perplexity with their company account with their enterprise account.

2:35:06

So we often say that we help developers make their app enterprise ready and today we're really thrilled to announce our series C which we've raised $100 million.

2:35:15

All right, >> we got a mallet. We got a gong to hit. >> Let's see. Let's see. Do we have the power?

2:35:24

>> Bring it down from the lamp.

2:35:25

>> It's like Thor reaching. >> There it is. There it is. >> There we go. Look at that. >> There we go. We can take this. >> Here we go.

2:35:36

>> 100 on uh two billion. >> That's right. That's right.

2:35:40

>> That is uh some great delusion >> from your side.

2:35:44

So we're going we haven't raised money actually in uh over four years.

2:35:48

I think a lot of folks kind of forgot about us to be honest as a company.

2:35:51

Um we were started >> back in uh beginning of 2019.

2:35:53

Um so this is over seven years we've been building building the company.

2:35:58

Really the news today what's new with us is the last year or two has been all about AI and we have found ourselves supporting all these AI companies as they're rapidly just explosively growing.

2:36:09

So whether it's it's like openai selling into the enterprise or you know claude like growing like crazy over the last few months or even cursor last year which you know like kind of came out from nowhere and took over um how people write software.

2:36:23

Workhorse is powering all of these and we're helping all of them take you know the functionality that they've they've landed in AI and actually enable them to go sell that into these big you know customers that otherwise they wouldn't be able to get um and essentially unlock revenue.

2:36:35

So, so over the last few years, you you joked that uh people forgot about you because you weren't uh raising what what were you doing?

2:36:44

I'm assuming you were you were at different points probably turning turning down investor interests like what what was your kind of mindset and leading the team through that period?

2:36:55

>> Yeah, we were just building.

2:36:55

I mean, I think at the end of the day to build infrastructure like we've created, there's no real shortcut for it.

2:36:59

you just have to spend day after day, week after week, month after month, solving tens of thousands of small edge cases.

2:37:07

Sometimes I say work OS, it's kind of like a ball of edges.

2:37:09

Like the whole thing is made of edge cases.

2:37:11

There's there's not really um just one way to, you know, quickly ship it.

2:37:15

It's it's in some ways like the anti-YC company.

2:37:16

It's a thing you can't build >> in like a couple months and launch.

2:37:20

It does take months and months and years and years to develop it.

2:37:24

>> Um early on we had a lot of amazing customers.

2:37:26

customers. I mean we um you know early on powered off for folks like Verscell and Web Flow and Carta and you know kind of the previous era of of cloud SAS we all know about um it's just what we've seen with the AI stuff is it it grows faster um the companies are adopted sooner within the enterprise >> and all this AI functionality it's

2:37:45

actually scrutinized by security IT people you know a lot more a lot more heavily like the IT people say no way we can use your AI unless it fits these you know security policies and so despite being started in the pre-AII era, work OS is actually kind of perfectly positioned to be in we are an AI business today without having been started uh started that way. >> Yeah. How have you how have you >> Yeah.

2:38:03

How have you how have you processed the overall SAS apocalypse narrative or or just vibe coding in general?

2:38:11

In many ways, if uh work OS can can kind of I feel like deliver kind of the dream of vibe coding, which is like maybe there's some uh some specific functionality you want to build.

2:38:22

you want to be able to sell it into the enterprise, but pre-work OS maybe that wasn't uh that wouldn't have been possible because you have all these edge cases that that enterprises are going to care about.

2:38:33

But how have you processed the entire narrative?

2:38:38

>> Yeah, when people are vibe coding stuff, uh you know, you typically don't want to vibe code your security part of your application.

2:38:44

application. And you you might vibe code the features of the UI like like the AI code gen is so good for doing UI engineering but when it comes to stuff like authentications or permissions or things around compliance or auditing that's maybe not the place you want to apply AI and and we've seen that

2:38:58

actually from uh a lot of our customers including our customers being the AI labs that are building the models to do this stuff they're even using using work OS um what we've seen from a lot of other businesses is there's no time to build it inhouse so in the previous era you might have had you like a few years to figure out enterprise. If you go back

2:39:15

If you go back and look at like Dropbox or something, you know, they they built for many many years before they did enterprise.

2:39:20

Today, what's happening is these AI companies, you know, powered by by the functionality, they go after enterprise pretty much immediately.

2:39:26

Like within the first year, they have to go up market.

2:39:30

And so there's no time to build it.

2:39:30

They just turn to us and we ship it for them.

2:39:34

>> So, so I mean truly how low is the bar?

2:39:36

Like if I'm if I'm in YC and I vibe code a piece of software that's like MVP but it works like Is is work OS like accessible to the point where it's self-s serve free for like how how low is the bar? >> Absolutely.

2:39:51

So we have an amazing free tier.

2:39:54

It's free up to 1 million users. >> Wow, that's a lot.

2:39:58

>> So even maybe you get a million users when you're in YC, but past that, you know, as you keep growing, what we charge for is the enterprise features.

2:40:04

And really we try to align our pricing model with you as a customer.

2:40:06

As you go close enterprise deals, you pay work OS.

2:40:10

It's it's kind of like Stripe, you know, you pay Stripe money when you go, you know, make money yourself.

2:40:13

Um, but it's so easy to use.

2:40:15

We have people that integrate, you know, work OS in an hour or two.

2:40:19

>> They're out there selling enterprise essentially as soon as they have demand.

2:40:23

And actually, just last month, we shipped a new capability that's even faster to integrate because it uses AI.

2:40:29

>> So AI is actually accelerating our customer base.

2:40:30

It's also accelerating how fast you can adopt and use use our functionality.

2:40:34

You essentially just run a command, it installs work OS >> and then KPI around that in terms of like integration timelines like it seems like they're getting shorter but what are we talking about like days of developer time, weeks?

2:40:47

>> If you run our CLI installer which uses AI to install, it takes roughly 7 to 8 minutes.

2:40:54

So, it's super fast and I've been doing this trick uh you know whenever I do like a sales call with a with a customer that's interested in using work OS essentially kick off the call and say hey go try this you know just run it in your terminal in the background we'll come back to it after I click around the dashboard a little bit >> and essentially you have a PC that's ready to go you know in the past

2:41:12

>> we we would still integrate fast you know but we would have to talk to their engineering team we get on the road map we we do an architecture call we have a amazing team of solutions engineers and developer success that helps plug stuff But I think what we're finding is AI is this accelerant not just in terms of market adoption but just making software easier to use and integrate. We even

2:41:29

We even have people that migrate using AI.

2:41:30

So they might using a different solution or having a homegrown thing >> you know you plug cloud or codeex into it and just rip through it. It's it's really wild. Yeah.

2:41:40

>> Uh what now is the job finished or or >> job's not done? >> Absolutely not done.

2:41:45

Uh in in some ways it's kind of like a new moment for the company.

2:41:48

I I told our staff at the beginning of this year, you know, um if I was to start work OS today, what I would essentially build is work OS for AI and specifically for AI agents.

2:41:56

So that's what we're building going forward.

2:42:00

You know, if you think about agents in the world, a lot of people talk about agents in different ways.

2:42:04

Really, what they're doing is displacing people doing work or or enabling people to do more.

2:42:08

You know, uh I think Ivan from from Notion has called agents the you're you're a manager of infinite minds.

2:42:14

the ability to go and control and kind of adapt all these different systems to do work for you.

2:42:18

The problem with agents is if you're spawning all these things, these workers to go do things for you, they essentially need permissions.

2:42:25

They need access to data.

2:42:27

And an agent isn't useful if you can't connect it to all your different stuff.

2:42:30

And you can't do that securely.

2:42:30

Last thing you want is an agent, you know, running wild.

2:42:33

I think there was a was a story on Twitter a week or two ago of um an agent an openclaw instance going rogue and deleting a bunch of email. You guys remember that? >> Amazon.

2:42:42

It was like Amazon also had I think some issues like head of alignment or something.

2:42:47

Yeah, >> it was a really wild look.

2:42:50

>> So, you know, you definitely don't want that for your personal account, but it's completely disastrous if you're doing that internally to your systems or god forbid it's doing it to your customers.

2:42:58

And so, this problem around what can agents do?

2:43:00

They're extremely powerful, but we essentially need to give them different types of permissions and guard rails and identity on top of it.

2:43:06

And so that's that's a new thing that we've been building at work OS with some partners building this new essentially identity fabric is what we're calling it that sits across everything and allows people building agents to have the connectivity but also the security and trust that uh is demanded by their customers and we hope with that >> it will act as an accelerant more into the enterprise.

2:43:26

It'll help more companies building AI do it in a way that's safe. >> Amazing.

2:43:29

Uh last question for me um how's the role of you said like integration engineer it sounded like almost forward deployed engineer.

2:43:35

How is that role changing at work OS now?

2:43:41

>> Well, everything at work OS is using AI.

2:43:43

Our customers are using AI.

2:43:43

We're using it every single role.

2:43:45

We have sales reps building stuff with cloud code.

2:43:46

You know, our finance team doing stuff.

2:43:48

We have hackathons going.

2:43:50

And so, it's definitely impacting the people that are working with customers every day in that way.

2:43:55

Um, I think the magic of cloud code and and this for deployed engineer is it essentially turns one person into a whole team.

2:44:01

So it kind of mimics having you know if you go sell product to like a bank like Deutsch Bank your giant company like Microsoft you would be able to afford having like 10 or 15 engineers go sit in their office and write code literally like post up with your laptop and like build stuff for them that that's the the original you know solutions engineer for deployed engineer.

2:44:23

What AI lets people do today is essentially have that same experience but have a tiny company do it you know like work OS.

2:44:28

It essentially lets us be that level of consultative, you know, impactful uh work in your organization.

2:44:35

The first step when we run our AI installer for customers, what it's actually doing behind the scenes is it's looking at your codebase and building a plan.

2:44:41

It's doing an architecture review by analyzing your structure and providing the best, you know, course of action going forward, the best recommendation.

2:44:48

And that I think we're just scratching the surface on that.

2:44:50

I mean, that's like >> it's so early.

2:44:53

It's so so early in terms of all this.

2:44:55

Um, so at the end of the day, it just makes people, you know, integrate faster, go to market sooner, ship faster, and just just grow up market sooner.

2:45:03

We we talk about work OS being the uh help making your product enterprise ready.

2:45:08

>> It's really an API that unlocks revenue.

2:45:10

It expands your TAM as an organization.

2:45:12

And when these companies go through that transition moment, say they just start getting product market fit and expanding, the last thing they want to be doing is is slowed down by the lack of these enterprise features when they have a big fish on the line. Yeah.

2:45:21

You know, and and what works does it turns that on almost immediately.

2:45:24

So if it's seven or eight minutes today, the big fish company of San Francisco. >> That's right. That's right. Yeah.

2:45:31

It's like a fishing line.

2:45:33

It's like we'll help you get it to vote faster. Yeah.

2:45:36

>> Well, thank you so much for taking the time. >> Awesome. Great. Great to get the update.

2:45:39

Congratulations to the whole team >> and we'll talk again. Great to see you guys. Take care. >> Bye.

2:45:44

>> Let me tell you about MongoDB.

2:45:44

What's the only thing faster than the AI market?

2:45:47

Your business on MongoDB. Don't just build AI.

2:45:49

own the data platform that powers it.

2:45:52

Our next guest is Adam Simon from IPG Media.

2:45:56

>> He's in the reream waiting room now. He's in the TV. Adam, how are you doing? >> Oh, >> hello. Welcome. Hold on. No, you're good. >> There we go. Now I can hear you.

2:46:07

>> Uh, >> what's happening? Great to meet you.

2:46:08

>> Thanks so much for hopping on.

2:46:09

>> It's great to meet you. Thanks for having me on.

2:46:11

>> Uh, please, uh, introduce yourself.

2:46:11

Give us a little background. >> Yeah.

2:46:15

Uh, I am, uh, Adam Simon.

2:46:15

I spent a decade as one of the top innovation executives at a global media agency, >> which basically means I got paid to be wrong about the future just a little bit less often than everyone else.

2:46:29

>> Um, and now I am out there consulting on where entertainment is going next.

2:46:31

Um, which in my opinion is off the screen and into the real world.

2:46:37

>> And I'm working on a book uh called The Immersion Economy about how immersive technology is uh set to supercharge the experience economy. >> Okay.

2:46:44

What did you what did you get right when you were predicting the future? >> Oh, good question.

2:46:48

Uh, I got right I I think that uh Netflix would win in streaming and I still maintain uh that after uh you know last Friday's news obviously I I I think that Netflix uh you guys covered this earlier but I think that Netflix is going to come out on top in this deal when we look back a few years from now.

2:47:07

>> Um that was that was something I called early.

2:47:09

I also called that they were going to pivot into um having an ad supported tier at some point as well as into gaming.

2:47:14

I think the Netflix behemoth will just keep growing. >> Yeah.

2:47:18

Uh walk me through a little bit more of like immersive economy off the screen.

2:47:22

Like that could mean Taylor Swift concert, that could mean ice cream museum.

2:47:26

There's a million things that happen off the screen.

2:47:28

Like how do you how do you describe like the territory?

2:47:33

>> Yeah, I mean it's it it is all of those things.

2:47:35

It is uh in my mind it the exciting part is all of those experiences that are about to get are already getting upgraded or are about to get upgraded with technology and some of that looks like immersive technology.

2:47:45

It looks like the sphere.

2:47:47

Uh it looks like what what Kosum is doing which I think is super interesting especially around like sports and and uh performances and stuff.

2:47:56

>> Um but I also think it means uh AI personalization in the physical world.

2:48:01

Um, we're, you know, we spend a lot of time talking about how AI is impacting software.

2:48:05

You guys certainly spend most of your days, I think, talking about that.

2:48:08

Um, but, uh, I think AI colliding with the real world, robotics is the place that most people go.

2:48:13

Um, which is understandable. It's super exciting.

2:48:16

Um, but I think there's an interim step before we see robots everywhere, AI powered robots everywhere, which is just our physical world, uh, being a little more responsive to us as individuals.

2:48:26

If you think about the physical world reacting to you in the same way that your uh your feed on social media might uh and starting to personalize itself in that way that triggers some nightmare scenarios.

2:48:38

But I think there's some opportunity to also do some really cool fun creative work there. >> Okay.

2:48:42

So I'll give you a specific example.

2:48:43

Uh we were in Montana uh we were in Montana this last weekend.

2:48:46

And we were at uh an ice race >> uh which was uh the the first event that they've done where basically they get a bunch of cars around an ice track and then >> you can take cars out and do some laps and then there's a bunch of spectators and things like that.

2:49:02

And I was I was you know my the event was incredible.

2:49:04

My only uh criticism is like there was kind of one screen over in one of the spectator areas and so uh in the area that that uh we were you you kind of you were relying 100% on your own vision to experience the event which was cool but at the same time there were certain angles and like a bunch of miscontext.

2:49:24

And so in an event like how are you thinking about personalization in the real world?

2:49:28

Is it we have smart glasses on and you have your own audio feed or like what does that actually look like in practice at something like a automotive event?

2:49:39

>> Yeah, I mean I think I think that that is a great example.

2:49:42

I think that u to your point we don't even need to wait for augmented reality glasses.

2:49:45

I think when they do come there'll be some really interesting exciting things we could do.

2:49:51

Um but just the ability to customize an audio feed you could do that with AirPods right now, right?

2:49:54

like lots of people already have them in their pockets.

2:49:57

We could be through people's smartphones broadcasting customized audio feeds to you uh through your AirPods just as a one example.

2:50:03

Um an an under sort of discussed feature of the sphere that I find really interesting is they have uh beam forming speakers.

2:50:13

So different sections of the sphere can technically be hearing different audio.

2:50:17

Uh they haven't really used this yet to uh to my knowledge.

2:50:19

Um, but the way that they talk about it is that you could have like different language tracks sort of beamed to different sections of the venue.

2:50:27

Um, and I think that that's the kind of thing that we're seeing built in to lots of venues.

2:50:32

And obviously that's not a personalized sort of onetoone experience, but you can also sort of see how that starts to move us closer in that direction and use some some technology that is, you know, not exactly new, uh, but to to enhance the experience and make it more accessible for more people.

2:50:48

How are you thinking about uh VR?

2:50:50

Uh we had uh Ben Thompson on our show earlier today.

2:50:53

He's been pushing aggressively for Apple to just let him watch any NBA game.

2:50:58

He basically figured out like it's like a $40,000 $40,000 worth of hardware that would just need to be installed at the different stadiums and then you could just He's like I don't need I don't need a separate live stream and commentary because if you just beam me into the stadium I can hear the commentary that's happening.

2:51:18

I can look up at the scoreboard.

2:51:19

I really just need that kind of like video feed. >> Yeah.

2:51:23

I mean, I think, you know, I I think long term there are some cases and some users who are going to want the more highly produced edited experience that Apple is is kind of leaning into.

2:51:32

I know Ben is super against that.

2:51:34

He just wants the the static camera.

2:51:35

It's like pretends that I'm there.

2:51:37

Um, but I I think that he's right in that Apple's biggest problem right now is scale and just having the volume of content that you could get with just static cameras.

2:51:49

And, you know, sports is is great and I I'm sure would sell a lot of headsets.

2:51:52

Um, but also just think about concerts and stage plays and any place else where just being a fly on the wall or being a butt in a seat, uh, you can't get there or don't want to get there physically.

2:52:05

Um, but being able to get there with your Vision Pro or your your MetaQuest or what have you.

2:52:08

Um, I I think that there's real opportunity there.

2:52:10

Longer term, I actually think that venues like Sphere and Kosum in particular um are going to create a pipeline for that kind of immersive content that will be able to to go straight to to headsets because if you think about what KSM is doing with the NBA, they basically have that camera already there courtside.

2:52:28

Um, and it's creating that immersive sort of shared realities, what they call it, experience in the COSM venues.

2:52:34

There's no reason you couldn't just port that over to a vision pro right now.

2:52:38

So, we're starting to see some pieces be put in place that's going to create a really interesting, I think, flywheel for immersive content.

2:52:45

And I sort of think about it as uh the venues like KASM being a kind of middle-class experience where if you have the money and the capability, you're going to go there in person, of course, but you're only going to do that for a few events a year, even if you have uh deep pockets.

2:52:58

Uh and obviously there's the at home viewing experience, which might be upgraded uh if we can watch it in immersive uh environments.

2:53:08

Um and then you have the option to go someplace like a KSM venue to get this sort of, you know, halfway experience.

2:53:14

I can still go out with my friends and be social and have a close to courtside experience without shelling out to to fly wherever the game is. >> Yeah.

2:53:23

Are there are there a lot more spheres caus uh experiences coming down the pipeline?

2:53:31

uh they they said they're super capital intensive so they're high risk but are from my view the response to to both has been great and Sphere seems to be >> somehow doing pretty well in uh in the public markets >> as of late.

2:53:47

>> Yeah, I think Sphere was caught everybody by surprise a little bit because that that first year there were there were lots of headlines about how much money they were losing, how much it costs not just to build but to operate.

2:53:57

Um but from everything I can tell they from the outside obviously they look like they have found a path.

2:54:02

Uh I think if you look at what they're they did with Wiz Wizard of Oz uh which was a surprise uh I think to everyone in Hollywood um it it is really successful.

2:54:14

Um they basically made like a fan experience for Wizard of Oz which gave them a reason to charge you over $100 a ticket to see a movie that you've probably seen dozens of times before.

2:54:24

And uh I don't know if it's streaming anywhere for free, but it used to be on television all the time, right?

2:54:27

It's it's like the classic movie that was was constantly available.

2:54:31

And yet people are shelling out over $100 a ticket to go see it.

2:54:35

Um they've sold over 2 million tickets to that already.

2:54:36

So I think it's an it's not exactly what they set out to do, but it's an interesting new path for them and clearly source of revenue. >> Yeah.

2:54:46

Uh uh the Brian Chesy from Airbnb was on the show uh fairly recently talking about how he is quite bullish on on IRL in the age of AI.

2:54:58

The the online world is getting very wild and intense.

2:55:00

Uh you can imagine some people just deciding I'm going to log off.

2:55:05

I'm going to go get go somewhere get away. I can see that.

2:55:09

I can also see the other side of it, which is that the online world continues to get more and more entertaining and some people just say like, well, I'm not I don't need to go take this trip or go to this experience because I'm perfectly entertained at home.

2:55:24

What indicators are you looking at to understand if if AI is actually disrupting some IRL experiences?

2:55:30

Would it show up as like Disneyland, you know, uh, visits dropping, which obviously could be a number of factors, like how would you where are you kind of like trying to see where there might be some disruption?

2:55:45

>> Yeah, I mean, I think I would be looking at things like live streamed sports and concerts and things like that to see are the ticket sales going down.

2:55:52

Everything we've seen to date is quite the opposite.

2:55:56

People are spending like crazy on experiences, right?

2:55:58

Um it's it's everybody thought that that it was sort of a revenge spending after the pandemic.

2:56:05

Um but that has has not played out.

2:56:07

Everybody is very excited to at least sometimes right like close their laptops, put their phone in their pocket uh and go out and do something fun.

2:56:14

Um I think what we're competing with is that people are they they want to uh be out there having experiences creating memories.

2:56:23

uh and that is not something that you're going to have the same sort of experience doing uh looking at a screen.

2:56:29

There's also some interesting um uh neurological research that I' I've been following um around just like how we sort of sync up and experience communal events.

2:56:39

Uh and it's just it's really fascinating.

2:56:42

But basically there's uh we we create stronger memories and we form stronger social bomb bonds when we're in person.

2:56:49

and it actually has a biological component to it.

2:56:53

It's not just sort of the emotional um aspect of being there together.

2:56:55

So I think you know maybe if the pandemic was the catalyst that we realized what what we were missing.

2:57:02

Um but I'm I'm pretty bullish on the fact that uh demand for um you know even techn technology or um optimized and oriented and enhanced experiences in the real world like just because we're out in the real world doesn't mean the tech goes away.

2:57:17

It just changes and evolves.

2:57:19

And I think that's where a lot of the exciting developments uh both for AI but also for for other technologies is going to be in the next decade.

2:57:27

>> Well, we got to do the next one in person with a live audience because just neurologically I feel like we'll have a we'll have a better time. >> Yeah.

2:57:35

There's also the the reality that uh many people are chasing IRL experiences to have something to show back on social media in this in this uh you know I I I try to remind myself you know we were at this this car event and >> everywhere you looked there were cars that if I saw them on the street I would just stop whatever I was doing and I was like trying to force myself to actually just like take it all in versus taking pictures.

2:58:03

But certainly there were some people >> that uh that that hardly were processing it live because they just wanted to share everything that they were seeing. >> Me too. Yeah.

2:58:11

>> And honestly that is the thing that I'm most excited about with with glasses whether like me ray bands or what Apple's going to release later this year or early next year.

2:58:20

>> Having the ability to capture content but stay more present in the moment I think is the best uh sort of marketing case for those products because you're absolutely right like we we want to be able to capture these things.

2:58:31

Um, but we've, I think, tilted a little bit towards uh doing things just for the sake of capturing them rather than uh actually enjoying the experiences.

2:58:39

And I look, I think everybody knows that even when you're doing it, to your point, you you're aware of it.

2:58:43

So, I think if there's the right catalyst, it'll get people off their phones and and more sort of focused on what's around them. >> Awesome. Totally.

2:58:51

>> Well, thank you so much for taking >> Yeah.

2:58:52

Great to meet and and come back on uh anytime. >> Have a good one. >> Awesome. Great to meet you. Thanks, guys.

2:58:57

Let me tell you about app loving profitable advertising made easy with Axon. ai.

2:59:01

Get access to over 1 billion daily active users and grow your business today.

2:59:06

Uh we have been keeping our next guest waiting.

2:59:08

We have Matias Wagner from Flux. ai.

2:59:10

He's the founder and CEO and he has some exciting news for us. >> What's happening? >> Welcome to the show. How are you doing?

2:59:17

Please introduce yourself to the company. >> Hi guys. Thanks for having me. My name is Matias.

2:59:21

I'm the founder CEO of Flux and you know we're building the first AI hardware engineer. >> Amazing.

2:59:25

And give us the news today. You raised some money. What happened?

2:59:29

>> Yeah, we closed the series B.

2:59:29

We raised a total of uh $30 million in new capital >> across congratulations. >> Led by ABC.

2:59:42

So, you know, super stoked and uh you know, stepping on the gas here now. >> Yeah.

2:59:47

Talk about the key problem, the key uh the key product, where you are in development, rolling out uh getting product in the hands of customers. >> Yeah, good question.

2:59:56

So I mean you know you've all heard it before hardware is hard. >> Yeah.

3:00:00

>> And uh we felt like this is like learned uh helplessness you know >> and we're taking the heart out of hardware right?

3:00:06

So if you think here what >> it's just wear now.

3:00:10

>> It's just where >> exact right >> but you know we we're starting with electronics here um cuz you know it's like a very standardized form factor and it's really at the heart of anything that's worthwhile building these days if you think about you know computers, robots, uh what have you.

3:00:23

Um, and the insight here really was like, right, that if you think about how easy it's become over the last two, three decades to make software, especially now with the with the AI boom, right?

3:00:36

>> Hardware hasn't gotten any easier to make. >> Yeah. >> Right.

3:00:38

Even though the supply chain is incredibly available now, I mean, I can like from anywhere in the world for like $20 can get a PCB manufactured in China and sent back in seven days fully assembled. Right.

3:00:49

>> And that that you couldn't do 20 years ago.

3:00:50

you had to be Loheed Martin or Apple or like a big tech company to afford that.

3:00:54

But now everybody can do that.

3:00:55

But the the design tools just don't exist and they're not accessible.

3:00:58

And that's what we're solving for.

3:01:00

>> So talk about the inputs and the outputs.

3:01:02

Obviously at the end I get some sort of PCB spec is that just a CAD file or a PDF?

3:01:07

Uh and then what am I actually putting in?

3:01:09

Is it just software that then I want to be translated into hardware?

3:01:12

Like at what level of abstraction are we operating here? >> Yeah, great question.

3:01:16

>> Yeah, great question. So look the vision is that you can go like if you think about catch you can go from like an a prompt to a poem y >> right or like clock code you can go from a prompt to a bug fix or a feature with flux right ultimately we want to enable you to go from a prompt to an iPhone

3:01:31

class device right now that's a long road right uh so what we are today is like you can make like what's like mid small to make complexity you can probably single shot today and everything else is going to be more iterative you know again like cloud code devon cursor all these tools very very comparable Sure, it makes sense. Uh,

3:01:46

Uh, who's the best customer for you right now?

3:01:50

Where do you see that going?

3:01:50

Are you going after startups that are iterating really quickly?

3:01:54

This is speeding them up or are you going to large organizations that already turnurning out tons and tons of PCB designs? >> Yeah.

3:02:02

So, it's a good mix, but most customers are SMBs.

3:02:04

What we're seeing is that there's like, you know, only about 25% of our customers have previously made a PCB themselves. >> Sure.

3:02:11

>> The others are technical.

3:02:11

There are mechanical engineers, firmware engineers, industrial designers, right?

3:02:16

But they didn't have the time, patience or resources before to make custom boards and they would obviously previously buy OEM boards, right?

3:02:21

If you think about like a favorite example I have, we have this customer, they make vending machines like snack, beverage vending machines and they would previously buy four or five off-the-shelf boards and integrate them into a single machines.

3:02:32

So think about the board that powers the display, the payment, the motor in here and so on.

3:02:38

Um, and now they can make like a custom like a single custom board with flux, right?

3:02:41

that's much cheaper to integrate into the machines on the assembly line, the machine breaks down in the field, right?

3:02:47

You know, where they're standing there like somewhere in the rain, you know, there's one board to replace and then, right, the board costs in pennies on the dollars compared to before, right?

3:02:54

Because what you're paying now is material cost, you know, to a manufacturer in China, right? I can do that for you.

3:02:59

And so that's kind of like here the the main driver, I think.

3:03:03

>> What's your favorite printed circuit board or favorite product that has a PCB in it?

3:03:09

>> Ooh, the space shuttle. >> The space shuttle. That's a good answer.

3:03:13

>> No, I looked this up the other day.

3:03:13

So, the space shuttle uh I think was one of the first boards that had like eight layers or so, right, at the time and it was handdrawn.

3:03:21

They didn't have software for it at the time, right?

3:03:22

Which is like crazy if you think about that.

3:03:26

>> Um, so yeah, this is like a fun anecdote. >> That's a great one. Jord, anything else? >> Wild, wild, wild.

3:03:31

Uh how have you seen anything exciting on the US side on the supply chain side >> on actual PCB >> manufacturing?

3:03:41

Is there I feel like we had somebody on the show last year that was trying to make uh PCB >> like we have send cut send now for uh for like machined parts and and things that you need to cut.

3:03:51

Is there is there a great company in America that's starting to think about PCB?

3:03:57

>> Yeah, like if we have our own Shenzen, I imagine it would have a variety of >> Yeah.

3:04:00

manufacturers like this to really because you want to get to the point where you can kind of sit next to the manufacturer and really speed up that iteration cycle. >> Yeah. Yeah.

3:04:09

I mean, look, a lot of people like us want that, right?

3:04:11

We looked into the details of like opening our own fap here in Fremont or so, but you realize quickly, right, making PCB boards is like the board itself is a chemical process.

3:04:20

Good luck getting the permits, right?

3:04:22

It's possible, but it's hard, right?

3:04:25

And so, these are kind of like the bumps you run into.

3:04:27

But I think there is exciting opportunities here, right?

3:04:31

>> Um especially because like if you're designing these boards with AI, right?

3:04:34

Then you can also optimize for the capabilities uh and the inventory you have at the factory, >> right?

3:04:40

Because the way to make this cheap to make is to like design it so it can be assembled automatically by machines, right?

3:04:45

But for that to work, essentially these pick and place machines, they have to have all the semiconductors you need uh on their on their rails, right?

3:04:52

Um >> and and that's difficult to do for humans to optimize for, right?

3:04:56

because it's like a quick changing thing.

3:04:57

But I think as we're automating all this, right, we can have the models designed towards what the what the factory can spit out in an hour, you know, for for almost zero cost. >> That's amazing. Congratulations.

3:05:08

Thank you so much for taking the time to come talk to us >> with with this pace.

3:05:12

I'm sure you'll be back on this year.

3:05:14

>> Yeah, this is amazing. >> All the progress. >> I hope so.

3:05:17

>> Have a great rest of your day. We'll talk to you soon.

3:05:20

>> Let me tell you about Labelbox.

3:05:20

RL environments, voice, robotics, evals, and expert human data.

3:05:24

Labelbox is the data factory behind the world's leading AI teams.

3:05:27

Up next, we have some exciting news from Zach from Cal AI. He's co-founder and CEO. Oh, is he not here?

3:05:37

>> We are getting him set up. >> Sorry about that.

3:05:40

>> We can head >> We can head to the timeline because we have an incredible uh announcement from Riley Walls, the new OpenAI staffer who uh launched Payphone Go.

3:05:49

It's uh effectively Pokémon Go, but for Payoneses.

3:05:54

Payones are strangely still licensed in California, he says.

3:05:56

So, I filed a FOYA request and got the full list.

3:06:01

Naturally, I made a game so you can now play. You create an account.

3:06:03

You get your unique player ID.

3:06:06

It's a nine-digit number.

3:06:08

Then, you use the map to locate one of California's payoneses.

3:06:13

Some are easy to find, some are not. Pick up the receiver.

3:06:15

you dial this phone number 88868366-97.

3:06:21

It's toll-free, so no coins required.

3:06:23

You can just go to any pay phone that you can find.

3:06:25

And then you call in and you enter your player ID and you collect them all.

3:06:29

You have to catch them all.

3:06:31

All 2,23 payoneses in California are at stake.

3:06:36

And Riley is on another absolute generational run with these projects. I love these. They are so much fun. Go check it out.

3:06:44

And congrats to Riley on another banger project. Well, >> all right.

3:06:48

>> I believe we have our next guest, but you >> But Zach is going to be later.

3:06:52

>> Zach will be later, so we're going to jump over to >> What's going on? >> How you doing?

3:06:56

>> Sorry for the confusion. >> Hey, how's it going? >> No problem. Nice to meet you guys. >> Good to meet you.

3:07:00

Please introduce yourself and the company. >> Yeah. Um, I'm Joan Rodriguez.

3:07:04

I'm from Barcelona, Catalan.

3:07:06

I'm the CEO and founder of Quiver AI. >> Amazing.

3:07:10

And at Quiver, we are you know uh building uh models for AI generation of SVGs. >> Okay. >> Um vector graphics.

3:07:18

Uh you can put a text prom and you can get a veto graphic >> or you can put an image and you can vectorize it. Love it.

3:07:25

>> So it's really good for designers, really good for icon generation, logos, and so on.

3:07:31

>> PNG to vector to SVG was such a hassle like a decade ago, maybe even like five years ago.

3:07:36

It's still it's still not perfect. What's the secret?

3:07:38

because I've seen these demos from I think the latest Gemini 3.

3:07:42

1 was doing some SVG drawings. It was pretty good.

3:07:45

What's the secret to going superhuman in SVG generation?

3:07:53

>> Well, that's a great question.

3:07:53

It's a pretty challenging problem and we are tackling this through a kind of new paradigm.

3:07:58

Uh so it's a code generation.

3:08:01

>> Okay, >> so an SVG is actually code in the back.

3:08:04

You know, it's kind of a bunch of code and you compile it and you can get an image out of it.

3:08:07

Uh and previous methods were mostly like tracing, you know, like you get an image and you kind of trace it.

3:08:13

But um yeah, I did this PhD in SVG and I invented this method where you can actually get really cool SVGs >> through code generation.

3:08:22

So you know, LLMs are really good at code. We know this.

3:08:25

And I just went nuts in my PhD to try to train these models. >> I love it. >> As soon as I could.

3:08:31

I imagine part of the benefit is that once you have an SVG, you can export that to Figma or Illustrator and then edit the and do the final if you're if your output's 99%.

3:08:39

Uh much better than getting an PNG that's 99% then you got to go and clean everything up and it takes a lot longer.

3:08:48

You can adjust the actual splines. Is that correct?

3:08:51

>> That's that's correct.

3:08:51

Yeah, you can edit you can change the colors.

3:08:53

You can maybe do animations later. >> Oh, sure.

3:08:57

>> Um yeah, Gemini is really good for that, too.

3:08:59

like animations is something that's popping up a lot in design X. So yeah, absolutely.

3:09:04

>> What do you what do you how are you thinking about uh competition long term?

3:09:09

Uh sounds like you guys can kind of lead in this space, but at the same time I imagine all the labs like kind of care about this category.

3:09:16

Do you imagine partnering with them uh or or having a different uh distribution strategy?

3:09:22

What does that look like?

3:09:26

>> I mean it's exciting to see all the good results.

3:09:28

They are also evaluating our ideas.

3:09:30

We have ideas to try and the big labs are you know showing really good results on the SVG space.

3:09:35

We are seeing that um if you just train a big model with a lot of data and just a bit of effort you are going to see really good results on this space and what we are doing yeah you're we're kind of doing this our own thing our own way with our own RL rewards one at a time and trying to learn how to do this the right way.

3:09:56

>> How do you think about the customer?

3:09:56

Do you want to integrate with a a piece of software that already runs and lives and breathes SVG or do you want to be sort of like this one-off tool that lives on a website that you know artists and creators can go interface with and then just take the assets elsewhere? >> Yeah, great question.

3:10:14

And so we started with a customer kind of consumer approach.

3:10:18

We built our website >> and we wanted to see what people do with these tools.

3:10:23

But at the same time, we build our API so that other tools can also interact with our um with our models and build on top of it.

3:10:29

We can see agents already being deployed.

3:10:32

We can see MCP servers calling Quiver >> and we are seeing you know all sorts of tools integrations with um all your um you know kind of stack of tools for designers. Yeah.

3:10:45

And that's what we are aiming for to be, you know, very close to the sack and and workflows of designers.

3:10:51

>> Yeah, this is Christmas for After Effects artists.

3:10:52

I've I've played with a lot of this stuff. It's a lot of fun. Uh, congrats, Jordy. Anything else?

3:10:57

>> Where uh where's the company based?

3:10:59

>> Yeah, >> it's going to be based in San Francisco. >> Cool.

3:11:02

>> But you're calling calling in from uh Barcelona or or That's right. Not yet.

3:11:06

In the process of >> last thing, you raised some money. How much did you raise? Yeah, we just raised $8. 3 million from less. Congratulations.

3:11:18

>> I'm so happy that this is funded.

3:11:18

This is uh this is very >> great.

3:11:22

Great use of your PhD time, too.

3:11:25

You're I feels like you're obviously just getting started with Quiver, but been at it for quite a while. It's very cool. >> Very cool.

3:11:31

We will talk to you soon.

3:11:32

Have a good rest of your day. Great to >> meet. Cheers. Beautiful. >> Goodbye.

3:11:35

Let me tell you about vibe. co.

3:11:36

where DDC brands, B2B startups, and AI companies advertise on streaming TV.

3:11:40

Pick channels, target audiences, and measure sales just like on Meta.

3:11:41

And we have our next guest in the race room waiting room.

3:11:46

Let's bring him in to the TV Ultra Jump. >> Zack, finally. How you doing? >> I'm good. How are you guys? >> Congratulations. Tell us what happened.

3:11:56

I want to hit the gong for you. >> Well, thank you.

3:11:57

After a year and a half journey building Cali, we were finally acquired by my fitness pal. Um, >> thank you. Thank you.

3:12:10

>> When when did you meet the team at Maui Fitness? >> Yeah. Yeah. Break down the deal.

3:12:14

>> We >> Yeah, we spoke to them initially back in May.

3:12:17

I was talking to a few companies, reaching out to them before going to college because it could have been if we got the right offer, it would have I think made sense.

3:12:25

I could have maybe had a normal college life for a little is what I was thinking.

3:12:28

We didn't move forward at that point with anything, but built those relationships and then kept in touch and when it finally made sense, we moved forward. >> That's awesome.

3:12:39

>> And and I I think I saw in your announcement you closed it while you were 18 and you're just announcing it now. Is that correct? >> Yeah.

3:12:46

So, we closed it in December when I was 18 and then I turned 19 in January.

3:12:50

So, now it's >> brutal, dude. You're gone. You're now >> It's over. It's over now.

3:12:58

>> Uh g give us the scale of the business, the monetization.

3:13:00

Uh I mean for describe it for those who don't know and then I and then I want to know like how did you grow this thing? How big did it get? >> Great question.

3:13:10

Cali is an app where you could track your calories just by taking a picture of your food.

3:13:13

And we did $30 million in revenue in 2025. >> Wow.

3:13:21

>> And we just finished January.

3:13:21

Yeah, I didn't check February, but January we did 5. 7 million in revenue.

3:13:25

It's a subscription model. >> That is incredible.

3:13:29

Uh, so heavily dependent on iOS in app subscriptions. Are you Android as well?

3:13:34

How how important has that been?

3:13:38

>> Android is definitely less important.

3:13:40

It's maybe a tenth of our revenue.

3:13:40

It pretty heavily is just iOS. >> Yeah. And then acquisition. Uh, what is this?

3:13:46

Influencer driven, ad driven, uh, customer. >> Yeah. Customer acquisition.

3:13:49

I mean uh uh obviously you're on the charts but uh are there viral mechanics like what is the growth engine for the app?

3:13:58

>> Early on it was mainly influencers and then as of recently the last maybe 6 months we've primarily continued our growth through paid acquisition on channels like Meta and so the two combined are really what's driving it. >> Makes sense.

3:14:11

What uh why what what are the kind of key reasons why My Fitness Pal wanted to pick you guys up?

3:14:16

I can think of a lot but uh what what what drove this?

3:14:24

>> Well, definitely many many reasons.

3:14:28

We have different audiences, different core audiences, but then we also have overlap >> and we both are very missiondriven.

3:14:36

We both just want to help the greatest number of people as possible become healthier.

3:14:40

And so since we were since we have this overlapping space, we realized that it would make sense to come together, share resources, and it would further the mission for both of us and really makes sense to just partner. >> Talk about accuracy.

3:14:56

I think there's some people that are skeptical that AI can take a picture of food and tell you how many calories are in it.

3:15:02

Uh how has that evolved over the year or so that you've been running the company?

3:15:06

Like how accurate was it at the start? How accurate is it now?

3:15:09

Where do you expect it to go?

3:15:12

The accuracy of Cali early on it definitely wasn't as good as it is today.

3:15:18

We initially started by just running CHBT by itself pretty much as the model.

3:15:23

You would just take the picture would put it into CHBT say how many calories is this? >> Yeah.

3:15:28

>> Output the response total rapper victory. >> Total rapper victory.

3:15:31

Honestly, congratulations. It's so amazing.

3:15:35

>> Well, that was like a month. Yeah.

3:15:36

>> And then we fine-tuned a model.

3:15:36

We built a more complex pipeline using like a ton of different models for different steps.

3:15:43

And so it's become more and more accurate over time.

3:15:45

At this point, we average over 90% accuracy for each food scan.

3:15:50

But you would actually be surprised only 30% of the scans are only 30% of food logs come from scanning.

3:15:56

The rest are like barcodes typing in in the database.

3:16:00

So that those are the most accurate methods.

3:16:02

>> That makes a ton of sense.

3:16:02

That makes a ton of sense. >> Very cool.

3:16:04