searchlore

0:06

[Music] Hey, hey, hey. Hey. Hey. Hey.

0:29

[Music] [Music] [Music] Hey.

1:14

[Music] [Music] Hey, [Music] hey, hey.

1:46

[Music] [Music] [Music] Triple glaze.

2:14

I see multiple journalists on the horizon. Stand by.

2:19

[Music] [Music] Heat [Music] up [Music] here.

3:01

[Music] We came to this world.

3:25

We came to this world to shape our future. We came to this world.

3:39

[Music] [Music] >> Large journalistic force headed towards you. Stand by.

3:54

[Music] Market clearing order inbound.

4:15

[Music] [Music] Team Deathmatch.

4:47

>> Delta team, you are clear to engage.

4:57

[Music] You're watching TVPN.

5:00

Today is Wednesday, August 27th, 2025.

5:02

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

5:09

Uh massive news from SpaceX last night. >> They did it. >> They did it.

5:15

Uh flight 10 made it back.

5:19

took off from Starbase, South Texas.

5:19

The uh the booster landed in the ocean in the I think the Gulf of America technically.

5:31

>> That's absolutely right.

5:32

>> And the actual Starship executed a successful flip maneuver, deployed a mock payload of a bunch of uh nonfunctional Starlink satellites.

5:41

Uh survived re-entry just barely, as we'll see.

5:47

It was a really dramatic video. It was amazing.

5:50

And then splashdown on target in the Indian Ocean. >> It was a full send. >> It was a full send.

5:54

It was It was a crazy crazy watch. Uh yeah.

5:57

Let's watch the recap video that the Wall Street Journal put together.

6:01

This is pretty interesting. >> Starbase, Texas. This is yesterday.

6:05

They scrapped two attempts on Monday and Tuesday because of uh some some problems with the the vehicle, but they got it off the ground. Look at that. They lit the candle. >> Go into space.

6:22

>> And so >> I wish we could get a better sense for just how massive it is from those angles.

6:26

>> And so that those are the dummy Starlink satellites.

6:28

The >> dummy that little motor conveyor belt on the left. >> Yeah.

6:32

Just the number of things that have to go right to actually release something from this rocket is insane.

6:39

And then this is the crazy part.

6:39

There's just a bunch of different Where did the stuff come from? >> Just take a hit.

6:44

We clearly didn't need it.

6:47

>> You would think that as soon as I saw I was like, "Okay, it's over.

6:49

The whole thing is going to blow for sure." It does.

6:51

It definitely doesn't make it back >> because if it's like losing parts, like every part should be important, right?

6:58

Like what's going on here?

7:00

>> Um, and so and they're also the whole time they were like testing the the flaps, like really pushing them to the limits, going farther.

7:06

There it is splashing down in the ocean.

7:08

you you were mentioning earlier there's a sound bite of them saying they were they were actually going further than they needed to.

7:14

So they were doing things like like okay let let's >> and then boom >> and then the >> convenient cut.

7:21

You would have thought the journal and the legacy media would have let that happen a little bit longer of the massive explosion. >> Yeah.

7:28

I mean the vibes the vibes going into this were rough honestly.

7:30

Uh lots of people with the Elon's been distracted by politics take.

7:34

Lots of people with the Elon too focused on AI romantic companions take uh we covered that.

7:41

>> You can imagine the conversation between Elon and Anie yesterday right before the launch. >> You got this babe.

7:47

>> I mean you know that uh you know that Anie uh apparently uh allegedly the uh the original Anie romantic companion model runs on uh the non-reasoning model. Just Grock 3, I guess. >> Just pure love. >> Pure love.

8:04

and and if if there's any uh evidence out there in the just market of AI chatting and companions is that you don't need a reasoning model when you are chatting with a romantic companion.

8:15

So uh a lot of like the backlash to the GPT5 launch was something like uh you know people were using it not just as a romantic companion but just as like a life coach, somebody to talk to.

8:24

It was very conversational and that quick back and forth was what people actually enjoyed. They like the tone.

8:30

They didn't care that it wasn't, you know, incredibly good at solving hard math problems or like the IMO or whatever.

8:38

What you got for me, Tyler?

8:38

I >> I think there's also something to be said for like um reasoning models kind of mess with the vibes, >> right?

8:44

It's like it's like the music thing where >> um >> sometimes you want a friend that's not going to think too hard.

8:48

They're just going to in the musical um like taste thing. Yep.

8:54

>> All the reasoning models gave like these weird answers because they would like subtly prefer, you know, artist names with numbers, stuff like that. Yep.

9:01

Although I mean I think the reasoning models actually had the best taste of all, but that's a bit of a hot take.

9:07

>> Well, we got to pull up this post from Red Bull Futurist. Very important. He says, "Super heavy. The effing goat." Wow.

9:14

>> And I just needed to highlight this cuz obviously Red Bull uh some of Red Bull Futurist work actually has uh went into this court. So absolute legend. A win for Red Bull. Really? >> For sure. For sure. Um, yeah.

9:28

>> I'm s It's actually surprising that that uh Red Bull and SpaceX haven't tried to collab yet.

9:34

You'd think that Red Bull would just be like, "We'll give you millions of dollars if you just put Red Bull branding over this." Like, >> why not? It'd be good.

9:41

It'd be good for both parties.

9:44

>> Even Even the Explosions are feel like on brand for Red Bull.

9:46

Red Bull's down for the crazy risk taking.

9:48

>> I could see Elon going with White Monster.

9:51

>> That would be really good.

9:55

>> They really should do that.

9:55

entire advertising is special and maybe you don't want to delete that.

9:59

>> I mean are would you go so far to say the white monster brand isn't special?

10:03

Would you say I guess it would be Red Bull? Yeah.

10:05

Um they uh >> the reason I was bringing up the reasoning models was because apparently um uh Elon wanted Anie to use the reasoning model like the the the AI the AI companion has to be able to do the hard math problems which is funny.

10:19

So, so, so maybe maybe she really was uh in use uh making doing rocket science and and calculating the last uh final decisions before this uh this rocket went out, but it was a huge success.

10:30

Um the Wall Street Journal has kind of the playbyplay here.

10:33

Uh SpaceX p SpaceX pulled off a smoother test of its Starship rocket, managing a more complete mission after setbacks earlier this year.

10:43

The Starship uh spacecraft flew through space after launching from the company's South Texas complex Tuesday around 7:30 p. m. ET.

10:48

The vehicle successfully deployed a batch of dummy Starlink satellites and re-entered Earth's orbit facing intense heat and forces before splashing down in the Indian Ocean.

10:58

This was the 10th launch of Starship.

11:01

I believe this was the this was ship like 37 or something.

11:03

They've built a lot of these.

11:05

Uh not all of them have launched.

11:08

Some of them have been scrapped.

11:09

Um some of them blew up on the test stand.

11:10

Like they really push these things super hard.

11:12

So, it's a 403 foot tall rocket.

11:15

They started testing this in 2023.

11:16

Uh there's been a series of explosions, mishaps, uh previous test flights that have been cut short.

11:22

Uh Elon Musk has much riding on the rocket envisioned to one day carry satellites, scientific devices, and eventually astronauts.

11:30

>> Just to put this into context, the White House is 70 feet tall. >> Wow.

11:34

>> So, this is >> Yeah, it's it's like the size of the Isn't it closer to the uh the Statue of Liberty?

11:39

I think it's like around that tall. 403 feet. I have to look that up.

11:43

Uh >> the Statue of Liberty is 305 ft.

11:47

>> Absolutely mo an extra 50 feet on that.

11:50

Uh for now, the company is pushing to show it can consistently fly Starship and experiment with its design.

11:55

Space up SpaceX set up Tuesday's mission to stress test parts of the Starship spacecraft.

12:01

We certainly saw that uh on display with the parts flying all over the place.

12:05

Uh seeking to give engineers information to continue developing the vehicle.

12:09

Basically the the the the calculus here is like how cheap can you make this thing?

12:13

If you use, you know, duct tape, will that work?

12:16

Will if you use tin foil, will work?

12:18

Like they really are pushing these things like crazy crazy hard.

12:21

You >> I think the other the other way to put this into context, uh not everybody is fortunate to be able to launch uh build and launch rockets for a living, but anybody that's built software knows that even if you're launching a relatively simple feature Oh, yeah.

12:34

>> it'll often times break and have issues and bugs and things like that.

12:38

it's a little bit um you have a little bit more leeway to ship stuff and make changes on the fly.

12:45

>> Um in this case, they're doing it in this incredibly public uh setting where uh the internet is going to have a a very strong reaction one way or another.

12:54

So the pressure is just insane.

12:57

>> So uh Musk wants both the Starship booster and spacecraft to be fully and rapidly reusable, more akin to an airplane than a traditional rocket.

13:05

SpaceX has made major strides with reusability with its workhorse Falcon 9 rockets, which are powered by boosters that can be used many times.

13:10

SpaceX faces self-imposed and external deadlines with the much larger Starship.

13:16

Musk wants to launch an uncrrewed version of the vehicle to Mars next year, but has said meeting that goal will be tough.

13:22

Of course, there's a Mars transfer window that only comes up, I think, every 18 months or so.

13:25

So, it's really critical to hit it in 2026, or else you got to wait till 2028, I believe.

13:31

Uh, in 2027, a lander variant of Starship is supposed to be ready to carry astronauts on a NASA moon mission through the AY's Aremis exploration program.

13:41

That's going to be really hard because I think they have to refuel in orbit and then also make it safe enough that astronauts can fly on it.

13:46

Like they have a lot to do there.

13:48

>> This is why we need to get Anie to be able to solve complex physics. >> For sure. For sure.

13:52

Uh, and so, you know, Elon's obsessed with saving cost, saving time and money. You should go to ramp. com. Save time and money. Time is money. Save both.

14:00

We are talking >> easy use corporate cards, bill, payments, accounting, and a whole lot more all in one place. Go to ramp. com.

14:08

Speaking of RAM, Ramp of course sponsors the Founders Podcast, which David Senra had a fantastic episode that just dropped called How Elon Works.

14:14

And it really contextualized how insane it must be to work on the Starship project.

14:18

And there's three quotes that I want to run through here that I absolutely loved.

14:23

So, first when an when an engineer told Elon the air cooling system for the Falcon 9 would cost $3 million, he shouted over to Gwen Shotwell, Gwen Shotwell uh to ask her what an air conditioning system for a house cost, about $6,000, she said.

14:37

So, the SpaceX team bought some commercial air conditioning units and modified their pumps so they could work a top the rocket.

14:45

And like my take is like like yeah obviously like 3 million is a lot. 6,000 is like nothing.

14:51

But like if you're spending $3 million on the air cooling system like it probably has been tested in aerospace environment.

14:57

It probably comes with a team to help you get it working reliably.

15:00

There's probably lots of nice to haves that go into that.

15:03

And >> yeah, this is this is I mean you know you would a lot of companies that have this much access to capital would not be fixated on things like this.

15:11

They would just say what's the best solution for our problem?

15:17

And so again that fixated cost like most billionaires with a space are just going to say well let's use the best of the best. >> Exactly.

15:27

Let's use the best of the best and let's just do one.

15:28

Let's do one let's do the first rocket and then let's worry about cost control later.

15:32

And that's very different.

15:34

So um you can probably like spending the $3 million probably speeds up the time to get a single rocket to space.

15:40

But that's not the goal here.

15:42

Elon's obsessed with building a system that will get to space >> like over and over and over again.

15:46

And so there's another quote here that's wild.

15:49

Uh Elon's just consistently questioning things.

15:51

This is David Senra talking um and trying to find the limit and then he's quoting Elon.

15:56

Why do we have to have four bolts there?

15:58

Who set that specification? Can we do it with two? They would say no. He said we'll try it. See if it fails.

16:03

And then on and on and on and he just moves down the line.

16:07

You need to be decisive or you're going to be dead.

16:09

Elon calculated that he made a hundred command decisions a day as he walked the floor.

16:13

At least 20% are going to be wrong, he said.

16:15

But we're going to alter them later, but if I don't make decisions, we die.

16:18

And so, like, it's very clear that looking back at the Starship tests, like there are so many gamles going on and so many design decisions that are basically irrational if you're just trying to get one to space, but you're really building up like playing the game on hard minimum bare minimum the cheapest possible rocket to get there.

16:40

And he's probably just adding cost just dollar by dollar like, okay, it completely blew up.

16:44

Let's add one more thing that might stop it from blowing up.

16:50

Like as opposed to the opposite which is like which is like >> that can never happen everything that stops it from any any system or feature or function or part that stops it from blowing up.

17:00

Start there and then maybe we'll pull one of those out if we're really sure that it can not blow up without it.

17:05

Instead it's like it's going to blow up.

17:07

Let's let let's add the minimum amount of parts to make it not blow up.

17:12

And so, uh, everyone in the space community always quotes this, uh, Latin phrase, add a Astra per Apera, to the stars through hardship.

17:21

I always thought it was just add Astra to the stars, but add Astra Peraspera is to the stars through hardship.

17:26

Through hardship specifically, and I thought that hit really hard.

17:29

Um, this this founders podcast is is fantastic.

17:34

There's a bunch of >> potentially one of the greatest episodes of all time.

17:38

>> It really is like uh there's this great story in here.

17:39

Um, well, it says Elon saved money by questioning the requirements when he asked his team why it would cost $2 million to build a pair of cranes. These are cranes.

17:46

They're supposed to lift rockets.

17:48

He was shown all the safety regulations imposed by the Air Force. Most were obsolete.

17:51

So SpaceX then goes to the Air Force and they start questioning them.

17:55

And it says that SpaceX was able to convince the military to revise them.

17:58

The cranes ended up costing $300,000.

18:03

Like that's like 85% savings. It's crazy.

18:05

He does this over and over and over again.

18:07

But the way he does it is just really really smart.

18:09

He's consistently comparing costs for parts, materials to other industries and other cases. Here's an example.

18:14

Elon consistently pressed his team to source components from non aerospace companies.

18:21

The latches used by NASA cost $1,500 each.

18:25

A SpaceX engineer was able to modify a latch used in a bathroom stall and create a locking mechanism that only cost $30.

18:34

And so again, it's like a 90 98% savings on on the latch.

18:38

And you just do that for every single thing and it adds up to something that when the economics change, it's not just a cheap rocket.

18:49

It's like it's the type of rocket that you can launch every single hour.

18:51

Whereas it's just impossible to underwrite that if you're if you're going to the taxpayer and saying, you know, we need >> and can birth an entirely new industry. >> Exactly. Exactly.

18:59

Uh there was another line in here I'm going to but but uh you guys should go listen to the episode but >> uh David highlights how Elon has this interesting problem where his employees like become so wealthy that they lose like the work ethic that like made the company what it is.

19:15

And so this frustration of like, okay, I'm, you know, we're making all this progress, uh, but then eventually, you know, early employees end up with, you know, a nice house, a vacation house, and they don't want to spend quite as much time totally uh at the factory grinding.

19:30

So, good problem to have.

19:33

>> Well, hopefully you enjoyed the SpaceX stream yesterday.

19:35

I think they used reream.

19:37

If you want to stream, you should use Reream.

19:39

We are powered by Reream.

19:40

One liveream, 30 plus destinations, multiream, and reach your audience wherever they are.

19:44

you can sign up for free.

19:45

Um, there's there's one last funny story in here.

19:48

The day before a launch, a final inspection revealed two small cracks in the engine skirt of the rocket's second stage.

19:54

That's that's the piece at the bottom.

19:56

Everyone at NASA assumed we'd be standing down from the launch for a few weeks.

19:59

The usual plan would be then to replace the entire engine.

20:03

What if we just cut the skirt?

20:06

Elon said, like literally cut around it.

20:09

Why not just trim a tiny bit off the bottom that had the two cracks?

20:11

using a big pair of shears. The skirt was trimmed.

20:15

The rocket launched the next day.

20:17

>> It's wild because another way to des if you had somebody working on your house and they were applying this approach, you'd be like, "What are you doing? This is so janky. >> Are you insane?"

20:26

It's crazy when it's part of this.

20:29

>> Took less than an hour to make the decision.

20:30

Three more principles combined that repeat over and over again.

20:32

Elon's anti-outsourcing, procontrol, and pro- daily iteration.

20:36

Um he uh and uh David has an interesting context here about uh offshoring.

20:42

By sending factories abroad, American companies saved labor costs, but they lost the daily feel for ways to improve their products. Elon bucks this trend.

20:51

He wants to have tight control of the manufacturing process.

20:53

He believe tents up in the in the parking lot. >> Yeah.

20:58

He he believed that designing the factory to build the car, the machine that builds the machine was as important as designing the car itself.

21:04

Uh it's a fantastic episode.

21:06

You should go check it out.

21:08

Um do we want to run through the history of SpaceX, the history of these launches?

21:12

We have Tyler Cosgrove with uh some some beautiful work on the on the TBPN yet to be sponsored whiteboard.

21:22

There could easily be a sponsor on there. Maybe Figma.

21:24

Think bigger, build faster.

21:26

Figma helps design and development teams build great products together. Get started for free.

21:29

uh for now uh Tyler take us through the history of SpaceX, the history of these uh various projects which are >> professor. >> Yeah. Okay.

21:37

So I think >> the intern becomes the professor >> class is in session.

21:41

>> A lot of people I think when when they hear these about these launches they're like aren't even really sure like what rocket is even going on.

21:46

So so I think you can kind of classify SpaceX rockets into like three distinct types.

21:50

At least are like the active ones that are still going on. So there's Falcon 9.

21:54

So this was like the first really like you know commercial like not just tests.

21:59

Um 2010 this is like you know the ninth iteration.

22:03

Um >> we were doing Falcon one before which was one engine right?

22:09

>> I think those were like mostly tests.

22:10

They weren't really reusable.

22:10

It wasn't really a business. >> Yeah. At that point. >> Um yes that was 2010. >> Oh perfect. >> All right. Can you still hear me? >> Yeah great. >> So okay.

22:21

So we have uh that that's Falcon 9.

22:23

Then Falcon Heavy comes out 2018.

22:25

This is essentially just um three Falcon 9 boosters just like strapped together. Yep.

22:30

>> Um but it's kind of a separate thing, right?

22:32

You can These are also um always built to be like reusable, at least some parts of it.

22:36

>> And that's a big piece of like the Falcon 9 has literally just nine engines kind of strapped together.

22:39

It's all the same engine so that they get the reusability there, right?

22:42

So they're bringing like one thing. >> Yeah.

22:45

>> Um and then we go down to Starship.

22:45

I mean this is what's currently being tested.

22:48

This is not like doing commercial missions yet obviously.

22:50

Um, but this is kind of the the next iteration.

22:54

So, this one, um, they're all kind of two parts, right?

22:56

But, uh, this one has 33 engines on the bottom.

23:00

>> Um, that's like called Super Heavy.

23:00

And then Starship is technically just like the the second stage. >> Sure. >> Um, okay. Yeah. Yeah. Yeah. Yeah. >> Yeah. Okay.

23:08

So, then let's go through let's first go through just like kind of general SpaceX launches. >> Yep.

23:13

>> So, here 2006, this is the first Falcon 1 launch.

23:16

This is what we were talking about earlier. >> Okay.

23:17

um kind of none of these like really do a lot until >> they have like three failures it took them and they were almost out of business.

23:24

They >> pictures of Elon in the >> Yeah.

23:27

where he's kind of sitting down. >> Yeah. Yeah. Yeah. >> Yeah.

23:30

>> And quadrilene aole out in the Pacific. >> Yeah. Yep.

23:33

>> And then we go to 2010.

23:33

This is Falcon 9 first launch. >> 2012.

23:37

This is the Grasshopper landing test. Okay.

23:39

So these are when um the first test of like propulsion landing, right?

23:43

which we kind of have seen later with the Starship where it catches it right >> lands down. >> Yep. >> Um 2012.

23:48

This is the first ISS cargo mission.

23:52

>> Y >> so this was um on the story of the grasshopper legs.

23:58

That's uh Doug over at Radiant Nuclear.

24:01

Now he worked on that project.

24:01

Uh it was like no one had done that before. Crazy decision.

24:06

And then Starship is unique in that Elon's basically factoring out the legs because if you take the legs off the rocket that reduces weight so you can get more payload to space and then instead of legs you have um you basically have a pair of chopsticks that catch the rocket.

24:22

Uh and so the legs never leave the Earth, right? That's the strategy. >> Yeah. Um so then >> ISS cargo. >> Yeah. ISS cargo.

24:29

Um this was the on the drone ship.

24:33

>> I didn't realize that took four years to get the drone ship right.

24:35

They have that crazy video we should pull up after this of like all the landing failures on the drone ship with I think it's like in the hall of the mountain king and it's just like crashes after crashes after crashes. Yeah. Okay.

24:45

So, >> and then we go to um this is the first 2017 the first reflight of a booster.

24:51

So, the first time they actually like >> kind of reused it.

24:54

>> Fully reusable at that point. Got it.

24:56

>> And then 2018, this is Falcon Heavy.

24:57

This is the three Falcon 9 put together essentially.

25:00

>> Y >> um then this is where they they basically just start to get the cadence record of launches per year.

25:03

Y you see those like uh those like uh timelapse videos if it goes from like one Falcon 9 launching to like one launching like every second like this is when they >> I mean it's pretty incredible.

25:14

I think the record before this was the USSR in I think it was like 1979 or something.

25:21

>> Yeah, it's pretty insane that they had that record.

25:22

Um but then yeah, we Starship.

25:24

So then let's go through >> Starship launches.

25:27

>> Let's go through Starship launches.

25:27

So first one April 20th, 2023. >> Yep.

25:31

Um, >> so it was fully stacked and it took off, but it blew up after 4 minutes. Is that right?

25:37

Basically, rapid unscheduled.

25:40

>> Not all of the engines started. There's 33. They didn't all start.

25:43

>> Um, but that was the first like kind of >> everything is put together. It it goes up. >> Got it.

25:48

>> Um, then the next test uh 2023 November, this is all all 33 engines started.

25:53

>> That was the big big upgrade.

25:53

Y >> um then we see um it reached orbital velocity. >> Okay. >> So another big step. >> Yep.

25:59

Um June 6, uh we see both stages kind of go up and then they both re-enter. >> Yeah.

26:05

So they both hit their max uh altitude. Yeah.

26:08

So they weren't they weren't trying to go any farther, but coming back is rough cuz you're going through the atmosphere and then they blow up on the way back. Yeah. >> Okay.

26:17

But they're making progress. Yeah. >> Cool.

26:18

>> Uh then we go to October 13th.

26:18

This is like kind of the famous Tower Cat. >> Yeah. I remember that.

26:23

>> This really blew up the >> the chopsticks. >> Yeah. Chopsticks. >> Yeah. >> Um okay. Okay.

26:26

Then we see uh November 19th.

26:29

This is the Raptor relay.

26:29

So the Raptor um >> and the tower catch.

26:32

Did it blow up in the tower?

26:34

No, they caught it, right? They caught it. Fully caught it. That was so crazy.

26:36

But then they probably have to like tear it down and like rebuild it because it's not like fully reusable yet, but it proves that they can actually >> That one was wild because it was so sci-fi, but it had been almost programmed into people's brains that that was like a thing through sci-fi.

26:51

the reaction broadly like people in tech were you know pretty blown away but at the same time it just it felt like something that space company.

27:00

>> The really crazy thing about the tower catch was that it it emphasized how high cadence the actual uh process of reusability can be because I believe that the the the the lower stage the booster goes up and is only in the air for like a couple minutes.

27:15

It's not actually that long.

27:17

Like the the upper stage was in space for like a full hour.

27:23

Like the the mission yesterday was super heavy, but super heavy just goes up and is back in like four minutes or 10 minutes or something like that.

27:28

And so you see it go up and then come right back and get uh and and uh and get caught in the tower and then they can just rotate it over, start refueling it, and it can just bring the next thing up.

27:38

So like the pace of iteration is is the the system is designed to be like not launch every day, it's like launch every 5 minutes.

27:45

It's like truly like airport level. >> Yeah. Um so then >> spaceport.

27:50

>> Let's go to first block two launch. Okay. So block two. >> Block two.

27:54

>> Um that's like when you look at the Starship like basically the top section.

27:58

>> Um there's been a couple it's kind of like this part.

28:00

There's been a couple iterations of this.

28:01

So this was kind of the next big like upgrade. >> Yeah.

28:04

>> Um and then we see March 6, we see a payload attempt.

28:07

So So this is what you're talking about earlier where they try to >> open the doors. >> Open the doors.

28:11

Put out They're not real Starlink um satellites, but they're like, you know, mock ones. Yep.

28:16

>> Um then we see Super Heavy Reflight.

28:16

So So this was um Super Heavy again is this booster.

28:22

That's the first time they like actually reused the same one.

28:25

>> Caught it and then put it back up. Wow. >> Yeah.

28:27

Like it it didn't actually go back up, but they like re ignited it basically and it worked. >> Wow.

28:32

>> Um and then this one just yesterday was uh payload deployment and then we talked about this earlier, but they both stages had >> controlled um splash down in the ocean. >> It's amazing. >> Y >> what a journey.

28:43

Well, congrats to every small note. Uh the the it exploded.

28:48

>> But uh next time, >> next time. Next time. >> All that matters.

28:51

There was a controlled touchdown.

28:54

>> Thank you for the breakdown, Tyler.

28:56

>> Feel free to keep the helmet on for the rest of the show.

28:59

>> And feel free to sign up for Vanta. Automate compliance. Manage risk.

29:01

Prove trust continuously.

29:02

Vanta's trust management platform takes the manual work and security out of your security and compliance process and replace it with continuous automation.

29:09

whether you're pursuing your first framework or managing a complex program.

29:12

Um, Philip Johnston was clearly inspired by what's going on but a little bit frustrated.

29:19

Wanted to highlight this.

29:19

Nothing pisses me off more than VCs proclaiming they don't invest in deep tech companies outside of LA and SF.

29:24

Starlink is quietly building one of the biggest cash generators of our time in Redmond, Washington.

29:30

It's also where Koopier, AWS, and Azure are.

29:32

80% plus of Sat of all satellites launched globally in the last year were designed, built, and tested in Redmond.

29:38

There's a reason we put StarCloud Inc.

29:40

in Redmond and we're on the lookout for more amazing talent.

29:43

So, good luck to Philip Johnston.

29:45

Hopefully, the VCs come around to the good work that folks are doing in Redmond, Washington.

29:50

>> What if we moved Redmond to the Gundo? >> We could. We could.

29:54

Um I think there's probably a um for all these companies there's uh there's assets in in in all the different places.

30:02

the different places. I I haven't heard of a company really uh going deep into South Texas and like because a lot of the Gundo companies are beneficiaries of the work that SpaceX did in Hawthorne and Elsa Gundo and in fact I think that there's a space company that's in the

30:18

maybe it's Rivian is currently not a space company but Riven is currently owns the original SpaceX factory in Elsagundo because they've moved out because they obviously they scaled up um but a lot of the Gundo companies are beneficiaries of the work that SpaceX did in Elsa Gundo to kind of bring back that industrial capacity. Of course,

30:36

Of course, before SpaceX, there was Lockheed and uh Northrup and a few other primes that were in the area and still are.

30:42

Um, but the Gundo's been like a great home for defense tech broadly.

30:47

Uh, maybe Redmond's the next one.

30:49

Maybe maybe you got to do some if you're doing satellite R&D, got to get a Redmond off uh an office.

30:54

if you're doing launch stuff, you want to head out to South Texas because they do seem pretty friendly out there.

31:00

At the end of the the SpaceX stream, the the announcer kind of uh like thanked all the different parties was like, "Thank you to the county where Starbase is because Starbase is like this chartered city within the county and thank you to the people of Texas and the people of this and that."

31:15

>> Well, they got a stock exchange now.

31:17

>> They do yesterday, right? >> Got NY Texas. >> Nice Texas.

31:20

something something down something happening in the uh in the Lone Star State. >> Yeah.

31:25

I haven't been to Texas in years.

31:27

Last time I actually went was for David Center's thing. >> Really? >> Yeah.

31:30

Uh other than that, I don't find a lot of random time to go there.

31:33

It's been a lot of New York, a lot of SF, Little Miami, but >> I have been a bunch.

31:39

>> When was the last time you went? >> Last year.

31:41

I think I went three or four times for different portfolio companies. >> Yeah. >> Great place. Great place.

31:46

Do we have the video of the uh the the SpaceX explosions over time?

31:52

They have this montage reel. I'd love to play that.

31:54

Uh if you can find that as I tell you about graphite.

32:00

dev code review for the age of AI.

32:03

Graphite helps teams on GitHub ship higher quality software faster.

32:07

>> All right, I'm throwing it in the timeline.

32:09

This is the graphite video started >> that I could find.

32:12

We will uh be talking about venture debt and what's going on in that market.

32:16

But um in the meantime, I believe we have a video of SpaceX's >> This is from SpaceX themselves.

32:22

How not to land an orbital rocket booster.

32:27

>> Yes, this is the video I want to watch.

32:29

Can we go full screen on this? This is great.

32:31

Um, and just im immense uh immense confidence to release something like this that on the face of it makes the company look bad, but actually demonstrates what makes the company look great, which is uh not being afraid of failure. Let's play.

32:46

>> The top comment is this is 100% the most expensive. >> Comes down so fast.

32:52

>> It comes down so fast.

32:52

It just the first one is just like didn't stop at all. Here we go. >> Uh-oh. Oh. Uhoh. Uh-oh. Oh no. Wow. There we go. >> Beautiful. But >> this is great.

33:09

>> Unfortunately, >> the smoke's still coming off >> that. Yeah.

33:14

How is that not just going to blow?

33:16

That seems dangerous to me around there.

33:19

>> Yeah, that's a lot of confidence in your product for >> especially for something that like it it actually blows up. Yeah.

33:23

Here it is on the drone ship, I believe.

33:25

Just smashing down constantly. This one trying so hard.

33:29

lands and then it's definitely going to blow. There we go.

33:33

You know it's going to blow up >> every time. >> The music.

33:40

>> Yeah, the music's great.

33:40

I thought they had a different one with all the mountain king.

33:43

Wait, what is coming off of that?

33:47

>> Why didn't they use the chopsticks? >> Yeah.

33:50

Tyler, do you know why they didn't go for chopsticks on this one?

33:51

Instead, did the >> like originally? >> Yeah. Yesterday.

33:54

Can you look up why they didn't go for for chopsticks uh grabbing?

33:58

cuz it seemed like they were able to do that a few flights ago.

34:00

I'd be curious to know why they didn't do it.

34:03

>> This is brilliant marketing.

34:04

>> Yeah, >> it's the Yes. And >> this is so good.

34:07

>> The approach just tilting all over the place. >> Why is it tilting? It's not even going.

34:12

Is it just >> It's a ship on the sea.

34:14

>> On the ship and it's just about to fall. >> This is crazy. Oh, it actually made it.

34:19

It seems like this one's coming down. That was a tough one. >> Solid. >> You call it solid. It's going to explode.

34:26

You know it's going to explode. This is insane. [Music] So good.

34:36

[Music] Oh, could watch these all day. [Music] >> Not bad.

34:45

>> Um anyway, uh on to venture debt journal.

34:51

Um no, you know what's really burning up the timeline?

34:54

what is the the audience is clearly celebrating uh Gulfream Aerospac says join us in officially welcoming the world's first the world's longest range aircraft to the skies the first 8,200 nautical mile G800 has officially been delivered >> longest range business aircraft >> to a customer at our Apple Wisconsin completions facility. Look at that plane.

35:19

So let's give it up for Gulfream Aerospace.

35:22

They don't have the luxury of having the quite the number of explosions that >> they said they said to join us in welcoming the world's longest range aircraft to the skies.

35:31

We are happy to join Gulfream in welcoming the world's longest range aircraft to the skies.

35:38

Anyway, let's go to venture debt.

35:38

Uh there are some changes in the market.

35:40

Um >> yeah, the journal has a story.

35:43

The asset class has shifted to late stage lending as companies remain private for longer.

35:48

Venture debt firms are shifting their focus to larger deals with mature private companies as more startups stay private.

35:53

For longer venture debt value hit a record 53 billion in 2024, up from 27 billion the prior year and more than a previous high of 42 billion in 2021.

36:03

Of course, uh SVB would have been a big uh player.

36:05

So, it's actually still, you know, given given um given how significant SVB was in that market, it's uh pretty incredible to see the growth. >> Yeah.

36:18

>> Um as the value of deals rose, however, venture debt deal count fell to,300 transactions in 2024, the lowest level since 2016.

36:27

The trend continued during the first half of 2025 with a little more than 19 billion in venture debt deal volume according to a separate report by Capital Adviserss Group. Uh great great name.

36:37

Very very very generic, but if you can own it, it's powerful. >> Yes.

36:42

>> Uh debt must react to equity, said Stefan Spazik, director of debt placement at Capital Advisors Group.

36:48

Venture debt providers have have to target stable companies with solid revenue if there is a lack of equity or sponsors decide to pull back on equity commitments to their portfolio companies.

36:57

He added >> I didn't fully understand this because it feels like there is very much not a lack of equity right now but is like what is he >> well I mean he's saying >> I mean he's saying uh there's hope that's the big risk right if a company running in the red and they raise their next round we can get into >> Yeah. Yeah. Yeah. Yeah.

37:17

Yeah, if the if the current backers aren't doubling down at series C, D, E, F, um, which is the vast majority of companies like there, we we focus on the companies that successfully raise every single round.

37:29

Uh, anyway, uh, uh, startups and growth companies are turning to venture debt as an alternative to raising what is often more expensive equity capital that stands to dilute existing shareholders in an environment where traditional exit opportunities remain heavily constrained.

37:40

Venture debt providers say credit also allows venture investors benefits such as retention of company board seats which would have to be given away in an equity financing.

37:49

The quote liquidity constraints on the part of venture equity are still a big driver in the market said David Sprung the founder and chief executive of Runway Growth Capital adding that he expects the venture industries liquidity constraints to continue into next year.

38:03

Runway typically writes checks of roughly 30 million and it lends to tech, healthcare and consumer.

38:09

Earlier this year, for example, data and AI company Data Bricks announced a 52 5.

38:14

2 billion uh you never know with Data Bricks, they might be raising 52 52 >> 52 billion >> announced a 5.

38:22

2 billion credit facility shortly after final close.

38:26

>> We want to be 50% debt to equity on this company on this high growth tech company.

38:33

>> 150 billion company now or something. It's over hundred.

38:34

It's It's >> Yeah, they they just used the the the greater than symbol, >> which is like who's counting it?

38:42

>> The last The last deal was a $10 billion series J. >> Series K. >> They did the K. >> They did the K.

38:50

>> Uh the company said it plans to use the debt for general corporate purposes.

38:51

The debt financing was provided by a consortium of investment banks including JP Morgan Chase, Morgan Stanley, and Goldman Sachs. Let's give it up.

39:00

>> And Blue Owl's in and Apollo Global's in and Blackstone's in. Just a couple names.

39:05

>> They really they really got the murder.

39:07

>> Technology companies are looking to diversify their capital sources and debt is increasingly attractive as a non-dilutive way to fund growth and lower the overall cost of capital. Yeah.

39:15

Uh says >> as companies scale and stay private longer, we see a significant opportunities to support high-quality businesses with flexible credit solutions.

39:22

Uh earlier this month on the smaller side, Runway committed 20 million to Swing Education, an online marketplace that connects schools that substitute teachers, which is backed by firms including Apex Partners and Reach Capital.

39:34

Runway's commitment consisted of a first lean term loan and a revolving credit line.

39:38

Uh what's interesting is like we hear a lot about the venture debt boom in the context of like uh the Hadriens of the world or like you know hard tech industrials like makes a lot of sense to have relationships with debt providers because you're buying a lot of land or you're buying a lot of equipment and like you want to finance value. >> Yeah. Yeah.

39:58

Even if you're buying, you know, when you when you read about Crusoe building a massive data center, it's like, well, of course, there's going to be some credit involved because a piece of that business looks like a traditional mortgage and it looks like a real estate deal almost.

40:11

And so, you want a certain a certain tunch of capital to be focused on earning real estate like returns as opposed to every paying for everything with venture uh capital.

40:21

Um, so since the 2023 collapse of SVB, one of the most active venture debt providers, more lenders have entered the marketers stepped up their existing lending, creating a more competitive landscape.

40:30

Borrowers not only have more appetite for debt, but they now have more choices as well.

40:34

One firm that has capitalized on the opportunities created by SVB's collapse is longtime venture debt lender Hercules Capital. Uh, great name.

40:42

Uh, which typically invests around 40 million in debt deals.

40:46

The firm's deals generally have a 15% loan to value ratio, meaning the borrower's loan volume represents 15% of its total value.

40:52

Hercules is focused on tech and life sciences.

40:57

Earlier this year, Hercules entered into an agreement with publicly traded biotech company Moon Lake Amunotheraputics to provide up to 500 million in non-dilutive capital.

41:06

At the close, there was a $75 million draw down with additional tranches becoming available upon achievement of certain specified uh milestones.

41:14

Uh our biggest competition is the equity market.

41:17

Again, uh this should be obvious to everyone listening, but uh founders are kind of going out and saying, "I need say I need 100 million.

41:26

Do I want to do I want to take the um potential risk that comes uh with an equity financing even if I'm going to be um or sorry potential risk that comes with a uh leveraging debt um even if I'm benefiting from the lack of dilution or do I just want to go um and raise a fresh primary round?

41:45

Um, anyways, uh, >> yeah, >> good to good to see the market, uh, recover.

41:54

>> The the thing that stuck out to me about this was the fact that the deals are getting bigger.

41:59

So, venture debt value hit a record of 53 billion last year.

42:06

The prior year was garbage, 27 billion in 2023.

42:10

Obviously, high interest rates, major pullback in the startup and venture community, uh kind of healing from the uh from the turmoil of like the SVB crisis.

42:23

>> March 10th, March 2023 was when SVB went bankrupt.

42:26

So, it's like it's like, yeah, of course, 2023 was going to be a bad year.

42:29

TW uh 2021, two years prior, was up at 42 billion.

42:32

So, we went from 42 billion down to 27 billion, but are now at a new high water mark of 53 billion.

42:40

But what's changed is the fact that the deal count fell to 1,341 transactions in 2024, which is the lowest since 2016.

42:50

>> More lenders doing less deals. >> Exactly.

42:52

Bigger deals, fewer deals.

42:52

Um, you know, more embracement, more concentration, more focus on the power law winners.

42:59

And that makes sense when you see all the capital that's flowing into the really really big companies, OpenAI, Anthropic, the Crusos of the world that are um able to marshall like a lot more capital.

43:11

There used to be a time in the SVB era where you could go do a $10 million series A and get $2 million in venture debt and you would just add that on top to kind of build the relationship, build the credit, stretch the balance sheet a little bit more, extend the runway, maybe you pay it was always like kind of it all goes into one cash bucket and you kind of spend it however you want.

43:31

spend it however you want. But uh in but it it kind of set you up to stretch your financings a little bit further and a lot of companies just by default got on that on that on that train pretty early and I think that uh SVB leaving the

43:47

market obviously uh really led to consolidation and so we're seeing venture debt go way way up the stack to to you know a5 billion ventured you know credit facility for data bricks if you remember $20 million series like like venture debt round. Like that's pretty

44:04

Like that's pretty big for swing education.

44:07

>> Uh Bench, if you remember, got in a rough spot with venture debt and ended up being acquired by that that company employer. com.

44:15

>> Um I guess they had they were a decent business but had tripped some Yeah.

44:21

>> debt covenants and were forced to abruptly shut down. Yeah.

44:25

>> And then they were ultimately acquired. >> Yeah. Yeah.

44:26

And and that's the thing is like you say like oh well I don't I don't give the venture debt lender a board seat but you know they do have debt covenants.

44:33

They do have the they're at the top of the cap table even above the preferred.

44:38

Uh if there's a liquidation they're going to get their money out first.

44:40

And so um even though they don't have a board seat they do have incredible leverage over the company and they actually have a sort of control that should be definitely considered.

44:51

Um >> it's very one could call it senior >> senior. It is senior.

44:54

Um, the other interesting thing is, uh, Hercules is, uh, is targeting 15% loan to value ratio.

45:01

Uh, that feels higher than what I've seen in the past.

45:04

So, uh, loan to value means if your company's worth $und00 million, they would loan you up to $15 million in venture debt.

45:12

In the previous era, where I was more familiar with this, it was like you would raise $20 million series A at a $und00 million post.

45:22

the VC takes 20% of the company in a board seat and then >> back on low single digits of >> two million in venture debt.

45:28

Three million in venture debt like loan to value of like 3% 4%.

45:31

Um but I think this I think this reflects um them being more selective about going into companies that are not just postproduct market fit and have some like VC hype behind them like they have real >> rep.

45:44

You don't you don't we don't know what the terms are on these 15% loan to value deals.

45:49

It could be that you need to keep a multiple of the of the loan on your balance sheet. Yep.

45:54

your balance sheet. Yep. totally otherwise we're going to >> uh you know require >> and and also they probably have done some sort of like uh private equity breakup uh valuation to say that you know even if this doesn't even if the company completely plateaus and all the

46:10

VCs bail there's still base case you know we can get 15% of the value out here as opposed to the the the previous era which was maybe more risk on in the sense that they were backing earlier companies but in in another way more risk off by because they were writing so so much lower. Um, this is the

46:26

Um, this is the >> Jason Jason O' Conor says, "This show is like a Buddhist statue I have in the background to hopefully bring in good luck and fortune." >> Thank you, Jason.

46:35

Jason, exactly what we're going.

46:37

>> We're praying for you and your capitalist endeavors.

46:38

Always >> I want to pull up this chart on Poly Market of Space Space X Starship fully reusable in 2025.

46:46

Of course, yesterday was a major market mover.

46:50

Um it does not look like anyone anyone had any inside information because uh the market spiked right as the launch was successful uh it jumped from uh 11% chance that this that Starship is fully reusable in 2025 uh to 71% now the market's hovering at 40%.

47:12

Um, and uh, the the rules here are on February 28th, Elon Musk posted that it was likely Starship would become fully reusable in 2025.

47:20

The market will will revol resolve to yes if SpaceX or Elon Musk announces that Starship is fully reusable by December 31st, 2025.

47:32

Otherwise, the market will resolve no.

47:34

So you are you are um you know uh riding on you know SpaceX and Elon's comms which of course will be hotly debated and obviously there's not a pure binary in terms of what reusability means.

47:45

Like clearly this rocket came down and exploded. They need to rebuild it.

47:49

But even when it comes down they still need to swap out parts and fix things.

47:52

Like they they were saying that during this last SpaceX flight they deliberately took off a bunch of tiles just to kind of like see if they could do it.

48:00

And so like you can tell when as the rocket comes down like a lot of those tiles got roasted and so they might have to, you know, swap those out.

48:08

There are elements that aren't reusable, but >> they're definitely going to swap out the stuff that was flying around, >> all the insulation or something.

48:16

That was one of the wildest uh wildest bits that I've seen.

48:21

Uh anyway, >> well, we have this chart here.

48:24

>> Yeah, we have a chart on uh we we we love to trust the experts on this show.

48:29

Uh we are specifically trusting the experts on the White House Intel deal.

48:33

Uh if you haven't been paying attention, living under a data center, uh the US government is now the proud owner of 10% of Intel shares.

48:40

You, the US taxpayer, now have a small slice of Intel, which also means you have a smaller slice of Figure Robotics because apparently Intel Capital invested in Figure Robotics.

48:50

So um >> if you were worried that you didn't have exposure, you can >> Yeah. Yeah. Yeah.

48:56

You can you can you can uh you can send the uh the layered SPV emails that you've been getting to the archive and uh and just be riding with Uncle Sam on that one.

49:05

Um but we uh did a little whirlwind tour of who is saying what about the White House intel deal semi- analysis seems pretty bullish.

49:14

Uh gave the latest uh the last it gives the US the the last US chip maker an actual chance to catch up spur the domestic sector.

49:23

the the the bid here is that um the the Intel story with the government is not going to stop here and that there will be some sort of pressure or incentive applied to other American tech companies to buy from Intel.

49:36

Uh Jordan Snyder, friend of the show over at China Talk is also bullish.

49:41

He says it's Intel's only chance is government support.

49:43

The US needs domestic chip fabrication capabilities and this is is a good path.

49:47

Uh Ben Thompson has a slightly positive bias.

49:50

He he he pulled out the steelman term uh and kind of steelanded.

49:53

He said national security and the economy concerns are too accurate.

49:57

Uh too acute to not keep Intel foundry viable.

50:01

Um although there he does cite uh Scott Linskum who's extremely bearish.

50:06

Uh and this says this will lead to widespread inefficient capital allocation in the chip sector.

50:12

That of course is the fear.

50:12

The Wall Street Journal is also somewhat bearish.

50:16

Uh, pitting Essential US partner Samsung and TSMC against a government-owned competitor is a bad idea.

50:21

I was talking to a few other folks that both they they predicted this, but they didn't think it's necessarily the good. >> Yeah.

50:27

And it has it has ripple effects if if other companies are getting the uh are getting strongly encouraged to use Intel's services. >> Yeah.

50:35

Could make those companies less competitive potentially.

50:37

Um the the uh Asianometry was constructive on the on the deal said could make the deal work by using Intel as an Nvidia second source.

50:47

And of course uh this could potentially hurt Nvidia's margins because the CPUs that Nvidia makes believe at TSMC uh are probably high margin.

50:56

If they have to pay Intel, their margin will fall.

50:57

But this is a small piece of Nvidia's overall uh business and so it shouldn't shouldn't completely upend the the amazing business they've built.

51:06

Of course, Nvidia is reporting earnings today and we are looking forward to tracking.

51:11

>> I have a post here from uh Buho Capital Bloke, the third technology brother says, "The fate of the world is in Jensen's hands.

51:18

Good thing he's a psycho."

51:20

And he has an excerpt from an article here.

51:24

>> Uh that constant kind of relentless improvement looking down at himself.

51:27

He never rests on his laurels.

51:29

One guy told me that after a blowout quarter, Jensen came into the room and said, "This morning, I looked into a mirror and I said, "You suck.

51:37

He does these mental tricks to make sure he doesn't get over because he's so scared."

51:41

He's a student of history that the big thing for technology companies to start thinking you're hot stuff and then start becoming complacent. He is so scared of that.

51:49

So, he tricks himself to not do that. >> It's amazing. >> Um, >> yeah.

51:54

Um, so closing out our trust the experts infographic.

51:58

Bill Bishop is slightly bearish.

51:59

He says Intel fabs haven't proven they can deliver US government ownership in private companies. That's a slippery slope.

52:06

So we will be monitoring the slope to see how slippery it is.

52:11

Hopefully it's the uh the slope of illusionment.

52:14

What is the Gartner hype cycle again?

52:18

>> The trough slope of enlightenment.

52:20

>> The slope of >> the plateau of productivity. >> The trough.

52:23

So, let us know where you are on the Gartner hype cycle.

52:26

We are >> working on some more.

52:29

>> We're gonna crunch all the data.

52:29

Maybe we'll throw it in, Julius.

52:33

>> What analysis do you want to run chat with your data and get expert level insights?

52:35

I want to know how people are tracking on the Gartner hype cycle. >> Yep.

52:41

>> Alex Heath just shared a link.

52:41

Tik Tok's parent company is making more revenue than Meta and is worth less than 20% of its market cap.

52:51

>> So, there's some reporting here from Reuters.

52:52

Uh, Bite Dance, the owner of short video app, uh, Tik Tok is set to launch a new employee share buyback that will value the Chinese technology giant at more than 330 billion.

53:01

Driven by continued revenue growth, the company plans to offer offer current employees $200 per share in the repurchase program. The people said up 5. 5% from $189.

53:14

It each it offered them about six months ago.

53:17

Um the buyback is expected to be launched in autumn.

53:23

>> Uh and the latest buyback at a higher valuation will come as Bite Dance consolidates its position as the world's largest social media company by revenue with its second quarter revenue up 25% yearonear.

53:36

The jump resulted in the company's second quarter revenue hitting about 48 billion, most of which is uh is from the Chinese market as it continues to face political pressure to divest its US arm.

53:48

And uh apparently Tik Tok's US operation remains unprofitable.

53:51

So >> the US operation, >> yes, >> how is that possible? >> Good question.

53:57

They have so many ads and stuff and I thought they were taking a big cut of >> I think Tik Tok Tik Tok shop is wildly unprofitable. >> That's right.

54:05

They were subsidizing a bunch of stuff there. Wow.

54:06

Well, >> maybe it's got to go to Jensen.

54:09

Maybe it's got to go to Oracle.

54:11

Um, we'll keep monitoring.

54:13

>> Let's give it to Larry.

54:13

I'm in favor of giving it to Larry.

54:18

>> Anyway, uh, if you're going beyond getting your brand mentioned on TikTok, you're trying to get your brand mentioned on ChatBT, head over to Profound.

54:25

reach millions of consumers who are using AI to discover new products and brands.

54:29

Uh AJ over at semi analysis says, "Feels like the fact that Zuck poached only a small amount of researchers from Anthropic is really under discussed.

54:37

They have something that really works with respect to culture, mission, and values."

54:41

And >> and remember cursor poached the claude two people on the Claude code team.

54:47

They went over cursor for a couple weeks back at Anthropic.

54:51

I think it was only one week. >> One week.

54:53

That is a crazy short stint.

54:57

Um, uh, Droid says the market implied mandate of heaven.

55:00

I mean, yeah, they like the the culture at anthropic is extremely unique.

55:04

Like they really do, uh, believe that they are on an exponential curve.

55:09

They believe in straight lines on log linear graphs as they say.

55:12

they say. I was I was thinking about that quote the other day uh because if you look at the data center capacity and the and the how pre-training is scaling like it is a straight line on a log linear graph we're getting orders of magnitude improvements every few years uh and it does seem extremely predictable except the last time I was

55:31

looking at straight lines on a log linear graph it was COVID and that of course like completely flatlined at a certain point because there are only so many people that can get a disease and so I uh I'm not I believe the straight lines on log linear graphs, but I also believe that straight lines can become flat lines on log linear graphs and uh we'll have to keep seeing. What do you What do you think?

55:52

>> So then the question is like where are we on the >> on the S-curve?

55:55

>> Yeah, where are we on the S-curve? >> I don't know.

55:57

I think >> we're pretty early. >> Pretty early.

55:59

Uh yeah, I I I do think we're early, but I I think that there's uh I don't know.

56:04

There's just a there's like there's a number of like small percentage chance things that could result in uh in like that scurve hitting like everything from uh like bottlenecks in the supply chain that are somewhat intractable.

56:20

We run out of a certain like bottleneck product that causes a delay in the next oo of b of data center build out.

56:29

We could also see some regulation.

56:31

We could see the business model's not quite like the like the the business community could just panic and it could just become a meme that it's like oh that actually like you know we overinvested we need to pull back for a few years.

56:42

This is what happened during the the the the the dotcom boom like like a lot of those fiber investments made sense over a 10 year 20-y year period but they were overheated in the short term and so that just created like a freeze and all of a sudden we didn't get a continuous buildout or we did but we didn't get continuous capabilities. Um I don't know. >> We'll see. What do you think?

57:05

>> I mean >> I'm bullish. Yeah.

57:08

>> What do you think about the fact that it feels like uh model capability is is like very discontinuous?

57:15

>> Like it feels like >> How do you mean discontinuous?

57:16

>> How do you mean discontinuous? Like it feels like it feels like we get we like like we are on a we are on a smooth path of like bigger models chopping wood training things but like only this is the rune take like like uh as technology rolls out it rolls out in this like smooth curve but then you experience like a few discontinuities when like new capability comes on comes online where like the chatpt moment like there wasn't

57:39

really that much like like if you were just if you were to just look at the the straight line on a log linear graph like it would be very hard to say That's the chat PT moment because it was really like in between GPT3 and in between GPT5

57:51

and in between like you know like medium-sized models, big models and then even bigger models and and like you couldn't you couldn't see there's no kink in the graph at at at the chatbt moment. There is on the user side there

58:02

There is on the user side there is on the revenue side but there's not on the actual like scale of data center side.

58:08

It's not like we went from like nothing to something on the actual data center buildout side and that caused the like so it's hard to predict where there will be new breakout successes along the curve.

58:21

Maybe that doesn't matter. I don't know. >> Yeah.

58:23

I mean I don't think it matters that like I you're probably going to see another kink in user >> Yeah.

58:29

>> usage when school starts again and everyone's like, "Wow, GT5 is so good."

58:32

Because no one's ever used a reasoning model before. >> Yeah. Did you see this?

58:34

Did you see this post about someone said that GPT5 caused like a drop a major drop off?

58:38

Uh this is from uh Marie Martins.

58:41

The model giveth and the model taketh away user registrations coming from AI search.

58:49

ChatgPT was like this huge boom. You see this?

58:52

And then GPT5 launched and uh this is like way deeper in the slides.

58:56

Um >> yeah, I saw I mean there was another like somewhat similar chart.

59:01

Um, >> yeah, >> that was like going around a couple weeks ago where it was like basically you could like tie it to when like schools it was like in May or something then usage completely dropped off.

59:10

Do you remember this chart?

59:13

But then it was actually like oh this is just some like weird like this is like for one organization it was only like 100,000 tokens like it was obviously not the actual usage.

59:22

Yeah, I would be very surprised if if we're if we're at a phase of chat adoption where like the school usage is still like a major factor. Yeah.

59:30

Just because like the the install base is so big it can't just be students, right? Like >> Yeah.

59:35

And then so I mean I'm a little hesitant to like pull a lot of stuff from from that graph. >> Sure.

59:41

>> But it seems it seems very weird. >> Yeah. >> I don't know.

59:45

>> Anyway, um >> this is a super >> interesting chart from from Mary Martins.

59:50

She is the founder of Tally Forms. Yeah.

59:55

>> And yeah, basically calling out they were getting a ton of registrations from people just in chat GPT and then all of a sudden dropped off a a cliff.

1:00:04

>> Uh so yeah, we'll see.

1:00:04

There's some debate in the comments around how GPD5 started stopped adding uh UTM parameters to requests. >> Oh, interesting.

1:00:14

>> There could be an attribution Yeah.

1:00:14

uh issue there, but they're saying um >> yeah, I mean, we we definitely want her to zoom out and look at overall registrations and see if there's a a fall-off in those because this was this purely additive or was it just like an attribution thing?

1:00:31

And then um and then I'd love to know like is this actually like affecting her business or is or is Tally still growing at the same rate?

1:00:38

Um and and it's just like it feels like it's coming from a different from a different source, but in fact it's all the same.

1:00:46

>> It's not impacting PopMart CEO and founder Wang Ning.

1:00:49

He's now the 79th richest person in the world.

1:00:52

According to Morning Brew, his net worth has grown by 20 billion this year.

1:00:55

It makes him richer than names like Peter Teal, uh David Ter, and Steve Cohen. >> Steve Cohen. Wow. >> Leoo billionaire. >> Labu billionaire.

1:01:06

Um, and we helped, we contributed to this by buying a labu for Bishop >> for Bill Bishop's book. >> So, it's on the way. Bill >> Tashi. >> Yeah.

1:01:17

Very excited for that one.

1:01:18

>> Anyway, back to Enthropic.

1:01:18

Uh, Daario was talking to John Collison on Sneaky Cheeky Pike uh about how he approaches LLM economics in an interview with with John Collison.

1:01:28

I thought this was interesting.

1:01:29

So, Dario says, "Uh, get kind of burned.

1:01:31

There's two ways you could describe what's happening right now in the business model."

1:01:34

So let's say in 2023 you train a model that costs $und00 million and then you deploy it in 2024 and it makes $200 million in revenue.

1:01:42

Meanwhile, because of scaling laws in 2024, you also train a model that costs $1 billion.

1:01:48

And then in 2025, you get two billion of revenue from that $1 billion and you've spent 10 billion to train the model that year.

1:01:55

Uh so if you look at it in a conventional way, the profit and loss of the company, you've lost $100 million the first year, you've lost $800 million the second year, and you've lost 8 billion in the third year.

1:02:06

So it looks like you're getting worse and worse exponentially.

1:02:09

>> Uh if you consider each model to be a company, the model that was trained in 2023 was profitable.

1:02:12

Um uh you paid $100 million and then it made 200 million in revenue.

1:02:19

Uh there's some cost to inference the model, but let's just assume in this current tunish example that even if you add those two up, you're kind of in a good state.

1:02:24

So if every model was a company, the model in this example is actually profitable.

1:02:27

And I thought about this in the context of GPT4.

1:02:31

That was obviously I think that was like around a hundred million dollar training run, something like that.

1:02:35

And they've clearly made a ton of money back from that because GPT4 is the backbone of something that generates like billions of dollars in revenue.

1:02:41

Of course, they've also trained other models that have been used less and maybe were less additive.

1:02:46

Um but uh th this was framed as like, oh no, this is like a nightmare.

1:02:52

like he's exposing the truth that like these will never make money because they're just going to continually invest endlessly.

1:02:56

Um I didn't see it that way at all.

1:02:59

This makes perfect sense to me.

1:03:01

>> At some point you could just stop training the qu the question is can you just stop training new models and just profits of old models? >> Yeah.

1:03:09

Well, it it's more just like at at a certain point you run out of you run out of you kind of tap the capital markets because maybe you can't raise $10 trillion for to fund the next model on debt or equity or whatever.

1:03:22

Um, but there's no reason that you can't like in each of these models if you wait 10 years you can afford the next run, right?

1:03:31

Because you train the model for 100 million.

1:03:34

Let's say you were like I'm bootstrapping a foundation model company. Crazy. But you could do it. Here's how you do it.

1:03:38

You you you invest, you know, $1 in a model that makes you a dollar next year, $2 next year.

1:03:47

Then you take and you wait 10 years. You save up $10.

1:03:49

Then you do the $10 run and you expand until you do the $und00 million training run.

1:03:54

That makes you $200 million the next year and $200 million the next year.

1:03:58

And you know what you have after five years?

1:03:59

You have a billion dollars in profit.

1:04:01

What do you do with that billion dollars?

1:04:02

You train the next model.

1:04:04

Then you wait five years and then you train and then you've saved 10 billion.

1:04:07

then you train the $10 billion market.

1:04:10

>> This assumes that this assumes of course that you would have a competitive offering in the market and you wouldn't be getting >> Well, yes. Yes.

1:04:16

But but that is a broader market dynamic of whether the market is willing to put up like currently the market's 100% cool putting up $100 million into a foundation model company.

1:04:27

There were like 20 companies that raised $100 million to train foundation models.

1:04:30

There's only a few of them left.

1:04:31

Then those few, a bunch of them put up, you know, a billion to train the billion dollar model.

1:04:36

Now you have who's who's going to put up 10 billion to do the the the really big training run?

1:04:41

It's going to be a few labs.

1:04:43

It's going to be OpenAI, Anthropic, Meta, Google, right?

1:04:44

Um as we get to the hundred billion, yeah, we might have to wait a little bit.

1:04:50

We might have to kind of like acrue savings and pay that down over time.

1:04:53

And this is another thing that might cause a kink in like the log linear graph if you have to if you can't immediately marshall 10 trillion dollars, hundred trillion dollars.

1:05:02

Um yeah, like right now we are in the phase of like snap your fingers in the capital marshals because we're we're in this uh we're in this era where the the gains >> remember round was so overs subscribed they ended up I think doubling the allocation. >> Yeah.

1:05:15

And so this could continue.

1:05:17

>> The concern I mean the Yeah.

1:05:17

The concern is that you can't can't re it seems difficult for even the big labs to do another 10x on on uh these pending training runs. >> Yeah.

1:05:29

I mean that's what Stargate is, right?

1:05:31

It's a $500 billion project.

1:05:31

Um, and so that's probably like a hundred billion dollar commitment in the medium term.

1:05:39

We're getting into the trillion territory.

1:05:41

Um, maybe it continues forever.

1:05:43

May maybe everyone on Earth kind of just says like yes, we're willing to like everyone sell all your gold and convert that.

1:05:49

But I think if you're I think if you're on if you're on Wall Street traditional finance guy or girl and you're looking at this and and in in biotech you might have something where it's like hey we need to spend $5 billion developing this drug but then it's going to be good and we're going to be able to sell it for a really long time.

1:06:08

>> This is every having massively uncertain future capital requirements y >> is going to just be a concern and an asterisk on the business.

1:06:15

I think that's like Finnhub is is really the concern here.

1:06:20

You have exponentially increasing costs and just wildly uncertain future capital requirements. >> Yeah.

1:06:27

And so the if like the way to deal with that like uncertainty, those jitters in like the later stage investors who are who you have to talk to to marshall once you run out of MOSA who are willing to take wild swings.

1:06:37

Um the way to allay their concerns is show them that yeah, we spent 10 billion and we're making 2 billion in profit.

1:06:45

In 5 years we'll be able to pay that.

1:06:47

So you can, you know, give us the money now to pull forward what's obviously going to pay off again.

1:06:55

And we're confident that the next 10x is going to get us another 10x in in profit.

1:06:59

And so uh you can underwrite these at any level.

1:07:02

Uh right now we're underwriting them every single year.

1:07:06

Uh we might stretch out a little bit.

1:07:08

We might compress, we might go faster. We We'll see.

1:07:11

>> Well, but >> Dan Ratliff in the chat says, "Elizabeth Holmes has hit the timeline.

1:07:15

She has been posting ripping posts."

1:07:17

Uh she quoted uh quote quoted Brian quote tweeted Brian Johnson earlier.

1:07:23

Brian said, "Defeating death would be humanity's greatest achievement." Elizabeth says, "Amen."

1:07:29

Uh she gives another quote here.

1:07:32

There comes a point where we just need to stop pulling people out of the river.

1:07:35

We need to go upstream and find out why they're falling in by Archbishop Desmond Tutu.

1:07:44

Um, and yeah, she's ripping some comments.

1:07:46

She's getting active on the timeline.

1:07:48

>> Now, this is mostly my words. >> Not her, right?

1:07:50

This is her >> mostly my words posted by others.

1:07:51

So, I wonder if paper printouts of >> probably over the phone.

1:07:57

She just said um she was responding to Brian and said, "First they think you're crazy, then they fight you, then you change the world."

1:08:07

And uh somebody quoted and said, "She's going to launch the Theronos Memecoin, isn't she?"

1:08:12

And Elizabeth says, "Never."

1:08:14

So um anyways, having having fun um on the timeline. >> Yeah.

1:08:23

Imagine the thrill of having posts read to you over the prison phone and then replying, "Okay, yes, like that one."

1:08:31

>> She said, "Last time."

1:08:34

>> Uh, Signal said, "Uh, dreams do come true if you put your mind to it."

1:08:38

>> Uh, and Elizabeth Holmes followed him.

1:08:41

>> And VC Brag said, "Never deleting this app."

1:08:44

And Elizabeth Holmes said, "Last time I deleted the app, it looked different.

1:08:47

Where did the little blue bird go? What happened?" >> This is funny lore.

1:08:51

This is like an odd It would be weird to be offline for so long.

1:08:56

I mean, >> she's she's responding to an Elon Musk quote saying, "For my part, I never I will never give up." And I mean never.

1:09:03

>> And Elizabeth says, "This is one reason Elizabeth, this is one reason Elon Musk wins. He never gives up. That's what it takes.

1:09:10

It doesn't sound like >> it doesn't sound like uh Elizabeth Holmes is giving up either, but she seems to be a big Elon Musk fan."

1:09:19

>> I have been shocked by the out of home campaign. Have you seen this? >> The blood hurts.

1:09:25

>> I someone is working on a redemption arc for Elizabeth Holmes and justice for Theronos.

1:09:30

I think it's a promotion for the new company of some sort, but it's all over.

1:09:35

Yeah, there's a new company.

1:09:37

>> Uh I think they're outraising.

1:09:37

Um uh but uh the the uh the the billboards are all over LA and they've been up for a long time, like months and months and months. It's been crazy.

1:09:48

Um anyway, Ben Gilbert uh says, "If you think your app needs more polish, always remember Google Maps shift maps without Europe." >> That's awesome.

1:09:58

>> That is a crazy screenshot.

1:10:00

>> It's crazy that they didn't crazy. >> Yeah.

1:10:02

They didn't even think to put like the rough outline of Europe, you know, and be like, "Ah, we don't have details on it."

1:10:07

But instead, it's just like you can zoom out and you only see America, the States, and and England.

1:10:11

And we got Ireland in there.

1:10:13

Of course, got to keep Ireland in there.

1:10:15

Well, if you're building the next Google Maps, get on linear. Linear. app.

1:10:18

Linear is a purpose-built tool for planning and building products.

1:10:22

Meet the system for modern software development.

1:10:24

Streamline issues projects and product road mapaps building.

1:10:28

>> Travis Kelce right after the announcement launches American Eagle collection. >> Yes.

1:10:34

>> And uh Michael Mirllor says nothing is sacred. Monetize everything. >> Let's see.

1:10:39

Let's check in on the American Eagle stock.

1:10:41

It is up another 9% today.

1:10:45

So they're just on an absolute tear. Yes.

1:10:50

>> Um, >> everyone in tech has been focused on the the Taylor Swift wedding.

1:10:53

Um, particularly what it means for Nvidia earnings.

1:10:57

I have a little thesis here.

1:11:00

So, um, obviously Taylor Swift getting engaged.

1:11:02

That's going to draw a ton of attention towards the NFL.

1:11:05

We've already seen the NFL had a measurable Swift effect from 2023 to 2024.

1:11:10

They saw 9% gain in female viewership year-over-year and game level spikes like a 63% jump among women 18 to 49 for a Chiefs game.

1:11:23

So imagine this NFL and streamers they lean in broadcasters if they escalate moment capture more shoulder programming alternate feeds multi- language clips realtime highlight reels. What do you need?

1:11:34

You need real time media. It's an AI problem.

1:11:36

You need clipping, scene detection, player ID, small object tracking.

1:11:40

All of this is going to run on Nvidia GPUs.

1:11:42

Uh globalization, you need translation at the edge.

1:11:46

Classic speech AI billion for Jensen text to speech translation.

1:11:52

That's an Nvidia product right there.

1:11:54

They have a product Reva that does that.

1:11:56

Uh creators pile in UGC explodes, reaction videos, podcast shorts, consumer AI effects, noise removal, eye contact correction, studio voice, increasingly GPUensive workloads.

1:12:04

People are going to be running RTX PCs with Nvidia broadcast software to get to get this stuff.

1:12:12

Ad dollars also follows attention.

1:12:14

Incremental audience especially for for the new female demos.

1:12:18

This is going to pull premium uh premium dollars to sports ads budgets.

1:12:20

Uh this pushes ad tech recommener systems ranking creative optimization frequency control.

1:12:27

These are among the most comput hungry inference workloads.

1:12:30

and an Nvidia sweet spot.

1:12:32

They have Merlin uh for for for recommendations uh of ads and Triton for serving ads.

1:12:38

Uh stream quality is going to be an arms race.

1:12:40

The women the the Swifties, they're going to def they're going to demand 4K feeds.

1:12:47

Platforms are going to race to improve video.

1:12:49

Um they're going to need uh to accelerate this with CUDA.

1:12:53

This means more GPU accelerated pre and post-processing incloud workflows.

1:12:57

Uh, sports production is going to modernize.

1:13:02

Autonomous cameras, analytics are a blueprint for low tier games.

1:13:04

These techniques climb the stack for 10 pole events.

1:13:09

More AI in the truck, more GPUs at the edge in and in cloud transcode farms.

1:13:15

Hypers scale hypers scale capex is the real lever here.

1:13:18

AWS, Google, Microsoft, Meta, they fund the compute behind all of this.

1:13:23

Imagine 2026 capex guides are going through the roof.

1:13:26

Over 10, we're going to be in hundreds of billions of dollars just supporting the Swifty demand for NFL content powered by Exactly.

1:13:35

And so all this is going to translate into earnings, of course.

1:13:39

Showcases durable, broad-based inference demand in media and adds exactly the vector analysts want to see as the AI cycle tilts from training only to training plus always on inference.

1:13:48

So that's what the Taylor Swift engagement means for Nvidia's earnings today. >> Groundbreaking.

1:13:55

Gabriel says, "This is the music that Keith Reo listened to in 2019 while doing a thousand burpees and right before logging on to reply wrong to a political scientist posting about uh Vuda Judge's South Carolina polling results. Can we play this video?"

1:14:12

>> This video is hilarious.

1:14:12

Uh in the meantime, while they're pulling it up, I'll tell you about numeral sales tax on autopilot.

1:14:16

Spend less than five minutes five minutes per month on sales tax compliance. numeralhq. com.

1:14:23

And you know who's going to need to use numeral?

1:14:24

Travis Kelce's new American Eagle line.

1:14:26

If you buy some American Eagle, uh they said inside Travis Kelce's new American Eagle line, an ad with athletes including Sunni Lee and Keon Anthony.

1:14:36

The day after it was revealed that Travis Kelce put an 8 karat engagement ring on Taylor Swift's finger.

1:14:42

American Eagle had announced it had a limited edition clothing line with its sportsware brand True Colors.

1:14:46

AE uh X True Colors by Travis Kelce will launch in two drops today and on September 24th.

1:14:54

So get ready and make sure that you're paying your sales tax on true colors. com.

1:14:58

Michael Mafur says nothing is sacred. Monetize everything.

1:15:05

>> Let's pull up the the DJ that >> the video.

1:15:07

The video is going to be um almost impossible to find.

1:15:11

>> Okay, it's deep in the Oh, they got it. >> Got it. Let's pull it up.

1:15:13

Let's play the >> Get the audio. >> Get the audio. The audio is key.

1:15:21

Let's go full screen on this.

1:15:24

This is like so annoyingly I don't know. I was not a fan of this. Did you listen to this? You think it's good?

1:15:31

[Music] >> Very Reddit coded.

1:15:34

>> Yeah, >> but this >> I can't believe there's so many people at this putting up with this Mario song. What What event is this? It does. It does inspire me. You know what we need?

1:15:50

We need what's coming up next.

1:15:53

>> We need uh we need this uh You see the the It's like they blow this fog around. Look at this. >> Me, Mario.

1:16:03

>> Yeah, the those like uh CO2 canyon cannons for sure. Like that to blow this.

1:16:09

>> Okay, we got something good out of this. Ben, we need the CO2.

1:16:10

Whatever those Yeah, whatever those DJ equipment are. Oh, yeah.

1:16:14

During the lightning round when we were interviewing >> and Nick, can you look into the health impact of being in a studio space with CO2 cannons?

1:16:22

>> Regardless, this is very David Solomon coded. Yes.

1:16:27

Uh, you have to be a Mario fan to understand. Yeah, I I I get it.

1:16:29

Um, I don't know if this would be for me and a big crowd, but uh who knows? Maybe maybe it's fun.

1:16:36

It seems like they had fun and I'm happy for them. these days.

1:16:38

Not not big into paying to be in a crowd.

1:16:42

>> Yeah, that's >> pay to avoid one. >> Yeah. Yeah. Yeah. That's that's rough.

1:16:47

Uh anyway, let's go to Wilmanitis.

1:16:47

He says, "Hard to think of a piece of writing that's had a greater impact on my life than John Ludig's index mindset in 2021.

1:16:55

This pro-index tendency, writes John Ludig, pervades the private tech markets, startups, and even our culture through what I call the index mindset.

1:17:02

A focus on preservation over creation, optionality over decisiveness, general over specific.

1:17:10

Public companies are an obvious thing to index, but the index mindset manifests in many domains.

1:17:14

In wealth, you instead of risk risk seeking expansion, you focus on steady compounding preservation.

1:17:24

In venture capital, instead of concentrated bets, you focus on deployment pace and IRR.

1:17:28

In public markets, instead of active funds like hedge funds, which we'll talk about later in the show, we focus on passive funds and diversification.

1:17:35

In employment, instead of four plus four to fiveyear tenurs and being focused, you focus on one to twoyear tenures portfolio building, lots of logos on your resume.

1:17:46

Um, in re uh in re research and development, instead of being radically radical and innovative, you focus on being incremental and defensive.

1:17:53

And in your personal life, instead of spiky and committed, you focus on being well-rounded and committal.

1:17:59

Internet software companies are far less risky than they used to be.

1:18:03

Even at the early stages, there hasn't been a venture vintage since 2002 with negative median returns.

1:18:07

Big tech is now huge tech.

1:18:10

So, the VCs are not taking nearly enough risk because they're they're not losing money.

1:18:14

They should there should like it is it is an indictment of the venture community if there's not a single fund out there that has a negative or a single vintage that has negative returns because they weren't taking enough risk. >> Well, it's median. >> Yeah. Yeah. Median.

1:18:25

Uh but uh yeah, it's vintage.

1:18:28

So it's broadly of course there are funds that don't do bad but it's like as an asset class.

1:18:34

>> Well, we might be turning it around. >> Anything's possible.

1:18:39

>> Uh riskaverse people money have flooded in.

1:18:41

When people have something to lose, they protect their downside.

1:18:44

Think of wealth managers encouraging a safe mix of stocks and bonds.

1:18:47

The tech industry has too much to lose.

1:18:49

Uh, and I love Will Manitis has the most chaotic way to highlight I've ever seen.

1:18:55

You can just use boxes uh in iOS if that's where he's editing this.

1:19:01

But he's just draws you in, drawing all over it.

1:19:04

>> Drawing all over it, and then he switches colors at some point.

1:19:05

I have no idea why, but uh Ben Gilbert says he's so good. And I agree. Shout out to John Ludig.

1:19:11

Uh, Satrini says, "Time to bust out the nose sequel Negrini Negronis again."

1:19:17

>> So, I think this was a picture from 2021, >> I think.

1:19:20

So, there was a Tiger Global >> This is when I think it was roughly 30% of all dollars in the venture industry were going to founder dinners and little and little uh events like this.

1:19:35

>> Essentially, essentially, >> it felt like that at the time.

1:19:36

They had the NoSQL negroni. >> Yep.

1:19:41

>> And >> the tiger tonic, the machine learning margarita, the AI artificial inebriation special, >> which is a non-alcoholic beverage. >> Interesting. Interesting.

1:19:51

>> I say bring it back, Tiger.

1:19:52

>> Yeah, I think you could go further further.

1:19:54

We we we got to come up with our own little menu of funny things.

1:19:56

We didn't have a menu at our at our event in New York.

1:20:00

>> This was an interesting uh screenshot. This one from Finn.

1:20:03

ai, the number one AI agent for customer service, number one in performance benchmarks, number one in competitive bakeoffs, number one ranking on G2. Finn. AI.

1:20:12

Uh, >> I will never apologize for cutting you off >> for an advert for an advertiser. Continue. Jord >> event.

1:20:19

Uh, fewer people can bench 225 than are worth 10 million.

1:20:22

In the US, about4% of the population. Around 1.

1:20:24

3 million people can bench uh 225 pounds.

1:20:31

Globally, the percentage is likely below. 1%. >> Wow.

1:20:35

>> With some estimates as low as 007%. 2.

1:20:38

13 US households have a net worth of 10 million or more constituting approximately 1. 6% of all households.

1:20:44

If you can bench two plates, you're already stronger than 99% of people on the planet. Just maybe not richer.

1:20:53

And Zach V says what everyone is thinking.

1:20:56

only job is to do both and uh we firmly >> both are a grind but well worth it.

1:21:03

I it's uh the grind to 225 was was serious for for many on the team and I think almost all of us have gotten there and are now beyond.

1:21:13

It's uh it's well well worth it. Well worth it.

1:21:15

Um, >> Capital says, "Your stock trades at $59, but we can make it trade at 62 if we manufacture a crisis by changing the logo, then immediately change it back to the original logo."

1:21:30

Uh, this strategy uh I don't think this was their strategy, but uh feels like something out of Na Nathan for you episode.

1:21:37

Uh, and I would >> it's hard to it's hard to properly manufacture a reaction like that.

1:21:43

Of course, this is about Cracker Barrel. >> Oh, did you?

1:21:47

So, the Cracker Barrel story is in the journal today.

1:21:49

Logo change spawns Firestorm.

1:21:52

Um, it's Oh, is this something about it?

1:21:57

>> Cracker Barrel planned to celebrate a fall menu and logo makeover with a festive country music concert in New York City.

1:22:04

I had no idea this that's what they were planning, but its new logo stole the show and not in the way the company intended.

1:22:10

On Monday, the company apologized for how it communicated the changes, but didn't pivot from plans to keep updating the brand.

1:22:16

The chain replaced its longtime logo featuring a man in overalls leaning against a bar a barrel with a streamlined version featuring just the chain's name.

1:22:24

The move engulfed the restaurant in a culture war firestorm with some commentators online and some customers accusing Cracker Barrel of assuing its country charm and heritage for a sanitized image.

1:22:35

Critics have lobbyed have lobbed personal attacks on social media against Julie Fel Masino, chief executive of the nearly 56-year-old chain.

1:22:44

Uh she scrapped a beloved American aesthetic and replaced it with sterile, soulless branding, wrote the woke war room on X in a message shared by Donald Trump Jr.

1:22:55

Uh she should resign and be replaced with leadership that will restore Cracker Barrels tradition.

1:23:01

The fallout has shaved tens of millions of dollars from the company's market cap.

1:23:05

Uh spawned calls for boycots and risked the casual dining chains turnaround plan.

1:23:10

Cracker Barrel has defended its changes, saying all the all the more campaign was meant to honor its legacy while bringing new energy to the brand.

1:23:19

Our values haven't changed and the heart and soul of Cracker Barrel haven't changed.

1:23:23

The company said the new logo, the fifth iteration in its history, is a call back to the original one with its barrel shape and word design.

1:23:31

On Monday, Cracker Barrel said the dust up over the past several days had shown how deeply people care about the chain.

1:23:37

Trump said on Truth Social yesterday, "Cracker Barrel should go back to the old logo, admit a mistake based on customer response, parentheses, the ultimate poll, and manage the company better than ever before.

1:23:48

Make Cracker Barrel a winner again."

1:23:51

And they immediately responded, reverting it back to the old logo. Wow.

1:23:56

>> And if you had invested invested in Cracker Barrel and Nvidia exactly a year ago, you would have outperformed buying Cracker Barrel. >> Wait, sorry. What was that? >> 52% in the last year. In the last year. No way.

1:24:12

>> Um, and Nvidia is up 41 and a half%.

1:24:16

>> Well, if you're looking for exposure to Cracker Barrel, either either long or short, go over to public. com.

1:24:20

investing for those that take it seriously.

1:24:21

They got multi-asset investing, industryleading yields, and they're trusted by millions, folks.

1:24:25

Um, this weekend, says Jeff Morris Jr.

1:24:27

, Martin Casado dropped a tweet that took over VC Twitter.

1:24:31

I wrote about it and what being consensus versus non-conensus actually means in today's market.

1:24:36

Every weekend there's a new hero, says JMJ, uh, that takes over venture capital group chats.

1:24:43

This past weekend was clearly slow on the news front because what I initially read as a fairly standard take ended up dominating my timeline.

1:24:48

Uh this is we talked about this Martin Casado said the idea that non-conensus investing is where the alpha is is actually quite dangerous in the early stage.

1:24:56

Follow on capital tends to be more and more consensus aligned.

1:24:59

For whatever reason this concept triggered a bunch of investors.

1:25:02

We're living in an era where consensus categories like AI, crypto, robotics, biode, autonomous systems, manufacturing are big enough to carry entire generations of startups.

1:25:09

You don't need to invent a new category to generate massive outcomes.

1:25:12

And that is an interesting that is an interesting take because like we're there VCs are supposed to be like tech investors, internet investors and yet like consensus categories include crypto, bio, defense, manufacturing, AI.

1:25:26

It's like everything is now is now like backable.

1:25:31

backable. Um there was an interesting uh quote in here that I wanted to highlight something about uh you can still yes you can still find non-consensus companies within consensus categories but at some point calling ourselves non-conensus investors is just glamorizing the job take American dynamism as an example a

1:25:49

venture category branded just a few years ago by A16Z that now feels like it's industry in its in itself uh in Palunteer in 2003 was truly contrarian nobody else was building forward deployed CIA software Palanteer is one of one company that probably only Alex Karp, Joe Lansdale, and Peter Teal could have created. Two decades later, it's

1:26:06

Two decades later, it's more relevant than ever with a $370 billion market cap.

1:26:09

And was equally contrarian, but for different reasons.

1:26:14

2025, he says, "Fast forward, American Dynamism is a full-blown investment category.

1:26:17

Hundreds of startups and dozens of funds are chasing defense, aerospace, and manufacturing.

1:26:21

I live 15 minutes from Elsa Gundo, which has famously become a dense hub of these companies."

1:26:26

Um there was another one in here about uh about um D to C and saying something about like that category being even even even that category.

1:26:39

>> One one note here I I was talking with uh was talking with a a buddy over the weekend and he was like what's going on with Elsagundo are are or is it is it legit?

1:26:49

are these are all these companies uh are they actually going to build big companies and something um uh something Jeff says here.

1:26:57

Oh my my reaction was some a small number of them will be massive and most will have meddling outcomes or or uh will shut down.

1:27:07

Uh but that is that is what our entire industry is based on and it's okay but it's still good that we have this excitement and energy and momentum and and a center of gravity and Elsagundo.

1:27:22

>> Yeah, there was something in here I I I can't exactly find it but it was basically saying that like like consumer is is contrarian right now. It's not very hot.

1:27:31

Very few people invest in consumer DTOC companies.

1:27:33

But that's just in like Silicon Valley venture.

1:27:36

There are still plenty of funds both on the private equity side and on the venture side that are doing deals in consumer all day long.

1:27:44

Some brands have kind of expanded and shifted away from it, but there still are lots of funds out there that are doing that.

1:27:49

So, um it's a it's just a it's just an interesting uh noodling on um what what what stands for consensus or contrarian right now.

1:27:57

Uh anyway, let me tell you about Adio customer relationship magic.

1:28:01

Adio is the AI native CRM that builds, scales, and grows your company to the next level.

1:28:06

you can get started for >> with a new 52.

1:28:08

>> Do you want to do this John Woo post >> series B?

1:28:10

Yeah, this John Woo post is great.

1:28:12

We we should run through this.

1:28:13

So, John Woo says, "Do not trust millennials to run your brand.

1:28:15

Their vision of the way the internet works was informed by the web 2.

1:28:18

0, Facebook 2011, Instagram vintage filters era, and their priors have not been updated since.

1:28:25

The internet is no longer deterministic.

1:28:29

It's algorithm driven and not even among your own followers.

1:28:31

Your posts are constantly being shown to a panel of users.

1:28:34

If they like it, the panel is expanded.

1:28:36

If they don't, you don't get reach.

1:28:38

That means putting out a high volume of post isn't noise, spam, or quote low signal.

1:28:43

Volume is necessary so the algorithm can pick up your winners and make them viral.

1:28:48

You have to experiment with high volume.

1:28:49

Learn quickly and double down on your winning formats.

1:28:53

The cost of failure is zero.

1:28:55

The tree falls in the forest and no one sees it.

1:28:59

>> Content creation just no one sees no one sees the flopped posts. >> Yep.

1:29:05

The cost of experimenting uh but the cost of not experimenting is astronomical.

1:29:09

Sure, signs of incompetence in the attention age.

1:29:11

We don't want to put out noise.

1:29:13

We need to protect our brand. We need more followers. That's ignorance.

1:29:16

Lowquality posts are no longer distributed.

1:29:21

Follower count doesn't matter.

1:29:21

The news cycle is shorter than ever.

1:29:23

People will only remember your winners. Pump up the volume. Totally agree.

1:29:27

I think if you look, you know, on X specifically, if you look at your favorite poster, if you actually scroll their feed, they'll have a bunch of like flops. >> Yeah.

1:29:35

You'll think of a person as like, oh, every time I see one of their posts, it has thousands of likes and then you go to their account and you see so many flops and and and the real secret is that they're trying a lot of experiments.

1:29:45

When the current thing happens, they're putting up every possible variation of the current thing.

1:29:49

Uh, it's a good strategy.

1:29:49

I like this uh reply by Holen.

1:29:51

She says, "Millennials are still good brand builders, but the execution should be left to the terminally terminally online."

1:29:57

And so there is something about like it's probably not enough just to have like pure slop.

1:30:03

It's more like you need a seed of something good, some sort of north star and then uh and then the but the execution in the modern era does require a lot of aggression.

1:30:11

Uh anyway, um we have our first guest of the show joining in just a few minutes in the reream waiting room.

1:30:18

Uh we will get to uh the Numeraai story in just a little bit.

1:30:25

Should we talk about uh the app mafia quickly?

1:30:28

Uh if you haven't been following this, uh some folks that we're familiar with, one of them who did uh PMF or die, um uh have put out a course on how to build an app that makes money, don't raise venture capital.

1:30:42

They sort of took over the timeline with a viral video showing them driving Lamborghinis.

1:30:49

uh kind of all the all the classic hallmarks of the of the guru that you previously found on YouTube uh with the here in my garage that guy um and folks on Instagram.

1:30:59

Sort of the first time we really saw it on Twitter or on X um and it caused a lot of uh you know people people going back and forth.

1:31:07

Harry uh Gistent I don't know how to pronounce his last name, but Harry says um apparently I mentioned in the first few minutes of the AT Mafia course so here are my thoughts.

1:31:16

I went to college with Blake Anderson and know Zack and Hunter.

1:31:20

Two, there are a lot of threads going around about profit margin, etc.

1:31:21

It's not it's not that relevant given these kids are clearly netting a minimum six to seven figures, probably for the first time in history for anyone that age.

1:31:31

This represents a major shift in the education system that platforms like are have pioneered.

1:31:35

People are talking about audience IQ.

1:31:37

This course is clearly targeted at kids/ younger founders and certainly something I would have loved 10 years ago.

1:31:42

It also represents a shift away from ventureback consumer apps dying from over raising.

1:31:49

Hundreds of apps like Headspace Calm raised eight to nine figures and were stuck with companies that couldn't quite IPO but also couldn't cash flow for the founders.

1:31:57

There's an entitled socialist mentality of quote why do I have to pay?

1:32:02

The reality is if it's free it's not very valuable.

1:32:03

I have watched the I haven't watched the course but these guys are extremely sharp and can pro provide value to young founders.

1:32:08

They so they deserve to be compensated for their time.

1:32:12

Now there's a wrinkle where the course is now free. Is that right? >> Course is now free. >> Okay.

1:32:16

>> So um I you know I think they they the the original video was uh was in their words designed to elicit a a dramatic reaction designed to bring out uh hate. >> Right.

1:32:30

They're sitting there in a mansion.

1:32:33

>> Blake loves poking the bear of the algorithm.

1:32:35

Like he he's very good at that and very deliberate about that.

1:32:38

>> But yeah, I it it is fascinating.

1:32:38

I mean, and and then immediately people piled on and were saying like, "Oh, the apps must not be making a lot of money otherwise. Why would they do this?"

1:32:47

Which >> uh I didn't um it seemed obvious to me why they wanted to do it.

1:32:51

They want to build their personal brands and this is like a business model for building a personal brand. Yeah.

1:32:57

personal brand. Yeah. this it's a it's a >> what's interesting is I don't think I've ever paid for a course on specifically on like entrepreneurship or something broad like that but I have paid for courses on like very specific technologies like I went to a like a boot camp for iOS development where they

1:33:14

taught us objective C for two days straight and that had like a real cost and I was able to expense it against the business I also paid for like a course to learn Houdini a particular uh visual effects software and that kind of had I I I thought it wasn't like a get-richquick scam at all. They didn't

1:33:29

They didn't make any promises.

1:33:30

They just said, "Hey, we're going to teach you this software.

1:33:33

We're going to teach you this thing."

1:33:34

And so, buying books is great.

1:33:34

Buying courses on specific topics is great.

1:33:36

Um, a lots lots of what you can learn about entrepreneurship is just available out there for free.

1:33:42

You can read Paul Graham's blog.

1:33:44

You can listen to Founders Podcast.

1:33:46

There are plenty of resources.

1:33:48

Um, maybe this is useful.

1:33:50

You'd have to try it out.

1:33:50

Now, it's free so you can make your own decision.

1:33:54

Anyway, we have our first guest of the stream in the TV coming in. Welcome to the stream.

1:34:00

>> Announcement, >> Josh, good to meet you. How you doing? >> What's happening? >> Hello. >> Good to see you both. Hi, John. Hi, Jordy. Jay joining the group.

1:34:06

>> You sound fantastic today.

1:34:06

Whatever microphone you have hooked up is working flawlessly, >> similar to what we have here.

1:34:10

Uh, congrats on the news.

1:34:13

Break break break it down for us.

1:34:16

>> Yeah, so I'm Josh, head of Gusto.

1:34:16

Uh, we love serving small business.

1:34:19

Uh we've worked with a company called Guideline for a number of years, close to 10 years at this point.

1:34:25

>> And the news is that we are joining forces.

1:34:28

Guideline's going to be joining Gusto. >> Amazing. >> Amazing.

1:34:31

I think I've I've actually used Gusto and Guideline together for my entire career.

1:34:38

>> I've used Gusto at four different companies.

1:34:40

>> Company I started in college >> and we have an employee sitting over there that wants a 401k.

1:34:42

So this is You have no excuses anymore. Time to set up. >> Yeah.

1:34:49

Um talk about the the the prehistory here.

1:34:52

Uh the company's been uh what was it YC 2013 2012 or something? I remember 2012 2012. Yeah.

1:35:01

Uh I remember implementing it at my YC company in 2012 uh back when it had a different name.

1:35:05

Um and uh and I remember that that for a number of different uh HR functionalities uh I would have to kind of go to a different platform and then the data would be uh would be fed back.

1:35:20

>> What informed the strategy back then and then why are you shifting it?

1:35:22

It like it seemed like it was working for a long time.

1:35:25

So why why kind of what is it horizontally integrate I suppose is the term.

1:35:30

>> I don't know what all the business is.

1:35:32

We just we just try to make the life of a small business owner easier.

1:35:34

But yeah, first off, I mean, thrilled to serve you and your companies and and we're always trying to get better.

1:35:40

So, send me feedback on how we can get better, by the way.

1:35:42

Um, but yeah, no, I'd say we go way back.

1:35:45

We launched Gusto back in 2012 and our first product was payroll.

1:35:50

>> Um, if you don't pay someone, they quit.

1:35:52

So, it's the least optional part of the stack.

1:35:54

>> And so, we've made progress there.

1:35:54

You know, we serve over 400,000 businesses.

1:35:58

Um, we love helping tech companies, but I love >> a lot of businesses, too.

1:36:01

That's a lot of business.

1:36:03

>> But I also love reminding folks there's more dentist offices in the US than tech startups.

1:36:07

So, like very focused on that mainstream small business owner.

1:36:12

>> And uh, yeah, to your point, you know, we we listen to customers more than anything and and they guide us.

1:36:15

And so, they told us, hey, there's all these other pain points we want your help with.

1:36:20

>> Our second product was health benefits, but you know, we're not going to build everything ourselves.

1:36:24

And so when retirement came up and like generally speaking, what we like to do is take stuff big companies have and bring it to small companies, >> right?

1:36:32

So from almost a fairness lens, it's like we need to bring great retirement benefits to these small businesses. They deserve it.

1:36:37

And that was basically KB the one of the co-founders of Guideline.

1:36:41

It's kind of a funny story.

1:36:43

He was previously the co-founder of Task Rabbit.

1:36:45

Um and so back in like 2015, we were in the same building.

1:36:49

Uh they were on the second floor, we were on the sixth floor and we there's a slow slow elevator, like the slowest elevator you've ever seen.

1:36:55

And so we would sometimes just be on these long elevator rides and that's how we got to know each other. >> Yeah.

1:37:02

I was going to I was going to ask when did the conversation first.

1:37:04

>> People say you meet your you meet your acquirer like years before, but what they don't say is in a very slow elevator.

1:37:09

>> You're going to hang out for for minutes at a time. >> Yeah.

1:37:12

But yeah, to answer the heart of your question, um we started partnering with them right when they launched in 2016.

1:37:17

And yeah, there's an integration.

1:37:20

You can kind of connect the two systems.

1:37:22

Like we have tens of thousands of shared customers.

1:37:24

Um, you know, KB I can say this stat very proudly on his behalf.

1:37:29

Like over 10 around 10 billion dollars of retirement savings in the accounts of our shared customers.

1:37:35

>> Let's hit the side for that.

1:37:37

>> I mean, yeah, Jordy, you you hit the gong.

1:37:38

We we like to hit the gong around here for >> Congratulations. >> I like it.

1:37:43

>> I like hitting the I like hitting the gong for retirement funds. >> Yeah. Yeah.

1:37:46

I mean, that's that's that's again that's their people's money.

1:37:50

That's their retirement peace of mind.

1:37:51

But >> uh but yeah, what we what we realized is there's a couple things happening.

1:37:54

One, >> we can do even more together.

1:37:56

So, we're going to continue to partner with third parties on other products, but especially with 401k and retirement, like we're going to really go deep and and that's what guideline has done.

1:38:05

Um they're a broker dealer, so there's all these different parts of the stack where if you cut out better cost savings.

1:38:10

I'm sure you had investors along the way that said, "You're sending so many customers to Guideline.

1:38:17

Why don't why don't you just build this?

1:38:19

You know, if you need if you need more capital, we'll provide it."

1:38:23

What why why did that not make sense?

1:38:25

And and why why why uh why buy verse build in this situation?

1:38:32

>> So, I think all options are always on the table when you're building a company. Buy, partner, build.

1:38:36

Uh and like for what we're doing, it's going to be a lot of connected problems.

1:38:40

So, you know, how do we bring it together as one simple, easy to use product is the key.

1:38:46

Um, but I'd say, yeah, the catalyst for shifting from partner to being one company is like there's just product stuff we can do that we can't do unless we're combined.

1:38:54

I can give you a couple examples if that's helpful. >> Yeah, please. >> Yeah. >> Yeah.

1:38:58

So, like the journey of of building a company or the employee journey.

1:39:02

We think a lot about kind of the person by person, what happens in their life, but let's say someone gets uh promoted, right?

1:39:07

And they get a pay increase.

1:39:08

Um, so that's obviously coming through our system because it's going to affect their salary.

1:39:12

And in that moment, right, whether it's in the app or in an email, uh, the, hey, like you now have more money coming into your pocket, you know, your choice, but do you want to set aside a bit more in your 401k or someone has a kid, you know, in our system that's a dependent, but that's a human being.

1:39:29

Someone just had a kid, right?

1:39:31

Like maybe you want to go do something there in the context of retirement planning or something related to health benefits.

1:39:36

So, kind of thinking through the person by person journey piece of it, um there's just things we can do that we think will drive more folks offering retirement benefits.

1:39:45

There's also a compliance reason too which I can get into if >> let's get into it. We love this stuff.

1:39:52

>> So, so like government, right?

1:39:52

We're here to like abstract government and sometimes there's carrot stick that government does.

1:39:56

So, 41k is a good example of of a carrot like it's pre-tax dollars.

1:40:02

So literally the government is making it hopefully a no-brainer to like put aside money for retirement with meaningful amplification of of that money than just you know putting it uh post tax.

1:40:13

But they also have compliance requirements.

1:40:15

I think over now 20 close to 25 states have rolled out what's basically a mandate that businesses have to provide some type of retirement benefit for their teams.

1:40:26

>> And so that's good intent, but it's still stressful because there can be penalties or fines.

1:40:29

And so we're also excited that basically we can give peace of mind to more small businesses so they're not stressed out, definitely not getting fined or penalized.

1:40:37

Um, and if they want to or need to set up a retirement benefit, we can make it brand simple through gusto plus guideline.

1:40:44

>> What's the anatomy of a partnership look like in this category pre-acquisition merger?

1:40:51

merger? Um, we were talking a few days ago about the relationship between Apple and Open AI and we were kind of like wait like should Apple be paying for the right to have good AI on their phones or should chatt be paying because they did

1:41:07

all these extra queries and we were kind of and I think like the rumor is they net it out to be like there's no money transferred and I could imagine that sometimes that comes up sometimes it's like hey I'm getting you customers you got to give me a fee like what what are the structures? Do you have companies

1:41:20

Do you have companies that are in one camp and the other or do you can you share anything about the relationship beforehand?

1:41:24

I'd be interested to know just like channel sales is something I'm like learning about now.

1:41:31

>> So the first thing to start with is Gusto's entire engine of of growth.

1:41:33

You know, here's another metric. We can tee it up.

1:41:37

Uh but we're going to we're on track to add about 150,000 customers this year. >> That's a lot.

1:41:43

>> So is that businesses? That's small business. Wow.

1:41:46

That's a lot of businesses. >> Yeah. Yeah. Okay.

1:41:49

And so >> you're talking some guys that absolutely love business. So this is fantastic.

1:41:55

>> The best news I've heard all day.

1:41:57

>> We can get into our whole like what is AI gonna do in the world tangent if you want too. Interesting.

1:42:02

>> Um >> but yeah, we'll come back to that. >> Yeah.

1:42:05

So basically like we are an engine to go bring in more and more >> small businesses, especially new employers, folks starting out, right?

1:42:11

If you don't pay someone, they quit.

1:42:12

So, it's right off the bat like an important key thing for you to do if you're building a team or adding even your first employee.

1:42:20

>> Y >> and so, um, there's a lot of folks as you can imagine that would love to, uh, distribute through Gusto because of the the scale that we're at.

1:42:27

And so, what that looks like first is not us starting with the partner, but starting with the the small businesses and finding out like what are their pain points, who are the products, what are the products they like or want to use already.

1:42:38

They're voting with their actions that come in through our integrations.

1:42:41

And then we do a lot of vetting.

1:42:43

So I think more Apple approach than the Android.

1:42:45

We want to make sure quality experience, accuracy, reliability, and then like just this obsession with user experience uh comes through of whoever we're working with.

1:42:56

>> And then yeah, we'll do partnership deals across multiple product areas.

1:42:59

Time tracking is another example.

1:43:01

Sometimes we'll have first party like we'll do our own product and still have best of breed third party. >> Sure.

1:43:06

>> And then that's a negotiation.

1:43:06

There's usually if we're going to drive a lot of customers to someone, there's some type of rev split. >> Yep.

1:43:11

Um because we're creating value for them, there should obviously be value for us.

1:43:14

>> Yeah, that makes sense.

1:43:16

>> What is your what is your guidance for a ton of startups sign up for Gusto today when they're one or two people.

1:43:21

It's just the founders and and um what is uh what is your kind of framework for you know the right companies to use Gusto and and um because you know again companies today they might start with a couple founders today and be 30 30 people in six months, right?

1:43:39

people in six months, right? these things happen >> like we got to vibe code our own >> but how do how do you >> got to rip this out and vibe code our own >> are you counting the are you counting the AI agents as people are not is my question >> yeah yeah that's a good question um but yeah how are you how are you thinking of serving startups as of today it's

1:43:56

obviously from my understanding a lot of where you guys started but it's uh it's always evolving >> yeah I mean as a quick backdrop you're alluding to it three founders we all started here in Silicon Valley we went through YC you know I was born and raised here my my parents aren't from tech though they came from like uh my dad was a teacher, my mom was a teacher. Um and so

1:44:13

Um and so >> same here was a teacher. >> Oh yeah. Where did you teach?

1:44:18

>> Uh in high school that I went to >> high school. My my dad's high school. >> Very cool. But yeah, it was funny.

1:44:23

I think uh I don't know if I asked my dad this.

1:44:27

I don't think he would have had a great answer, but I do remember asking somebody back in the day. How do I pay?

1:44:32

I was had a company was starting to make money.

1:44:35

I was like, how do I pay myself?

1:44:36

And they're like, "Oh, you just set up like gusto and you just onboard yourself."

1:44:40

And I was like, "Oh, that's cool."

1:44:43

>> Well, we love serving small business.

1:44:45

Obviously, tech startups are a part of what we do.

1:44:47

We love serving them, too.

1:44:47

I guess from a like target customer lens, you know, broadly speaking, like one to 100 person companies is our sweet spot.

1:44:56

That's 98% of the employers in America.

1:44:59

>> But I always love like also it kind of blows people mind usually there there's six million employers in America, give or take.

1:45:04

Twothirds are less than five employees. >> Wow. >> Right.

1:45:08

Twothirds are less than five employees.

1:45:09

So, whenever I meet a small business owner and there are three people and they say, "I'm small."

1:45:12

I'm like, "No, no, you're not small.

1:45:13

You're like the common business owner in America. You are >> the average." Yeah.

1:45:19

>> Well, >> uh, quick thoughts on, uh, AI because you alluded to it earlier. >> Yeah.

1:45:23

I'm interested I'm interested specifically in how you think about surfacing AI.

1:45:27

surfacing AI. like do you want to put a chat box in some of in front of someone and then that's an interface to all the the the deterministic workflows you've built or do you want to just turn AI loose within your you know behind the

1:45:41

scenes development team and have flows that are working just to improve things reality check things like you're probably using machine learning for fraud detection for years is it an evolution of that >> there's a product manager inside of you um I would say yes to both and more Yeah. >> Um, so like we do a lot of complex kind

1:45:58

>> Um, so like we do a lot of complex kind of backend processing of either money movement, filings, documents.

1:46:03

There's a lot of ways to use AI tooling for productivity there.

1:46:07

>> Um, so that's great and we've always used technology for how we scale, right?

1:46:11

Like we serve over 400,000 customers today.

1:46:12

That's only possible through through using a lot of software.

1:46:14

through using a lot of software. Uh but from a customer experience lens um actually again both I really think there's this massive I mean understatement here but paradigm shift around how we use software and how software helps us you know when I got into software in high school like

1:46:31

>> what I loved about it is it felt like it gave you superpowers and then every time I saw like books on how to use a tool or I learned like someone would brag on LinkedIn I can use this tool it like really really annoyed me because software is not something you should have to learn how to use. It should just It should just work.

1:46:45

It should just make sense. It should be intuitive.

1:46:46

It should give you superpowers.

1:46:47

So conversational interface I think is going to be a pretty I mean again understatement but like >> much more accessible way for a lot of folks to use software and it'll work in combination with web workflows and stuff like that.

1:46:59

And so we have a conversational interface for Gusto called Gus. >> Sure.

1:47:03

>> And to your point like it can take actions, right? >> You get it. You get it.

1:47:06

Um but like you can like set up a sift shift schedule. You can go run payroll.

1:47:10

like you can leverage all of our APIs that we spent 14 years building and Gus can do stuff for you, not just give you information.

1:47:18

>> Funny story, my dog's name is Gus Gustavo and I have a friend who could never, he was so into uh B2B software.

1:47:26

He works in the B2B software industry that he didn't understand that my dog's name was Gustavo and for a long time he would call him Gusto.

1:47:32

I was like, I didn't name my dog after the software platform. Well, I have one.

1:47:38

>> Speaking of dogs, Bill Bishop is in the is in the uh Substack stream from Cynicism Cinism, and he says he's a happy gusto client.

1:47:46

So, you have another fan listening right now.

1:47:50

>> Thrilled to serve you.

1:47:50

>> I have one I have one question. Uh and I'm curious.

1:47:53

When when I got onto Gusto, >> Yes.

1:47:56

>> back in the day, was it was close to a decade ago at this point, >> I was I was confused that I had to hit the button to run payroll.

1:48:05

I was like, "Is this not a payroll platform? Why why do I have to?"

1:48:09

And I remember there was a couple times that I would get an email and it' be like I'd be like, >> you're like, "You need to run payroll now." And I was like, "Why?"

1:48:16

I I think you do you can automate it now. But is there >> Yeah.

1:48:21

Is there is there kind of a generational divide there in some ways where there's like a generation that understands like payroll is something that you hit a button and you run it versus payroll is just something that is just humming.

1:48:32

At the first startup I ever interned at, uh, the CFO hated me because I would always turn in my time sheets late and he's like, I need to run payroll.

1:48:40

And I'm like, just pay me the same every month. I'm fine.

1:48:42

Like, you can just put it on autopilot.

1:48:44

And they're like, autopilot doesn't exist.

1:48:47

This is 2006 or something like that.

1:48:50

>> So, we we rolled out autopilot for payroll back in like 2013. >> I remember using it. It was fantastic. >> But it was Yeah.

1:48:54

So, the key the difference here between both of your use cases was if you have nothing to input.

1:49:00

So if it's like let's say all salary versus some hourly um andor like no one took PTO.

1:49:04

But again there's there's inputs that go into payroll.

1:49:06

Now assuming that's all there like it should just work and that's the way it does work.

1:49:11

We way back when this is now 2013ish like we have uh Penny the Pig is kind of our like kind of a part of our company kind of brand culture mascot.

1:49:21

>> I can see Penny and Gus potentially beefing over who's the most important.

1:49:26

>> I mean they're they're >> mascots.

1:49:28

>> They're friendly with each other.

1:49:29

There's there's a friendly competition, but but we had to have the the gift.

1:49:31

It was a gift for like running payroll when you clicked run and it went too fast and we had customers back then go like it can't be that fast. Something is broken.

1:49:41

So we had the gift just run longer for the sake of it >> just to make people be like, hey, there's actually hard processing happening behind the scenes here.

1:49:48

>> And and the world's come a long way since then.

1:49:49

So people expect stuff to be fast, simple. Um we want to automate.

1:49:51

I mean a good example is 41k.

1:49:53

So if you want to go change the contribution that's being made like that can now be done easier by us being one company.

1:50:00

>> But yeah I we think eventually payroll becomes something like flowing water.

1:50:02

It just works >> and and you don't have to do much um direct action yourself. >> Fantastic.

1:50:08

Well thank you so much for hopping on >> chatting.

1:50:10

Uh acquisition makes a ton of sense and congrats to both teams.

1:50:14

>> We'll talk to you soon. >> Yeah.

1:50:15

And make sure you give that teammate a 41k. >> We will set it up. >> It's happening. >> Have a good one. Talk to you soon.

1:50:20

All right, >> and we have Keller from Zipline in the reream waiting room. >> Big news. Zipotle is now live.

1:50:26

Chipotle partners with Zipline for aerial delivery.

1:50:33

What more could you stream? How you doing?

1:50:38

>> Even more dramatic background this time.

1:50:40

>> Every time you deliver.

1:50:40

Thank you so much for hopping on.

1:50:43

>> That's our job, right?

1:50:45

>> Yeah, that was unintentional.

1:50:46

Uh anyway, uh how did this come together?

1:50:49

It seems it seems like the perfect partner.

1:50:51

Walk us through the the evolution that actually what went into the announcement.

1:50:58

>> Yeah, I mean we we we've been talking to Chipotle for a while.

1:51:00

We obviously have announced a few other restaurant partners.

1:51:02

Um I mean you know for reference like there was a time in my life where I lived on Chipotle.

1:51:06

Like that was my >> every entrepreneur every entrepreneur has like a Chipotle era.

1:51:12

>> Raise your hand in the studio if you've lived on Chipotle.

1:51:14

I think everyone's >> it's yeah you're getting more.

1:51:16

It's just a period of your life where you're getting more calories from Chipotle than anywhere else. >> Oh yes, absolutely.

1:51:23

>> I can distinctly remember uh those moments.

1:51:26

>> It's the base of the pyramid.

1:51:28

>> I also remember I also remember I feel like you guys would appreciate this.

1:51:30

You know, there was a time when I was living out of my car action like eating Chipotle and you know $10 you could get like two full meals out of it.

1:51:37

>> Then you could you could get like the tortillas too like those tortillas.

1:51:41

There's a lot of optimization that you can do.

1:51:44

And I I remember thinking that like one day I will know like it like how will I know if I'm rich?

1:51:52

I will be able to get the guacamole every time without feeling guilty about >> I have the exact same story.

1:51:58

When I went through YC in 2012, uh my business partner, my co-founder would always get double meat and I was like, "We're trying to save money.

1:52:07

We don't we can't afford double meat." And double beans. Double beans are free.

1:52:12

>> It was tearing us apart.

1:52:12

And then I finally just defaulted to like, look, if you if you're going to get it, I'm going to get it, too.

1:52:17

And we will both fall on the sword of higher burn.

1:52:19

And and to this day, if I >> moral hazard. >> Yeah. Moral hazard.

1:52:24

But to this day, if I eat out with him uh and we're at a restaurant, I'll just be like, get whatever he gets cuz I know he's going to get something ridiculous and he's getting the most expensive thing on the menu.

1:52:32

And at least then I get stuck with the half of the bill and I'm happy.

1:52:34

And I'm not like, ah, I got stuck paying for his expensive thing. Anyway, congratulations.

1:52:39

We love We love Chipotle.

1:52:41

Um what what else what what else can you share about the progress of the business?

1:52:45

It felt like you know classic overnight success as we love to talk about on the show but um >> also how did you get them to actually name it?

1:52:54

>> Their idea that was their idea not oursing.

1:52:57

oursing. I think you know look I I think that what we're you know obviously there's this like big transformation coming in AI and robotics and there are a lot of you know very fancy hyperscalers that get talked about all the time but the reality is that like all the basic businesses and brands that we use and love also want to be able to

1:53:14

take advantage of this kind of technology and so I think a lot of you know some some of the biggest brands like Chipotle are asking themselves like hey how can we use robotics automation to like accelerate or you know dramatically transform our own brand and the kinds of experiences we can provide to customers. So yeah, them choosing

1:53:30

So yeah, them choosing Zipotle, us trying to deliver for them.

1:53:34

I mean, I think that we did a delivery in 5 minutes yesterday from like order made to the thing arriving at a customer's house.

1:53:41

Um, that's not possible every time, just to be clear.

1:53:44

It does take time to make the food, but like, >> you know, I think that something showing up super super fast like that.

1:53:48

In fact, we've we've found it has been necessary because we we've only started really rapidly scaling these food deliveries in the last couple months.

1:53:57

It's actually necessary to tell customers to slow down and make sure to blow on the food.

1:54:00

We've had people accidentally burn themselves.

1:54:04

>> Um >> good problem to have. >> Yeah.

1:54:06

Weird weird thing, right?

1:54:06

People are just used to it being cold or smush. Yeah. Just not not that great.

1:54:10

So, um you know, really cool to see that starting to scale, especially for a brand that's like very near and dear to my entrepreneurial journey and heart.

1:54:18

Um I mean yeah you know other things that we're seeing >> even since we were last 10 weeks ago or something yeah a couple couple months ago um we have seen you know a problem that we were having in July was that the service was growing extremely fast between 25 and 30% week overw week in terms of flight volumes um in in in fact >> yeah like so you know Saturday new record high flight volume Sunday we actually beat Saturday by 20%.

1:54:46

This is just two days ago.

1:54:49

Yesterday, you know, Monday was a record day.

1:54:50

Tuesday was a record day. Really cool.

1:54:52

So, >> if you keep if you keep giving us records, I'm going to keep hitting the So, >> cooler with the records.

1:55:05

>> No, but even like even cooler is that um you know, right now the service has a net promoter score of 94.

1:55:09

So, you know, no very few services in the United States are close to that today.

1:55:14

um let alone delivery services.

1:55:16

Uh customers are ordering a lot of customers are ordering like three to four times per week.

1:55:22

In fact, we have many customers who are ordering a couple times a day via the service.

1:55:25

I think that you know when you realize you can get stuff in 5 minutes, it actually starts to change the way that you kind of live your lives.

1:55:31

It's giving people back like three or four hours a week when they can just be focusing on their kids or on their family rather than like stressing out with traffic and trying to get somewhere.

1:55:40

um the at this rate in a couple months um our deliveries in the US are going to exceed all of our global flight volume across eight other countries.

1:55:49

So yeah, US is growing super super fast.

1:55:52

Now we're launching a new Walmart Super Center every week at this pace.

1:55:57

>> Um it's just yeah it's I think uh you know I think the moment for automation and autonomous delivery has arrived.

1:56:03

>> So is is Walmart >> 10-year overnight success?

1:56:05

>> 10-year overnight success? Should I think about the expansion as driven by go where the Walmarts are and then talk to the city officials to get support and Walmart's kind of like the backbone or is it more like find other cities like Dallas me whether they're in Texas or not and then partner with companies and

1:56:25

businesses that are in those kind of friendly cities that understand the value that this can bring and the demand and then uh and then kind of roll out that way like is there a particular path that you're taking like how should I imagine the map of Zipline growing or the coverage map growing over the next few years? >> Yeah, that's a really preient question.

1:56:43

>> Yeah, that's a really preient question.

1:56:45

I mean, Walmart has been a key partner.

1:56:46

We're building a lot of charging infrastructure across these metros with them.

1:56:50

Um, we're also building independent charging infrastructure that's just designed to serve a lot of different customers at once.

1:56:55

So, for example, we just um opened a big charging site uh in uh uh Rowlet.

1:56:58

It's a small city inside da small town inside Dallas where like that one site is going to be able to serve I think upwards of 40 or 50 restaurants that are all within range and then from there we can deliver out to customer homes as long as they're within 10 miles.

1:57:15

So like one key thing to think about here is it's kind of a network optimization problem but we go a lot farther than normal delivery.

1:57:21

you know, where you have a human driving a car wants to go any from any given restaurant, you can typically reach 10 times as many people via instant delivery. Wow.

1:57:30

Um, as as via, you know, traditional delivery.

1:57:33

So, and again, um, >> you know, 10 times as fast.

1:57:35

So, we're really focused on going metro by metro right now.

1:57:39

A lot of our costs, you know, maintenance, it kind of happens on a metro basis.

1:57:42

So, we're we're going to go tall.

1:57:44

Like the goal is to go very tall in each metro.

1:57:46

We have a lot of metros now that are like super super excited and basically on the road map for us to launch.

1:57:52

Um um and and we're we're trying to build the capacity and launch them as quickly as possible.

1:57:57

But you know building this kind of infrastructure across the US it's like it's a big undertaking.

1:58:02

Um, and you know the other the other cool thing to mention >> just really quick is that like >> you know one thing I'm very very proud of more proud of every day.

1:58:10

I mean Zipline now has 120 million commercial autonomous miles is the largest commercial autonomous system on Earth. Zero safety incidents. That's the main point. Zero safety incidents.

1:58:20

So actually if you just look at the basic safety of cars globally the system is already provably about 10 times as safe as cars um in terms of fatalities or injuries.

1:58:30

Um and you know from that safety perspective I mean like you know zipline is actually right now about um we have we have already achieved a level of safety that is 2x our goal that we had set by the end of this year.

1:58:41

So like again safety and reliability is kind of everything.

1:58:44

It's like core of what we do and it's a big part of like why I think cities are starting to understand like this is a um incontrovertibly good thing.

1:58:52

>> Is this a problem for Pizza Hut?

1:58:52

Pizza Hut famously the buildings are huts.

1:58:54

You can't land on the roof.

1:58:56

How are we going to get pizzas from Pizza Hut in uh >> The important questions.

1:59:01

>> The important questions.

1:59:01

But but seriously, like is there >> I thought you were talking about the size of the pizza. >> Oh, no.

1:59:04

Because actually I want to know you are working on a pizza compatible.

1:59:10

>> We just we just cut it in half. >> Oh, great.

1:59:11

The company >> we already have we already have the box designed for a couple different pizza partners. It works great.

1:59:19

>> Should should uh should franchises be thinking about a flat roof? Is that relevant?

1:59:25

uh what does the actual infrastructure look like uh as and like should should restaurants and businesses be thinking about zipline compatibility as they roll out the next like version of their of their retail infrastructure or their their residential plan or the real estate plan. >> Yeah. So, it's pretty cool.

1:59:42

I mean, first of all, one of the things we announced just in the last couple months is something we call zipping points.

1:59:46

And zipping points are um really simple infrastructure. We pay for it.

1:59:50

We can drop it in 2 hours.

1:59:52

It does not have to be permitted. There's no construction.

1:59:56

As long as we drop a zipping point with a restaurant or a retailer or a hospital or primary care facility, they're instantly enabled with zipline delivery. >> Got it.

2:00:03

>> No power, no permitting, nothing.

2:00:03

Y >> do not have to do anything for the building.

2:00:06

Interestingly, we are talking to like a lot of hotels, a lot of apartment buildings and a lot of even these restaurants as they cuz for example, a lot of these big restaurant brands, they'll only have like two or three or four total designs of buildings that they'll build.

2:00:19

And um they are in fact now starting to take into account Zipline and they're starting to design specifically for autonomous delivery.

2:00:27

You don't have to do that, but you can.

2:00:29

You know, one big point I would make here is that like the thing that blows my mind, I mean, starting to see this like exponential takeoff from a demand perspective.

2:00:35

I would emphasize, by the way, we freaked out enough about demand four weeks ago that we turned off all marketing, all demand generation marketing in the company.

2:00:42

We turned off all the inapp notifications.

2:00:44

We turned off all the field marketing that we were doing.

2:00:48

and it basically made no difference.

2:00:51

Like we continue to grow 25 or 30%, you know, week over week.

2:00:53

I think that um what the reason this is happening in retrospect is kind of obvious which is, you know, we're now starting to deliver to um a lot of offices.

2:01:02

We're delivering to a lot of apartment buildings.

2:01:04

We're delivering to universities.

2:01:06

Like I think the product more viral. >> Yeah.

2:01:09

When you think about the virality of the product, I mean I'm sure this is already happening or will happen which is a my story about this. >> Yeah.

2:01:16

>> Yeah. my brain goes to some guys at a fraternity order a 30 rack gets dropped into the backyard and I don't know if you're doing that you're probably not doing that yet but uh that is a very viral moment that people will be taking videos of for a long time they'll >> you can you can see I mean

2:01:34

>> tons of people are doing Tik Toks and a lot of those Tik Toks are getting seen like seven eight million times we were actually initially asking customers to not um like you know put stuff on social just because like we wanted a chance to do an early access program and kind of like, you know, iron out the kinks. But

2:01:47

But yeah, even even even people who aren't customers are sometimes like videoing their neighbor getting a delivery and then putting it on TikTok and having it get seen millions of times. So, it is cool. >> Incredible. >> Amazing stuff.

2:02:00

Always always great to chat with you.

2:02:02

Shout out to Danielle, your or whoever's running your your comm's program because your remote your your remote call-in setup, I think, is the best of any.

2:02:11

>> This is this is the I mean, this is the manufacturing line.

2:02:13

So this this line is ultimately capable of producing about 55,000 aircraft a year and and that's scaling to that as fast as we possibly can. >> That's so many.

2:02:23

>> We have to we have to learn how to build drones in the US. >> Yeah. Yeah.

2:02:25

I'm very excited about doing it. >> Thank you.

2:02:28

>> Uh we will talk to you soon.

2:02:28

Have a great >> welcome here. >> Anytime. >> Talk soon. Kelly. >> All right. Cheers. >> Talk soon. Thanks.

2:02:35

>> Up next we have eightsleep. com. Get a pot five Ultra.

2:02:38

I put on a clinic last night. Jordy P.

2:02:40

My favorite soundboard Q. I got a 90. >> Oh, good for you.

2:02:45

I got an 84 >> roasted 100% quality, 88% consistency, 6 hours and 31 minutes. Slept. I'm on a run.

2:02:54

Got a 5-year warranty, 30 night risk free trial, free returns, free shipping. Head over to sleepp. com.

2:03:00

I'll be right back >> and I will talk to Will. How you doing? >> Hey, how's it going? Great to be back. You're >> good.

2:03:06

Jordy just plays a sound queue and steps away.

2:03:09

uh uh what's new in your world uh just uh yeah way uh what's what's the latest?

2:03:18

>> Yeah, so about half an hour ago we did a big launch of something we've been building for quite a while at Prime and Elect.

2:03:23

Uh so for those who are have not met me, I'm Will Brown.

2:03:25

I work at Prime and Elect for a uh open source research company as well as a comput platform.

2:03:29

Um and we released something today called an environment hub which is for RL environments as well as eval.

2:03:36

And uh it's something we've been kind of working towards for quite a while.

2:03:40

But I think it's also something that like a lot of people are talking about, but also like wondering like what does this even mean?

2:03:46

Y >> um like I I saw like over the weekend like Google Trends for RL environments like shot up.

2:03:52

>> Um and you guys had made some joke tweets about this recently that I thought were great about like >> Ender, that's not an RL environment you were in.

2:04:00

That's you place real Amazon orders.

2:04:02

Um uh it perfectly encapsulates my level of understanding here.

2:04:06

So, I'd love to go deeper.

2:04:08

Uh, it feels like a bullcase for distributed infrastructure generally because we're maybe like it's maybe it's additive, maybe it's maybe it's uh replacing, but uh certainly less of the focus is on like get a 100,000 H100s in one single cluster.

2:04:25

There's more to be done in different places.

2:04:28

But try and concretize it for me a little bit more like like what is an RL environment? how are they being used?

2:04:35

Uh what are the what are the what are the trade-offs in terms of compute and design of the infrastructure um that that actually delivers some sort of valuable product at the end of the day. >> Sure. Yeah.

2:04:48

So I think we can get to the distributed aspect of it but just like first I think like an environment is essentially it's an eval like a lot of the things people make as like popular evals like Sweetbench or like ArcGI like these are environments.

2:05:00

They're a thing where you have some input tasks.

2:05:02

You have some kind of harness and then you have a thing at the end that looks at what your model or agent in the harness does and then says a score at the end.

2:05:12

And so this is like the exact setup we use for evals, but it's also what you use for RL training.

2:05:15

And so when people are talking about, oh, we're like going to make a bunch of RL environments, what they really mean is they're making evals that are designed for like agents to uh do some task interactively. >> Yeah.

2:05:27

is the Amazon order like we saw that report that uh some folks are building like full digital replicas.

2:05:32

I mean I guess it's already digital but like a like a like a a simulated environment of Amazon.

2:05:38

com or Door >> DAC are are those kind of the power law opportunities that people are really focused on?

2:05:45

Uh what are some other examples?

2:05:47

Are people um building these environments for like you know old old enterprise software systems that people want agents to interact with like how wide is this is the scale of RL environments right now?

2:06:01

>> Yeah, it's pretty huge actually.

2:06:01

So I think I think there is like these kind of like flagship website sorts of things, but I think there's also like a long tale of like increasingly niche applications that go down towards like these very fine grain narrow things all the way up towards like broad domains where models really struggle because we don't have a good like interface for them.

2:06:19

Like I'd say Excel is one of it's like everyone uses Excel but like no one is really like an LM power user in Excel because it doesn't integrate very well.

2:06:27

>> Um no we don't have the cursor for Excel yet as a broadly deployed thing.

2:06:29

Yeah, >> I think it's easier to have an environment for a terminal or for a coding agent because the harness is a little simpler.

2:06:36

And so that's one sort of thing you'd want as like an environment and that's one of the sort of things that like the mechanizers of the world I imagine are like building these sorts of things.

2:06:43

Um I'm sure that like the labs all have their idea of like what tasks their customers want models to be better at.

2:06:49

They're currently not being used for as well as just like what sorts of products they're people trying to build.

2:06:54

So like you can build the greatest agent harness in the world, but like if your model hasn't been trained for it, like some models might just like not be very good because it's not very ergonomic for them.

2:07:01

And so >> you can think of environments as like a way of taking a model like having a interface that you really want your model to be great in like a a harness and then just like a feedback loop that lets it get better at being a model or agent in that harness. >> Yeah.

2:07:16

Help me understand the Excel case more specifically. Uh Copilot exists there.

2:07:21

We're hoping that Microsoft rebrands it as Clippy eventually.

2:07:23

But uh there are also like seven different startups that are building uh you know cursor for Excel.

2:07:29

uh help me understand the context of verifiable rewards and what that would look like in Excel because if I if I'm trying to do an analysis, I'm trying to understand a company's financials, build a DCF, like there are templates, but then to actually get the correct valuation, that's something that like even if I have the perfect DCF and it tells me that Nvidia is worth five trillion, like I can't verify that against the market.

2:07:56

I could wait a year but the DCF could still be right and the market could be wrong.

2:08:00

So um like what does verifiable reward look like?

2:08:03

What are people actually building towards in that environment you think? >> Right. Yeah.

2:08:07

>> Right. Yeah. So the easiest recipe which uh like we've done work on as well as I've seen a lot of papers about this and it also just kind of like the obvious only real thing you can do is have a correct answer >> where it's like there is a gold standard like someone made this ECF y

2:08:22

>> um and we the agent doesn't see the finished one and it's like >> new enough that it's out of the training distribution from pre-training um like past six months is perfect >> and then you want to train your model to like be in the harness where it could make the DCF >> and then look at what it produced. And

2:08:36

And then you have all these different things you can check of like did it get all these fields right. >> Yeah.

2:08:41

>> Um and this those evaluations are a combination of like uh like vibe checks spot checks when humans are looking at it for like this is how kind of people have been doing historically is you just like you build a thing you try it out you look at it you try to make a list of what's wrong and that doesn't really scale.

2:08:56

scale. to scale it, you need to automate this evaluation trunk um where you want to have a thing that comes out like a DCF, a spreadsheet or a paper or a piece of code and have a way of programmatically checking this and this checking usually is going to involve LMS in the loop >> maybe fine tuned LMS or customized LMS

2:09:13

for grading whether the thing has been done correctly and then this like you got to iterate on this meta process a bit to have a good grader um and see like does the does the result of your grader match your human vibe check your spot check based on having some kind of experts in the loop. But what you then

2:09:26

But what you then want to do is take that kind of evaluation process and freeze it as a kind of piece of code >> and now you can plug this into your harness.

2:09:35

So the harness plus some like input task of like do this for every company um and your like gold standards for every company of like what a report should look like made by some analyst.

2:09:44

You're grading against like the golden answer and now you scale this up.

2:09:48

>> What was the industry standard before you rolled this out?

2:09:50

would would like the actual AI researchers or or AI engineers who were going to uh do reinforcement learning on a model in a particular domain actually design and implement the environment with the reward all internally and it didn't exist as an external function whatsoever.

2:10:11

external function whatsoever. Was that the status quo >> uh in terms of like open source infrastructure or like the in terms of like >> methods broadly >> very few people have been doing this the amount of people who have successfully trained a large scale agent model with

2:10:25

reinforcement learning is very small yeah I imagine >> it is not broadly deployed like the companies that are like >> on the periphery of maybe doing it are like cursor perplexity >> yeah so >> it's not a thing that like yeah when I hear a lot of like we're an AI agent

2:10:40

startup what that really means is like Maybe you're leveraging another agent or some sort of agentic API in that particular case and then maybe yeah you might be you might train your own um and maybe that's the is there a significant cost to this like does this affect if I if I just take the landscape of like

2:10:58

everyone who's building cursor for X right now that implies that at some point they like if we see this bifurcation and the and the market for cursor for X really is you know highly frag fragmented and there's not just one foundation model company that just eats everything. Um, are you going to see

2:11:15

Um, are you going to see small RL uh reinforcement learning environments with verifiable rewards in every single subniche of SAS that these companies are going after and would they have like costs to that? >> Is it expensive?

2:11:30

>> Uh, so build like I think you have to build the harness anyways.

2:11:31

I think one of the hardest things to scale is building good evaluations and rubric criteria.

2:11:36

Y >> if you can do the evaluation bit um >> depending on like how so one thing is like the infrastructure problem of like doing this for like a frontier scale model >> uh is not broadly accessible.

2:11:47

It's not the thing that you can just like put in money and it comes out.

2:11:50

That's what we're working on building.

2:11:53

>> Um >> but you can do it with like these like tiny models like the non-mix.

2:11:56

So kind of technical thing is like >> to serve models at scale you really want them to be like large mixture of experts models that are efficient for inference.

2:12:05

the current ecosystem for tooling for doing that yourself is not great.

2:12:08

Um most people are doing like llama and quen experiments and these models are like they're good models kind of but like they're not really what you want to deploy for a serious application. >> Okay.

2:12:19

So >> um yeah is that uh c can you give me a like a rough order magnitude for the for the level of cost that it would it would if I have a problem a specific domain where I think I need to uh train an agent in an RL environment.

2:12:35

I have the verifiable reward.

2:12:37

my AI researcher has, you know, defined the harness and the reward and I go to you to actually do uh the the RL on top of maybe an open-source model.

2:12:48

Am I am I talking about tens of millions of dollars of of GPU cost?

2:12:54

Like like try and ground it for me a little bit more like >> uh like I would say this is like napkin math. >> Yeah.

2:13:02

Like if you want to do like a serious run that would like meaningfully improve a model at like the Deep Seek or Kimmy scale, >> we're talking like hundreds of thousands for like a big boy serious one.

2:13:12

Um you can do a lot for thousands actually.

2:13:16

>> So it's kind of like Yeah.

2:13:16

Like those in terms of raw comput. >> Yeah.

2:13:19

Like when when people were like I fine-tuned llama on this thing and now it now it does this funny thing or I fine-tuned this image model on my face.

2:13:28

like that those used to be like kind of like proumer level projects.

2:13:30

like proumer level projects. Now we're getting into the okay this is like a enterprise level effort but it's probably not going to you you don't need to call up SoftBank to get it done >> right I mean there's like a there's a whole spectrum of these things that's why it's hard to like say like what's the bucket because you can like you can

2:13:47

fine-tune llama on your laptop >> but you also you can't fine-tune like Kimmy or DeepSeek the big ones on your laptop um and also it's like how long running is the task how many samples do you need how much do you really want to crank it like it's one of those things where you can kind of just like get more as you dump in more compute. >> Yep. >> Yep.

2:14:05

>> Um but currently people are spending a very large fraction of their total compute costs on experimentation and rebuilding the same Like there's not it's not easy to do this stuff like getting your GPUs to work correctly, getting your like libraries to work correctly, um deciding what hyperparameters to use.

2:14:22

like people spend and this is one of the reasons that like the labs have needed so much comput and why the expenses are so high is that like they have all these researchers and the researchers are all doing experiments and each of these experiments is like thousands of dollars of compute.

2:14:35

Um and you multiply this over a year and 100 researchers that's a lot.

2:14:41

Um >> okay sorry uh you you can finish. Sorry. >> Oh sure. Yeah.

2:14:47

Just that like I think there's some of these pieces that like once you can get the the hard questions answered once Yeah.

2:14:51

you don't need to keep redoing them for every new environment and you can kind of >> have some very cheap spot checks like like a kind of one shot eval like how good is the model at the start is kind of be a bell weather for like is the running going to work. >> Okay.

2:15:04

Loosely related to somewhat of the concept of generalizability.

2:15:06

I want your reaction to this Rune T this Rune post.

2:15:11

Uh Run said, "My bar for AGI is an AI that can learn to run a gas station for a year without a team of scientists collecting the gas station data set."

2:15:22

What do you think about that as a bar for AGI?

2:15:23

We don't need these RL environments for specific uh for specific tasks. What are your thoughts?

2:15:32

>> I mean, one thing you might have that look like is the like I think that could be answered by an agent who realized it's not good at gas station stuff.

2:15:38

Y >> and figures out how to kind of create its own training environment.

2:15:41

It's like, okay, I need to practice.

2:15:42

The same way that like if you want to learn how to code, you got to go find stuff to practice on.

2:15:47

>> Maybe that's one way of thinking about what that level of AGI means of like truly self-learning, like build your own tasks and curate your own tasks in a way that allows you to check whether you've been doing them correctly or not. >> Yeah. Yeah.

2:15:58

It does feel like we're in this era where there are a few really obvious things that we want to RL on and get really good at.

2:16:05

And you see that with the IMO and the math and the deep research reports and they're very reliable.

2:16:10

But then once you come up with some random task, then you need the gas station data set.

2:16:15

And so yeah, maybe the future is is uh go and find those autonomously and and set up the reward function train and then iterate and bake that in. Um >> yeah, but sorry. >> Yeah, sorry. Go ahead.

2:16:29

>> Oh yeah, just say I think one way to do this is just like the environment can be the thing you're going to serve like cursor is an environment.

2:16:34

Y >> um lovable is an environment.

2:16:36

Um all these things people are already kind of building things that could be environments if they connect the like the wiring correctly, but like >> I think we're not quite there yet in terms of having this be a thing that people are scaling and now like every company wants to like hire like environment builders and RL people. Yeah.

2:16:55

>> And it's like you're not they're not all going to hire people for this.

2:16:57

We need to kind of better ways of like scaling this out to more people and like infrastructure as a service kind of.

2:17:03

>> Yeah, that makes a ton of sense.

2:17:03

wanted to ask about uh Daario's comments in his interview with John Collison talking about uh basically I I'll read the quote.

2:17:12

He says there's two different ways you could describe what's happening in the model business right now.

2:17:14

So let's say in 2023 you train a model that costs 100 million and then you deploy it in 2024 and it makes 200 million of revenue.

2:17:21

Meanwhile, because of the scaling laws in 2024, you also train a model that costs 1 billion and then in 2025 you get 2 billion of revenue from that 1 billion and you've spent 10 billion to train the model.

2:17:30

So if you look in a conventional way at the profit and loss of the company, you've lost 100 million in the first year, 800 in the second, and 8 billion in the third year.

2:17:39

So it looks like it's getting worse and worse.

2:17:40

If you consider each model to be a company, the model that was trained in 2023 was profitable.

2:17:43

You paid 100 million, then it made 200 million of revenue.

2:17:47

There's some cost inference with the model, but let's just assume in this cartoonish example.

2:17:51

Um, and so Finn Twit has just been like freaking out about this, like being incredibly bearish around things.

2:17:57

It's obviously a b a bubble.

2:17:59

um you know and just kind of worried you have exponential uh increases in cost and then unpredictable kind of like capital uh needs in the future.

2:18:09

Why do you think they should stop freaking out given that you were just over at uh at Goldman >> uh Morgan Stanley but yeah so um >> like >> there's an element of the freaking out that's like kind of a good point which is like you can't keep doubling every year. Yeah.

2:18:28

at some point there's a plateau.

2:18:29

Um, and you can grow every year.

2:18:32

You can still have like exponential growth, but the percentages are not going to there's only so many people to adopt AI products.

2:18:36

Um, and once they adopt it, like if they're going to be spending more and more, it has to be delivering like a very large multiplier of new value per year.

2:18:44

And there's some laws of physics of like how good models are going to get, how like fast we can possibly make the chip.

2:18:51

Like the chips are only going to scale so quickly.

2:18:55

quickly. we can only build so many GPUs year after year with the current like rate of growth and so like I do think there will be kind of like the space we end up in where like stuff broadly works quite well there are like a few big labs

2:19:09

who have very good models that are like you can deploy them and then there's also like this kind of paro curve of like how much can you spend uh how fast does it need to be how hands-on do you want to be for it um like how like uh narrow is your like thing you want to do. So if your thing is like oh I want

2:19:27

So if your thing is like oh I want to build a great browser agent, this is super broad.

2:19:31

There's so many things you can do on a browser.

2:19:32

If your thing is I want to build an agent for writing Rust, this is much more narrow.

2:19:36

Um, and so like if you only if you're making like the product for rest developers like the uh cursor for example, you can make a model that's just as good as quad for opus that's much smaller and much cheaper most likely um because the the focus is like more narrow.

2:19:52

And so this is kind of how I think about generalizability is like you can spend more on compute, you can make the model bigger.

2:19:59

These are all different axes to spend and you can also make it like more narrow and go deeper on one thing.

2:20:03

And so we're going to have a lot of knobs to turn I think. >> Yeah.

2:20:08

>> Are you broadly long RL environments, short normal pre-training data?

2:20:12

Do you have a do you have any like extra color on the relative value or like where the lowhanging fruit is across those two?

2:20:23

>> I'm long both, but I think they're going to kind of turn into the same thing.

2:20:27

Like like one thing for example that people have been doing for like pre-training recently is that people are just using like deep cigar one and generating tons of like reasoning samples from it >> and then mixing those in at the end of pre-training and calling it mid training >> and it's like is in some sense this is RL because it's just taking the juice from doing RL on like deepse 3 but it's also pre-training.

2:20:48

Um, and so I think one thing that we're excited about with environments and scaling these is like you can get a lot of good data out of these by you have a like a task set.

2:20:56

You have the ability to generate agent data inside of this task set.

2:21:02

You have a filter for throwing out the bad stuff and keeping the good stuff.

2:21:06

And so can you just like do trillions of tokens of this and put it into pre-training? Why not?

2:21:12

Um, >> do you have insight into what's happening in the video or image world equivalent of RL environments, verifiable rewards, etc.

2:21:23

Like, uh, I I was just shocked by the level of quality of text in images and chat GPT.

2:21:29

Uh, we're seeing the same thing with Nano Banana from Gemini.

2:21:34

Uh, and it felt like it felt like images in Chetchup particularly was like uniquely good at Studio Gibli and uniquely good at text.

2:21:45

And that felt like I was I was almost like putting on my tin foil hat and saying like they're using Photoshop here and they're doing two different layers or something or like there's something else going on here that's just like a re one really good model. Like it feels spiky.

2:22:00

it feels particularly good at text and particularly good at cartoons, but it hadn't gotten like way way better at, you know, super photoreal imagery or whatever.

2:22:08

Um, so do you have a do you have any view on how you translate all the stuff that's happening in the in the agents and and textbased LLM world into sort of the diffusion and uh and image or video world? >> Sure. Yeah.

2:22:21

>> Sure. Yeah. So like I am not a diffusion expert or anything but like I >> I imagine it's a mix of essentially these kind of environments where you have some greater >> as well as like good old fashioned RHF like the stuff people were doing for like instruct GPT jackp era where it's

2:22:36

like you have like up vote down vote um and then you're training a roar model to like check these and I think the image domain is easier to kind of spot errors >> whether it's a human or an automatic grader like if text is wrong you can like do OCR on the image get the text back out and see Is it the right image from the prompt? Um, and so you it's

2:22:52

Um, and so you it's this is like pretty easy to verify versus like if a chat GBT answer is like slop, how do you verify that it's slop?

2:23:02

Like what's the algorithm that checks if it's slop or not? Like that's pretty hard. >> Yeah.

2:23:05

>> And so in some sense like the image models were just like >> for a while they were bad at things that we could very easily measure.

2:23:10

Um, and it's getting it's gotten harder to measure the things LMS are bad at much faster.

2:23:16

Um, and so I think some of this is like the oldfashioned tricks that made Jaxvt original version as good as it was being applied to the image domain for kind of patching up the the obvious fixes.

2:23:29

>> Uh, give me your hot take.

2:23:29

How much uh how much do you think Meta is paying Midjourney with that new deal they they announced? >> It's a good question.

2:23:38

>> You think it's I do you think it's nine figures a year? >> Hopefully.

2:23:42

I mean like a >> feels like Feels like it would be >> more maybe. I don't know.

2:23:46

They're spending crazy amounts on everything.

2:23:48

>> That's what I was saying.

2:23:48

And it makes sense for MidJourney to want to have like a low-key announcement about it, but there's like another team that would have been like, "We just signed a $500 million a year, 10year deal, or something something obscene obscene that uh that looks more like an acquisition.

2:24:08

Even though clearly uh from a from a dollar value standpoint, even though clearly um they're going to continue to operate independently.

2:24:16

>> Yeah, I would kind of think of it as like midjourney's version of doing an API business.

2:24:19

So like it could just be based on usage to to I'm sure there's a big picture component of it, but like think of like what you can kind of do the napkin map by saying like okay, >> what are other image providers like charging for API usage?

2:24:30

charging for API usage? um like Val or like uh um replic replicate or like all these other services like you can kind of map and math the cost or of an image or back it out from mid- during subscription prices and how much they give you and then say like okay how many of these are these people going to be doing if they're doing like generative ads for Instagram that's a lot if they're doing like every person has the

2:24:52

ability to apply it to their Instagram posts or their stories that's a lot and so you can the numbers can get crazy pretty quickly and I think it does kind of depend them like like if it's just ads or if it's just to like certain

2:25:05

customers or it's not like deeply integrated then it like the scale is not as crazy as if it's like no you're really getting like midjourney stuff everywhere for everyone instantly at volume. Yeah, I was uh I was thinking when uh

2:25:17

Yeah, I was uh I was thinking when uh Studio Gibli moment happened uh that the response from Meta should be to pre-render or pre-generate a Studio Giblly of every single person on Instagram's profile photo as a Giblly because that's basically what people were doing in Chat GBT.

2:25:39

>> Um like pre-generate that and then just when you when you open Instagram it just says, "Hey, we did this for you. uh do you want to share?

2:25:46

And that would probably be like the biggest day of usage on Instagram, but it would also probably bankrupt the company because it' probably be like $50 billion worth of inference or something like that. I don't know.

2:25:56

>> It's a lot of those images are expensive. Yeah.

2:26:00

>> Got some breaking news.

2:26:00

Nvidia beat on revenue and earnings per share, but is down 5%. >> Wait, why?

2:26:08

>> Um they didn't beat hard enough.

2:26:08

They didn't beat hard enough.

2:26:11

>> Can we ring the gong for the beat?

2:26:12

>> I mean, let's bring it's Yeah.

2:26:12

four and one respectively.

2:26:18

>> Nvidia's Nvidia's uh like a both a competitor and a supplier partner, right, to you like >> Yeah.

2:26:25

I mean, we sell we resell the GPU, so like we like we're big Nvidia fans.

2:26:29

>> So, you're rooting for them. That's great. >> Yeah.

2:26:30

And I think Nvidia has been very friendly to like the ecosystem of players.

2:26:34

Like Nvidia is not trying to be the like single you can even like buy GPUs on their website.

2:26:39

Like they don't sell GPUs to people.

2:26:41

They they sell them to like data centers and like big companies. >> Yeah.

2:26:46

And I I know that there was that news that they were coming for the Neoclouds with DJX Lapon and it felt like it felt like they were dipping their toe in that area, but it didn't feel like this was existential for anyone else in the in the ecosystem.

2:26:58

It felt somewhat additive.

2:27:01

I don't know if you have a take, but >> Oh, yeah.

2:27:04

So I mean I think it's also like a different like they're going to kind of offer that as like a very premium like white glove sort of thing to certain enterprises. >> Sure.

2:27:12

>> I think we are much more on the end of like >> get all the data centers we can find like partner with every Neocloud >> and have like really uh cheap pricing and then like build features on top. Yeah.

2:27:23

>> Um where part of this is like doing like core research.

2:27:25

So like we like have friends at Nvidia like we talk to them about like they're they release a lot of cool like open source um like research stuff and so like that is like very in spirit of like the sorts of things we are like aiming to do and I think we're like friendly. >> Yeah.

2:27:41

Uh are there any RL tasks that you think are like truly intractable?

2:27:46

something that like is is fundamental to humans creativity or comedy or something like what what what's the Mount Everest of of developing an RL environment around?

2:27:56

around? I think the hardest is stuff that and a friend of mine was tweeting about this earlier today actually um one of our collaborators who uh from Evals who's doing some cool open source eval work for computer use but things that need a human in the loop to have a like fine grained accurate simulation like stuff that if a human if models are not

2:28:15

good enough at replicating human behavior yet and that human behavior is key to the environment then you're not going to have an environment that faithfully captures the task because there's a kind of chicken and egg Um, and so like Twitch streamer was one example of like um like having a model that can like be good in Twitch chat kind of requires like an accurate sim of Twitch chat. >> Um, and it's like how do you build that?

2:28:37

>> Um, and it's like how do you build that?

2:28:39

Or like a physical streamer.

2:28:39

Like there's a >> this real time interaction problem you need to kind of solve before you can get started because the the scale at which these things move, you can do it if an LLM is simulating the user. Yeah.

2:28:50

>> You can't do it if you need a real human. >> Yeah.

2:28:53

Yeah, I feel like the uh the the time horizon is really it feels really really tricky to simulate when you think about like the the the impact on a life of a certain behavior.

2:29:03

I mean we struggle with this with like drug development understanding like does something in childhood affect you as a retired person.

2:29:10

It's like you, okay, now you have to create a simulation of the entire human body to understand that if you did this thing at 18, it's gonna cause you to, you know, uh, have your like knee blow out when you're 65 or something like that.

2:29:22

Feels much harder to simulate uh, one shot.

2:29:23

But maybe maybe we'll get there. Who knows?

2:29:27

>> Yeah, at some point you just got to let the error of time run forward and see what happens. >> Exactly.

2:29:31

Well, thank you for hopping on.

2:29:33

>> Always great to see you.

2:29:33

Congratulations on >> Check out the Prime Environments Hub. We are live now. uh or on Twitter X. See you around. >> Send it, chat. >> See you. >> See you guys. >> Talk to you later. Bye. >> Bye.

2:29:46

>> And yes, uh we have the Nvidia uh market chart here down uh 2. 75%.

2:29:50

Is that what I'm seeing here? Correct.

2:29:55

>> Um let's see how Nvidia is doing as well.

2:29:57

Um >> but uh high expectations, but you know, after hours it's down 1. 42%.

2:30:01

It's bouncing back right now.

2:30:03

The company's trading at $179 a share was at $181 a share.

2:30:09

Something tells me Jensen's going to shrug this off.

2:30:12

If you look at the six months, um the uh the stock has just been smoothly climbing upwards.

2:30:18

Um anyway, we have our next guest, Julia Steinberg, in the studio. Welcome. How you doing? >> Welcome back. >> Hello. >> I am doing well.

2:30:29

I was thinking about wearing like a little hat that said Chinese Communist Party on it to, you know, establish what I was talking about. Yes.

2:30:38

But unfortunately, I came up with this idea 20 minutes ago.

2:30:40

And because I am dealing with California infrastructure and not China infrastructure, I could not get it delivered to my house in time. >> Yeah.

2:30:48

You are you driving from LA to SF or vice versa? Where are you going? >> Yeah.

2:30:53

So, uh, this this may be a bit TMI for the good listeners of TBPN, but I was rearended a few weeks ago.

2:30:58

So, my car has since been fixed, but I must drive it up to San Francisco.

2:31:04

>> And I it's actually sort of thinking about with the drive north.

2:31:06

It's a lot of what Dan Wayne talks about in breakneck is sort of like, look, China's able to build these good things.

2:31:12

And I was thinking about I'm like, why do we not have a functioning highway for Highway One? >> Yeah. >> Why?

2:31:20

like PCH, Pacific Coast Highway.

2:31:23

It's one of the most iconic American landmarks. Yeah. >> One of my favorites.

2:31:26

Personally, it's ridiculous that it shouldn't work.

2:31:28

And I tweeted it just sort of out of frustration.

2:31:31

Like, I'm really frustrated that I have to, you know, do this ugly, annoying drive where my car is going to smell like cow for 6 hours rather than see the beauty of my favorite thing.

2:31:44

As a kid, my my parents, we'd be on a road trip.

2:31:46

My parents would say like, "Okay, we're going through the the the cow zone.

2:31:52

We're that big cattle ranch." >> Yeah.

2:31:53

You have to turn on the recirculating.

2:31:56

And I was I would just be like, "I'm hitting the window down."

2:31:57

And I would >> I if I were my child, you would have been like booted out of the car window at that point. Um >> I was surprised.

2:32:05

>> I was surprised. Uh so PCH between Malibu and LA was shut down after the fires for a while and then it was open to like specific uh people based on your address and then it got actually open pretty quickly and I think that was

2:32:22

probably nuome knowing that like this stretch of PCH is like critical to my popularity in the state because there's just a lot I mean for for no reason other and celebrities drive on that route and if they're and if he doesn't get it open so it was open ahead of schedule. >> He was taking a victory lap of like

2:32:43

>> He was taking a victory lap of like >> the one is not really critical to many people's commutes.

2:32:48

It's more of the >> Well, especially up by It's Big Su that's shut down, right? >> Yeah, Big Su.

2:32:53

So, I can drive from San Francisco or however far north on the one to Big Su, but I can't.

2:32:57

The sort of Cambria region is cut off that >> sort of it fell into the ocean.

2:33:02

And a lot of people on Twitter were sort of I wasn't expecting the tweet to blow up, but people on Twitter were like, "How can this like girl expect mother nature to want this freeway here?"

2:33:12

I was like, "This is absurd.

2:33:14

Like, we've sent men to the moon. we can't have a bridge.

2:33:17

Like I'm not a structural engineer, but Dan Wang in his book Breakneck, he was like, "Oh yeah, the Chinese government has built x amount of like the bridges like the top 10 some of the biggest bridges in the world in this like record short amount of time."

2:33:32

And it's it's frustrating to me that we can't do that.

2:33:36

We somehow is admitting that like as a civilization, we're just going to let the mudslides win. >> Yeah. Here's the solution.

2:33:42

If you want one the highway one to open, you got to put a Harris Ranch out there because that's the highlight of I5.

2:33:49

You stop for the A5 Wagu ribeye.

2:33:52

You get a big steak at Harris Ranch.

2:33:56

This was the highlight of my trips up and down.

2:34:00

>> Harris Ranch is the really smelly ranch.

2:34:02

>> That's the really smelly one.

2:34:02

It's >> I didn't realize you can pull over and get a meal.

2:34:05

>> Oh yeah, >> I didn't realize that either.

2:34:07

>> Oh, you're not doing the five correctly then. Yeah.

2:34:09

When you're driving from LA to San Francisco, you uh run out of gas halfway because you're driving a terrible car.

2:34:14

This is my experience in like 2012.

2:34:16

Um and you always stop at Harris Ranch.

2:34:18

You get a big steak and it's fantastic.

2:34:20

>> Yeah, I think most people are are flying by.

2:34:22

>> The other alternative that is actually scenic and nice is there even though we don't have a high-speed train.

2:34:25

We do have a train that goes through >> it takes 24 hours.

2:34:29

Like it just takes >> skill issue. Get a book.

2:34:31

I thought you were I thought you were into books.

2:34:35

>> You're the GM of books. >> I do like books. the general manager.

2:34:38

>> I like being able to read books while I'm not in a moving vehicle. >> Oh, it's so nice.

2:34:43

>> This is one of my more controversial book takes is I like >> being sitting still. >> Okay.

2:34:48

>> When I'm reading a book, planes are fine.

2:34:49

I can read many books on planes.

2:34:52

I've read many books on planes. >> Okay.

2:34:53

Well, speaking of books, take us through your review of Dan Wang's book. >> Yeah.

2:34:57

So, I sort of was introduced to Dan Wang about a year ago.

2:35:00

Dan Wang about a year ago. I was actually on the five listening to one of his articles put through some voice reader technology um how technology grows which is how I first got acquainted with the ideas of process knowledge which is basically that

2:35:14

engineering requires a great deal of like verbal uh transmission from one person to another or sort of one like elder me uh elder mentor to a mentee in the United States we've like lost a good deal of that process knowledge in how to make things because we've offshored it to China. Yep. Yep.

2:35:33

>> And Dan Wang really in and breakneck does a really a fantastic job of describing how these communities of engineering have maintained this process knowledge, this ability to build things which in the United States we've just voluntarily shrugged off and lost.

2:35:47

He I would say like the big idea if you're going into the Atlantic, this is the exert they chose was that rather than thinking of the United States and China as a capitalist society versus a communist one or a socialist one, we should think of the United States as a lawyerly society and China as an engineering society.

2:36:06

So in the United States, a lot of our elites go to top law schools.

2:36:11

You know, Supreme Court clerkship, one of the most prestigious things you can get.

2:36:14

Uh maybe even Supreme Court is one of the most prestigious positions you can get.

2:36:19

Whereas in China, the pilot bureau is made out of engineers first and foremost.

2:36:23

Uh I think Cinping doesn't have that big of an engineering background, but pretty much most people in the pallet bureau do.

2:36:28

And in the United States, when we focus around rules and regulations, what's getting us from point A and point B is like, well, what is this going to look like?

2:36:37

Is this going to violate people's rights? Can we do this?

2:36:41

In China, it's get from point A to point B.

2:36:42

It's like, okay, how do we do this?

2:36:44

And building something from the ground up. >> Yeah.

2:36:46

I was listening to Casey Hanmer on Door Cash and he was talking about how like he's trying to install solar panels in the desert and he needs to get some like environmental report that says that like if he puts a solar panel over it, it'll kill this tuft of grass that like a bird might come and eat.

2:37:00

Meanwhile, like if he was building like a chemical plant, it would be like fine because like like the chemical lobbies have like lobbyed to like get it make it easy.

2:37:08

And so it's just like nonsense.

2:37:09

But my my question is um the the the the engineering uh what what's the phrase he uses?

2:37:18

Engineer >> engineering society.

2:37:21

>> Engineering society and lawyerly society, right?

2:37:22

Um engineering empire or something.

2:37:25

Uh so I I understand that and it's very it's very alluring like I like engineering more than I like lawyer lawyers I suppose.

2:37:32

Um but my prior my prior is that America is number one and America's the best. Oh.

2:37:38

Uh, should we really try and change horses in the middle of a stream here?

2:37:42

Maybe the beauty of America is that we have so many lawyers, uh, keeping us great. >> What do you think?

2:37:50

>> Well, my my dad, who's probably watching right now, is a lawyer.

2:37:53

So, shout out to the lawyers keeping America great.

2:37:58

>> But, >> I guess the bigger question is like, is there a world where where we say, "Okay, let's let's try and pivot from lawyerly society to more engineering focused.

2:38:04

we get halfway there and and we actually are worse off because you want to just like lawyer max or engineer max and if you're if you're halfway in between it's just a nightmare. I don't know.

2:38:16

>> Well, how I view it is that the United States is very good at engineering maxing in the private market whereas China is very good at engineering maxing just having the state do it.

2:38:24

You know, Chinese Communist Party.

2:38:28

>> Well, both is there >> both but the state capacity for engineering >> the state capacity is a lot bigger.

2:38:32

But it's also like this is my biggest issue with breakneck is that it it was just like way too keynian.

2:38:36

It was like oh it's so great that the government is you know basically getting these people in the middle of nowhere in China to dig holes and by dig holes I'm sort of exaggerating a bit but it's digging these bridges that no one is going to use or these airports in the middle of nowhere that have fewer than half a dozen flights a week.

2:38:53

Like I just don't think that's a good use of market resources there.

2:38:57

It's it's a sort of interesting question that Wing raises though I don't think he says it straight up is like is capitalism even a useful sort of metric in the 21st century and I would say absolutely but we need to prove that in the United States by doubling down on free markets.

2:39:14

>> Free markets could bring us great things like a beautiful bridge uh that would make PCH work.

2:39:19

I I'm a huge believer in market incentives.

2:39:22

I think the issue that Wang sort of highlights though when he talks about siege and paying not wanting like a services economy is just that in the United States maybe the culture is a bit broken where our best minds want to do I don't know like B2B SAS rather than building bridges or building nuclear power plants or other infrastructure.

2:39:44

I do think the sort of culture is changing a bit.

2:39:46

We're returning a bit to wanting physical manifestations of American excellence that it's not just my app is like 0.

2:39:55

02% more productive than your app in delivering an AI girlfriend.

2:39:57

But I think that overall what Wang talks about in China's sort of engineering society, it it mixes two it conflates two things.

2:40:08

It conflates engineering and it conflates culture.

2:40:11

And when the government is in charge of culture and they can say everyone from this district should want to be a guitar manufacturer which is true in one of the districts in the book that he talks about which makes guitars which is pretty random considering that there's not really a historic guitar industry in this city uh in Gujo.

2:40:29

It's just something that they sort of was like we're just going to make guitars here and it happened and they got very good at it. >> Mhm.

2:40:36

But with the United States, culture is a lot more organic.

2:40:40

It's harder for the United States to say, "We're going to make this value really well adopted."

2:40:47

And this is something that's sort of bad about China, too.

2:40:48

And uh Dan Wang talked a lot about the one child policy in China.

2:40:53

He talked a lot about zero COVID in China where people >> John presented the multiple children policy, >> the three have three kids policy, and Dan rejected that roundly. Got destroyed.

2:41:05

I I like the three kids policy.

2:41:07

I think I think it's a >> I think it's a good policy.

2:41:11

>> They have the money, there's incentives, and and if you if you really really tilt the field the playing field in the favor of having more kids, like people will have more kids. >> I agree.

2:41:21

And it's like talking, of course, you know, I have to bring up Tavis, uh the engagement that happened yesterday.

2:41:26

I was talking to my boyfriend at the end of the day and I was like, Jake, like I I have to tell you something, but you're not going to be interested in this at all. >> Okay.

2:41:33

And he was like, "Oh my god, like what did she do this time?

2:41:35

Like did she get another car accident?"

2:41:36

And I was like, "Pravis and Taylor got engaged."

2:41:40

He's like, "I don't know who those people are."

2:41:41

And also like I'm really not interested in that.

2:41:43

>> He doesn't read the Wall Street Journal because Taylor >> Come on, you guys need to write something about >> Taylor wrote a 2014 oped about Spotify about from that.

2:41:51

>> Also, I want to know the I want to know what this engagement means for America.

2:41:55

>> It's a good question from from Arena from you guys.

2:41:57

But it's like this it's one of the weird things.

2:41:59

It's like a Taylor Swift baby boom.

2:42:03

>> Like how China is it's not who's a Chinese pop star who's going to get pregnant from a football player and have kids. >> I don't know. That's a good question.

2:42:12

We'll have to get to the bottom.

2:42:13

>> I do think there should be a Taylor Swift Orchid collab.

2:42:14

You know, she's on the older side.

2:42:16

I hope she has many kids.

2:42:18

Orchid could be a sponsor, you know, there and having D1 athlete that's also world famous superstar kids. >> That's true.

2:42:25

I mean, they're they're certainly they don't shy away from uh the economic opportunities associated with their um with their announcements. That's for sure.

2:42:37

>> No, but I but I do really think that the the Tavis sort of engagement announcement, the the royal wedding that will happen in a few years, like I unironically do think this will cause a baby boom.

2:42:46

Um because pe people are very mimedic, women are very mimetic and if number one pop star has kids, >> they found the answer.

2:42:54

Jackie Chan, born in Hong Kong, has two children. That's the answer.

2:43:00

We just need to promote Jackie Chan's children in China.

2:43:01

That will cause the boom.

2:43:05

>> Well, maybe the thing with Nepo babies, like Nepo, the sort of like weird take on Nepo babies is that we need to have more Nepo babies and just the out of them.

2:43:13

So it's like your kids, if you are successful, your kids will be successful, too. >> Yeah.

2:43:18

>> Because the line about kids now is like, "Oh, if you have kids, especially if you have kids young, they'll draw away from your success." >> Yeah.

2:43:24

>> But if I'm like, "Oh, my child is heritably going to be >> like a kung fu superstar because I am a kung fu superstar. I totally have kids." Force them to do that.

2:43:34

>> I mean, that certainly happens in Hollywood.

2:43:34

There's a ton of celebrities that have had that have kids.

2:43:36

I mean John David Washington uh is Danzel's son uh played football and then was in Ballers I believe and tenant.

2:43:45

>> I I have a sort of funny story about that.

2:43:47

I grew up in LA and I was on the debate team at my high school and there was a lot of sort of overlap between girls who wanted to do the debate team and girls who wanted to do the school play.

2:43:56

And this one girl whose parents were both very very famous like super A+ list actors was deciding between doing the debate doing the debate team and doing the school play.

2:44:05

And our debate coach, he said to me, he's like, "Well, it's usually pretty easy to convince them that they're not going to have a famous career as an actor and they're a lot more likely to become a successful lawyer.

2:44:15

But if both of your parents have stars in the Hollywood Walk of Fame, then that's a lot less likely." >> Makes sense.

2:44:21

Sorry, I'm prepping a chart.

2:44:25

Jordan, you have anything else? >> Chart for what? >> Our next guest.

2:44:29

>> Oh, >> who's in the ream waiting room? Sorry.

2:44:31

We've been keeping >> It's always great to see you.

2:44:31

We got to come on more often.

2:44:33

Yes, we should have scheduled more time, but we have we have someone else in the waiting room, so I'm >> Well, next time I expect a custom chart to be pulled up.

2:44:40

>> Next time you're in LA, come do the show in person.

2:44:42

>> That I said I said I'm here. >> Okay, fantastic.

2:44:44

Fix the car, drive over the accident. >> Yes. >> Thank you. Appreciate it.

2:44:51

>> Always great to catch up.

2:44:52

>> California to fix the one so we can drive scenically in the future.

2:44:54

Thank you so much for hopping on.

2:44:56

Our next guest is in the reream waiting room, >> Olivia Moore >> from Andre Horowitz. How are you doing? >> What's happening?

2:45:05

>> Hello, it's great to see you guys.

2:45:07

Thanks for having >> great to see you.

2:45:08

>> I was just sharing with the team the news today, but you break it down for us first.

2:45:12

What are you announcing today? >> Yeah.

2:45:15

So, we're announcing what we call the consumer AI top 100, uh, which is where we literally Thank you.

2:45:19

I appreciate when other people are this excited about data as I am. >> We love data. We live for market maps. >> Yes. >> Yes.

2:45:28

This one has I would say it's market map adjacent in that we every single website and every single mobile app around the world >> in kind of descending order of traffic and then we pick out the top 50 in each that are AI native companies to just get a look at what consumers are actually using.

2:45:45

>> I consider myself a bit of a list connoisseur myself.

2:45:47

Actually, I enjoy making lists from time to time.

2:45:49

Uh I'm sure you get yourself in hot water with this list.

2:45:53

People are going to say, "Take me off this list.

2:45:54

I want to be under the radar. Put me on this list.

2:45:55

I want uh I'm doing better than that other company on here.

2:46:00

Uh talk about the methodology.

2:46:02

Uh how confident are you?

2:46:04

Is there an element of like vibe and curation?

2:46:06

Is this like the New York Times bestseller list or is this >> No, there's no element of vibes. It's all data.

2:46:11

So we literally for the web list we go we have a provider called Similar Web and we literally just look at monthly number of visits. Yep.

2:46:17

We pick the first 50 that are AI.

2:46:19

And then on the mobile list, we do um sensor tower and we pick the top 50 by monthly active users, but we do get lots of uh fun texts, emails, inquiries of people either wondering about their rankings or wanting to be on the list.

2:46:34

So, I'm happy to inform everyone that it is completely unbiased.

2:46:39

>> There are some things on here that stand out to me as uh completely unsurprising.

2:46:43

ChachiPT number one, Gemini number two, Deep Sea, Grock up there.

2:46:46

Uh that seems obvious to me.

2:46:48

Uh stuff that's standing out are specifically >> tool like hyper specific tools like remove background, remove BG.

2:46:56

I've actually used that tool.

2:47:00

>> Janitor AI is number eight. >> Yeah.

2:47:02

What else stuck out to you as surprising? What did you learn? >> Yeah.

2:47:05

Every time there's um apps that are kind of like literally single purpose, remove. bg is a great example.

2:47:12

It was a company that was acquired by Canva but grew to like 32 million monthly active users literally by being the best product to upload a photo and get the background removed from it.

2:47:21

And it's a catchy URL that people remember so it keeps getting usage. >> Yep.

2:47:27

>> The janitor one is a good call out too because one of the surprises for me list after list is how many companion products make the list.

2:47:33

Like >> it's always a dozen or so each time.

2:47:37

Janitor is one of them and one of the highest ranking ones behind character.

2:47:40

So it doesn't seem to be >> this looks cuz you don't know.

2:47:42

It looks very uh grock x aai.

2:47:47

>> I would have thought janitor would be like back office B2B like I I I thought that would be something to clean up my data warehouse.

2:47:53

Uh but you're telling me that it is in fact a companion of some sort. >> Interesting. I also see spicy chat. ai.

2:48:01

That sounds like something we shouldn't push pull up on this stream. Cushion.

2:48:02

ai has a has a heart there that seems little uh suspicious. Juicy chat.

2:48:08

These there are a lot of these uh use cases there. Um is this a new thing?

2:48:13

Is this the first time you've done the top 100?

2:48:17

Uh like what are the biggest movers that you're noticing or or that you expect?

2:48:22

>> Yeah, this is the fifth list.

2:48:22

We do it every six months.

2:48:24

So we've done it for about two years now.

2:48:25

We started it like six months after Chat GBT came out.

2:48:27

Um >> interestingly, there's 14 companies that have made every single list.

2:48:33

Most of them are in kind of creative tools.

2:48:35

So think like Midjourney and 11 Labs. Sure.

2:48:39

>> Quite a few of them are general LLM players like FGBT, Perplexity, PO. >> Yep.

2:48:45

>> Um, one of the big I say surprise, but probably not a surprise to anyone who's been on Twitter is like Vibe Coding completely exploded on the list this time. So, Lovable, >> sure.

2:48:55

>> Wasn't even on the last list and now they're like number 23. Replet's also on there.

2:49:01

>> It's just below the cutoff.

2:49:01

Cursor is on there as well.

2:49:04

Um, and it's interesting because like you would think that de the developer market is much smaller than the consumer market.

2:49:10

So for developerf facing products to make an app like this is pretty impressive and it shows that like a a significant majority of developers are using them.

2:49:19

>> Well, I think the way that I look at that market is there's so many people in the world that have had an idea for an application without the engineering co-founder like the CTO.

2:49:33

And so that's just like decade decade plus now of like pent up demand for people that had an idea for an app and then all these companies whether it's lovable or replet or things like that have people on TikTok saying like you can make an app now or they're running ads that say like you can make an app right and so that's just like a flood of demand for people yeah non-developers becoming developers for the first time.

2:49:57

>> How are you thinking about Meta's strategy?

2:49:59

I noticed Meta AI ranked 46th.

2:50:04

That feels pretty low compared to the amount of effort Mark Zuckerberg is putting into that project, although it's early days.

2:50:08

Um, also the Meta AI app, I went to go test it.

2:50:11

I realized I had it downloaded because I had bought a pair of Meta Raybands and they kind of, you know, iterated on that product to turn it into Meta AI.

2:50:21

But, you know, LLMs are vended into Instagram and yet Instagram's not on here.

2:50:26

Uh, do you think that you'll see more companies that are in the AI bolt-on narrative era of their business kind of jump on here?

2:50:37

Do you want to keep them separate?

2:50:37

Is it What is your thinking about all of that?

2:50:41

>> Yeah, Meta is really interesting. It's funny.

2:50:43

The the larger tech companies I think have really uh stepped up in the past year or so on the AI front.

2:50:47

Like Google is probably the best example of this.

2:50:51

Um, Gemini was number two on the list.

2:50:55

>> They also had the AI studio on the list. Notebook LM on the list.

2:50:57

Google Labs which is where you get VO. >> Oh wow.

2:51:00

They have four four companies on four four separate entities on the list which were all ranked separately with their own traffic.

2:51:09

>> Grock had a big debut on the list from from X formerly Twitter.

2:51:11

Um >> Meta I think has struggled a little bit both on web and then on mobile.

2:51:17

I don't know if you guys remember the big scandal where they realized that a bunch of the users were accidentally posting their very private post to the public feed. Yes.

2:51:27

>> And so usage fell off a little bit. >> Interesting. Yeah. Yeah.

2:51:32

>> So they didn't even make the mobile list, but to me it's like if you guys have tried the Meta AI products, they're a little bit underwhelming.

2:51:37

Like quality of the images generated, the they did the weird celebrity companion thing.

2:51:41

So now that they're working with MidJourney, I would expect to see them make some improvements here, but it might take a little while.

2:51:48

>> Yeah, even the MidJourney thing, it just feels so natural in the other apps.

2:51:50

And if you have a billion people opening one app every day, like where Meta has been successful has been put stories into Instagram.

2:52:00

Put videos into Instagram.

2:52:03

They used to launch sidec car apps like Facebook had an app called Facebook Camera that was supposed to compete with Instagram and then they were like we got to shell out for Do you think Mid Journey's ranking gets hurt because people still use it in Discord or is or is they shift because it feels low at 28? >> Sure. Yeah. Yeah.

2:52:20

Yeah, they just recently kind of rolled out the web.

2:52:22

Um, >> I think that's true for both MidJourney and Sunno and a couple other of the creative tool apps that also have Discords.

2:52:30

It's funny because in the first iteration of this list, like no one had a website, everyone had a Discord and now a lot of those companies are really trying to shift their traffic over. >> That's interesting.

2:52:39

Yeah, I bet you could do a different list of just like the top Discords because all that's very quantitative and available and um even relying relies on even uh further data.

2:52:49

Um, Jordy, do you have anything else you want to take through this?

2:52:51

Um, I'm wondering I can continue. >> Go for it.

2:52:55

>> Um, I I I'm just interested in uh understanding uh your predictions for what the next version might look like.

2:53:02

What are you tracking on the earlier stage side that isn't quite ready for breakout adoption in ter as measured by unique monthly visits, but we might see an uptick of in the next year or two on the consumer AI top 100?

2:53:15

Yeah, I'm definitely expecting to see more in proumer and productivity.

2:53:19

We're I feel like both the models are just now reliable enough to actually do work for you or getting there and they're also agentic enough to like Perplexity Comic can draft an email and put it in your Gmail >> drafts folder which is amazing. >> Yeah.

2:53:36

Um, so we saw a couple companies like Manis make the list this time around that thesis, but I'd expect to see things like Fixer, Serif, >> and others in that category.

2:53:45

Gen Spark is another big one that I would not be surprised to see on the next list.

2:53:51

>> The other what's been happening to all the companies that actually were GPT rappers where they were for a while you could build a business to a few million of ARR probably more just basically.

2:54:06

I remember one >> just advertising against chat uh talk to PDF and >> yeah there's those but I'm saying there was other ones that would be called like open chat and it would be >> oh sure yeah yeah when you would search chatgbt >> and they would have inapp subscriptions

2:54:21

>> literal rappers GBT >> that was a loophole that was open on mobile so we we've only done the mobile list I think four times now and it was complete chaos the first three times because the app store was not policing >> what we call police there which was like

2:54:36

all of developers mostly abroad that were just kind of ripping off the free version of chat GBT putting it in an app and acting like it was the same thing you got if you paid for chatbt premium >> y >> uh and they were called things like chat and askai or chat GBT or chat GB GTP or

2:54:53

something but they finally cracked down on that over the last six months so now we have a mobile list that's like >> maybe more representative of true products but it's all like beauty cam filters now and scar do your homework and get answers. >> Oh, really? >> Oh, really?

2:55:07

>> What's your read on uh how Apple's done from a curation standpoint?

2:55:11

There's like a lawsuit happening right now between XAI and OpenAI and Apple.

2:55:15

I mean, this is probably going to get printed out and blown up on a big chart in court soon because Grock looks great here.

2:55:23

Um, but yeah, I mean, I I I'd love to know your thoughts on on on how Apple's curating uh the the app store.

2:55:32

Apple is a little bit struggling overall with AI.

2:55:34

Like we've even seen this with Siri and the fact that they're now potentially going to turn to Gemini and Google Google to fix it for them, which would be kind of unheard of.

2:55:41

Um >> I think they do an okay, but not a fantastic job with with curation.

2:55:45

It's definitely not like a payto-play situation from what I've seen or what I've heard.

2:55:50

But I also don't think that they're necessarily kind of at the cutting edge of of what's truly the best products that are coming out next that people should be spending time on. >> Yeah.

2:55:59

They also >> they should hire you. Yeah, they should.

2:56:02

Uh they yeah I mean they also obviously uh prioritize they editorialize in what they recommend and so I would be surprised if they're pushing companionship anytime soon since the lineage going back to Steve Jobs was this is a very clean productivity creativity >> well I would imagine gets promoted midjourney I would imagine gets

2:56:22

>> Apple coms is going to be every every day there's an article in the New York Times about how somebody had some something tragic happen to them while getting advice from a model and so Apple coms doesn't want to be like we recommended you recommended companions and the companions went haywire just very >> Anyway, we got to hop on with our next guest. Thank you so much for joining.

2:56:39

Thank you so much for joining. >> Thanks for having us. This is so fun.

2:56:43

>> Thank you for list maxing.

2:56:44

>> Thank you for list maxing. It doesn't get set.

2:56:46

>> Thank you for reading it.

2:56:46

I'm >> Let's hit the gong for the big list.

2:56:51

>> We hear we hear the gong is is too loud. >> Too loud microphone.

2:56:54

Sorry for your ears, everyone, but great great to catch up.

2:56:58

>> Thanks for hopping on. We'll talk to you soon. >> Talk soon. >> Bye.

2:57:00

Up next, we have Flo from lindy. ai.

2:57:05

Uh, big announcement today.

2:57:05

We will bring him in from the reream waiting room.

2:57:08

While we are waiting for him to join, let me tell you about adquick. com.

2:57:11

Out of home advertising made easy and measurable.

2:57:13

Say goodbye to the headaches of out of home advertising.

2:57:14

Only adqu combines technology.

2:57:16

Out of home expertise and data to enable efficient, seamless ad buying across the globe. Flo, how you doing?

2:57:21

Sorry for keeping you waiting.

2:57:23

What's new in your world? >> Yeah, good. Don't worry. Uh, I'm doing great.

2:57:27

Uh, you know, it's a it's a good day to build B2B SAS. Yes, let's go.

2:57:31

Let's hear it from BB SAS.

2:57:34

>> Finishing out the show with some just amazing news. Tell me more. >> Yeah.

2:57:39

>> Uh we we just announced a VIP coder today. I just saw Olivia here.

2:57:41

She mentioned Lovable and and Rocket and and a few of those. They've been blowing up.

2:57:48

Uh what we announced today is is is a VI coder that can test its own work. >> Okay.

2:57:52

>> And it's kind of it's kind of insane to say it because like you would assume like of course you can, but actually none of those products check their work.

2:57:58

Like imagine a software engineer that's just like here and it's like it doesn't work. Did you even test it?

2:58:01

And like oh no I didn't do that.

2:58:03

So basically these vodels rely on you to to test the work yourself and then like work with them to debug it and it's like it ends up being a lot of and painful work.

2:58:10

Um we >> are you talking about are you talking about generating a test suite in code or spinning up a web browser and interacting with the with the vibe coded product? It's it's the latter. Thanks for asking.

2:58:29

Is it okay if I share my screen?

2:58:29

I feel like it'd be better.

2:58:31

>> You can share your screen, but we are live.

2:58:33

So, anything you share will be shared with the internet for burned into the training data world manager. >> Please.

2:58:40

>> Um, >> yeah, Vibe Code does something right now. Let's see. Live demos. They never go wrong.

2:58:44

>> Well, I'm not I'm not going to do like a live live demo because it does take like three minutes or something.

2:58:47

So it' be a bit boring but like you know singular example >> uh built a scrabble checker >> and this is when we realized it's actually an emergent behavior that comes out of this agent that we gave a computer to.

2:59:02

It was like all right I built this scrabble wheel checker for you. Uh I'm going to open it. Okay.

2:59:06

I'm going to type a wheel that exists like thunder.

2:59:10

And yeah it tells me it's valid.

2:59:11

Now I'm going to type a wheel that doesn't exist like ZQXW.

2:59:12

And oops it tells me it's valid as well.

2:59:15

Um I notice there might be an issue.

2:59:17

It's showing ZQXW as valid even though it shouldn't been valid.

2:59:21

Let me fix the API to improve the validation. Fix the API. Try again. Doesn't work. Um okay.

2:59:25

So that's it does have its own computer, its own browser, and it sees the work and can click around and and actually test its work. >> Yeah. Very cool.

2:59:35

Um what what do you think the landing uh the landing not landing page but like uh what do they call it like the like the the landing market the first the first like product market fit moment?

2:59:47

Do you have any glimmers of hope?

2:59:49

Like is this going to be used by restaurants who want to spin up a landing page?

2:59:52

That was kind of like the previous boom of it wasn't vibe coding but when you saw folks uh stand up landing pages with uh you know Squarespace or or any of those Weebly like there was a whole boom in that.

3:00:03

It feels like there's another boom in in vibe coding right now.

3:00:07

Uh do you have expectations about where the sweet spot is?

3:00:11

Because I I would imagine that people aren't going to build like, you know, they're they're not going to go out and raise a hundred million dollars for a company that's built on the back of a vibe coded solution.

3:00:20

It's more of like a prototyping tool right now.

3:00:22

But then for small businesses, this could be something that they really rely on for a long time.

3:00:27

Like talk to me about the shape of the early customer.

3:00:30

>> I think that's actually the perception we're trying to fight.

3:00:32

I think the reason why VIP coding so far has been limited to these landing pages is because the tools have been so limited and landing pages are just like the the the fastest dumbest thing you can build.

3:00:43

Same for like this perception that like oh it's like a cute prototyping tool but you're not actually going to build your company on this.

3:00:48

You know we we the thing that is unlocked by this new paradigm of something as simple as like testing your work.

3:00:55

If you can't test your work you can only build very very simple stuff.

3:00:58

If you if you test your work, we're seeing instances of like this running for hours and building like one of the craziest examples we've built internally.

3:01:05

We call it like Lindy BNB.

3:01:07

So, it's like an Airbnb clone and it's functional.

3:01:10

Like you can you can select things, you can book them, you can put your place up for for for rent. Uh it's functional.

3:01:16

Um you can you can sign in, you can sign up, it's got Google, it's got all of that stuff.

3:01:20

So, um, you know, we're actually seeing a lot of, so yes, we're seeing the landing pages, but now we're also seeing a lot of like internal tools.

3:01:27

So, you're almost getting into like retool territory. >> Sure.

3:01:31

>> Because people are like, I'm just going to hook it up to my API and I'm going to build like oh calculators and like various internal tools for the for the team in like a couple of minutes. >> Yeah.

3:01:38

Do you think of this as like a separate product uh or does does this like kind of sit on top of everything that you've already built?

3:01:44

Um, and can you is there a way to like hook in the traditional like lindies into what you're vibe coding?

3:01:54

>> No, it's it's super integrated with the rest of the product that we've built, which is like agents obviously.

3:01:57

And so that's the whole value prop is like, hey, building your website is step one.

3:02:02

Like once you've done that, you actually have to build the rest of your business.

3:02:05

Like I I joke like building your website is sort of the equivalent of getting your your your business cards.

3:02:08

Like, okay, yeah, >> that you did that, now you need to like build a business.

3:02:11

So like now that you built your website, Lindy can like market it for you, find customers, do the entire customer support for you fully autonomously.

3:02:18

I I have want to actually launch Lindybnb and have it fully because I've got enough of one business.

3:02:24

I don't want to run like Lindb, but have it fully managed my agents, you know, and see how far we can go. >> Yeah. Yeah, it's great.

3:02:31

Uh yeah, keep sharing demos.

3:02:34

I'm I'm excited to check them out. Anything else, Jordy?

3:02:37

>> Yeah, just I I uh I'm interested.

3:02:37

you know, clearly I think it's almost contrarian at this point to launch another prompt to app builder, right?

3:02:47

And and you're insanely uh insanely intentional about everything that you do.

3:02:52

Um do you think that the category actually needs a bunch more people to try to be experimenting here? Right.

3:02:59

You see some of these revenue ramps.

3:03:00

I think a lot of entrepreneurs would assume that the market is is run away.

3:03:04

I should focus on something else.

3:03:06

But you've clearly made a different decision.

3:03:09

from the traction we're seeing there is plenty of room in this market.

3:03:12

I think it's it's like the very very very beginnings of this market as crazy as that sounds. Yeah.

3:03:18

>> Also and you know we've not told that story yet to anyone.

3:03:20

Uh the way this came about was like we actually built the V1 of this by accident.

3:03:25

So you know we had this released this release that like I went on the show a couple of weeks ago where we were like hey we gave a computer to Lindy and then we realized after we gave it to computer I was like oh it can build websites and you can QA them because that's what happens once you have a computer you can do everything.

3:03:39

everything. So then we we've basically packaged it up and released it as as LD build and and improved it obviously but uh I think it's basically an emerging capability of the platform and the reason why we've decided to release it is like well one we were like it's

3:03:52

it's already working and two we then started to look into uh the other products and we were like I can't believe none of them checks their own work like this is actually done that it's like a massive deal and and when we AB tested like okay same prompt on like lovable reply to Lindy this scrub

3:04:09

example for example every VIP coder out there it's a simple example it's like a very mini app every single VIP coder got it wrong every single one including Lindy like it got it wrong first shot as you saw and then the difference is like Lindy then tried it on work saw that it got it wrong and fixed it

3:04:26

>> so I I I think it's it's day one for this product >> that's very cool well congratulations on the launch keep us >> last uh last very quick question where are we on the Gartner hype cycle >> peak of inflated expectations trough of disillusionment, slope of enlightenment, or plateau of productivity. I have this conversation every day with

3:04:46

I have this conversation every day with with a different friend.

3:04:48

I I forgot the different labels here, but like look, you know, >> are it's basically are we are we are we coming down from the peak towards the trough or are we headed up the slope of enlightenment?

3:04:59

>> I think we're heading up the slope of enlightenment. >> Wow. Bullish. Bullish. Bull market continues. >> Very bullish. Very bullish. >> Very bullish.

3:05:06

>> Well, thank you so much for hopping on. We'll talk to you soon.

3:05:08

Have a great rest of your day. >> Thanks everyone. Bye. >> Cheers.

3:05:11

>> Uh well, if you >> shout out to Ilhan Vani.

3:05:15

>> Yes, shout out >> in the chat.

3:05:17

>> Yes, >> he is in Boston University >> at Boston University.

3:05:20

>> Let's hear it for being >> watching with the coast.

3:05:24

>> Five others >> having a window in Silicon Valley. Um >> bullish.

3:05:27

>> Well, let me tell you about Wander. Find your happy place.

3:05:29

Book a wander with inspiring views, hotel graded men, dreamy beds, top tier cleaning, and 24/7 concier service.

3:05:33

It's a vacation home but better folks.

3:05:35

Um, not gonna find a Wander on Lindy's vibecoded Airbnb.

3:05:38

He's got to solve getting good inventory.

3:05:43

That's what Wanderers done.

3:05:43

Um, Jeff Dean is I I don't want to say he was inspired by us, but he posted a beautiful uh oneofone uh AI generated trading card of him playing soccer.

3:05:55

He's the chief scientist at Google, of course.

3:05:59