0:02
I see a large IPM [music] on the horizon.
Search this video
0:09
You're [music] surrounded by journals. Hold your position.
0:22
[music] Overnight success.
0:28
The place of 30,000 double [music] right [music] there [music] information. declaring order inbound. >> Let's just roll.
1:09
We are surrounded by journals. Hold your position. Come on. Get up. Trust the experts here. [music] Five. Focus.
1:29
We are expert founder building.
1:43
I see multiple journalists on the horizon. Stand by. [music] Express. UAV online. Blaze. Double blaze. [music] Triple blaze. Double kill. Fight. Wrong.
2:30
Cup is [music] [music] team deathmatch. We are experts. triple blades.
2:59
That's [music] just wrong. Right.
3:07
>> Clearing order inbound.
3:13
[music] We [music] are surrounded by journalists. Hold your position. >> Strike one. [music] >> Strike two.
3:43
[music] Activate golden retriever mode.
3:57
>> [music] >> Marky clearing order inbound. Five.
4:19
I see more journalists on [music] the horizon. You're watching TVPN.
4:36
Today is Thursday, September 3rd, 2026. We are live from the TV.
4:43
Ultradom, the temple of technology, the fortress of finance, the capital of capital.
4:46
Let me tell you about ramp. com. Time is money. Save both.
4:49
Easy to use corporate cards, bill pay, >> accounting, and a whole lot more all in one place.
4:55
>> It's model mayhem, folks. It's model mayhem.
4:57
We got tons and tons of new AI model releases.
5:00
It's a great week to be into AI.
5:03
I think all the lab leaders, they got together.
5:05
They said, "You know what? >> People just love AI. Let's give them more.
5:10
Let's all team up to launch new AI for them the same week so that everyone has something just to be happy about, you know." >> That's right. That's right.
5:19
>> Uh we got Anthropic Fable 5. 1, we got Muse Spark 1. 3, we got Gemini 3.
5:22
8 8 flash open AI's tease an Astra is a GPT6.
5:28
There's a six where the S goes in one of the videos.
5:31
People will figure it out.
5:33
But the model mayhem is continuing.
5:36
Everyone got back from their long summers, their vacations.
5:37
They said, "We got to we got to launch something new.
5:43
We got to update this stuff." >> And Grock, Claude. >> Yeah.
5:48
>> And we're all down this morning.
5:52
>> People thought Astra might have escaped.
5:54
kind of a deflock moment for AI maybe.
5:56
What do you think's going on? >> Possibly.
5:59
>> Do we actually understand why all of the different models went down at the same time?
6:03
Because it's easy if it's like AWS went down and it took down a bunch of >> East US East one.
6:11
>> Why is Gemini down then?
6:12
>> No, no, Gemini wasn't down.
6:14
>> Oh, Gemini was never down. >> Oh, okay.
6:16
Okay, okay, that makes sense.
6:18
So, Amazon clearly very critical to the to the global internet.
6:22
Uh, good luck to the folks over at Amazon that are fighting the good fight.
6:26
Keep >> Couple big announcements.
6:27
One, >> which we got >> Shimat's birthday. >> It is >> big 5.
6:32
>> Should we sing full happy birthday?
6:33
>> I think full happy birthday.
6:35
>> Happy birthday to you. Happy birthday to you. Happy birthday, dear.
6:44
>> Happy birthday to you. [applause] >> Fantastic.
6:48
>> And allin conference is coming to Los Angeles, I believe, couple weeks. Very exciting. [applause] Sign up. >> Wait for our invites.
6:54
So, wait for our invites.
6:56
>> I think we I think you've been invited to pay and go if you want to go. >> Good to know.
7:00
Uh, >> equally important. >> Yes.
7:04
>> TBPN's Road to Christmas. >> Christmas.
7:05
[laughter] How many days? >> 112 days out. >> 112 days out. >> 12 days out. >> 112 days.
7:11
>> I got to say it's feeling like it's going a little slow.
7:15
>> Yeah, it I >> I wish there would I wish I was feeling more pace.
7:19
>> Okay, think about it this way.
7:19
We're only 13 days away from double digits. That's a big moment.
7:24
That's a moment everyone's gonna be talking about on the road to Christmas.
7:28
>> When we get to 99 days till Christmas, that's when you can start a countdown. >> That's big.
7:32
>> If you get a really big advent calendar, you can basically start that. >> Yeah.
7:36
>> Typically advent calendar starts December 1st. Why not 99 days away?
7:41
>> A Q4 advent calendar would be pretty elite by TVPN.
7:44
Little treats along the way help you hit those KPIs. >> That could work. That could work.
7:49
Uh, let me give you the roundup on the model mayhem that's going on.
7:52
First, let me tell you about CrowdStrike.
7:53
Your business is AI, their business is securing it.
7:56
Crowdstrike secures AI and stops breaches.
7:58
So, uh, Anthropic launch Claude Fable 5.
8:01
1 alongside the restricted Claude Mythos 5. 1.
8:04
Google released Gemini 3.
8:07
8 Flash and a cyber security focused version. That's good news.
8:11
Uh, Meta released Muse Spark 1. 3.
8:15
So, uh it might not be that much of a surprise that Enthropic seems to have the strongest model of the three with Fable 5.
8:20
1 scoring 66 on the uh artificial intelligence in index.
8:23
Uh that is the bar chart that everyone has been posting uh in this cycle.
8:28
It feels like uh we're sort of maybe getting to the end of the benchmark era.
8:34
It feels like when these models are released, it's much better to solve a novel math problem or do something else.
8:41
The Pelican on the Pelican on the bicycle is still still one of my favorites. Yeah, I love that one.
8:48
But yeah, um >> it changes everything. >> It does.
8:50
But the benchmarks, you know, they've been accusations of bench hacking, odd hard to interpret many of them.
8:58
>> Very very low trust in benchmarks.
8:58
And at this point, everyone has had enough experience using various models. Yep.
9:04
>> They have their own sort of internal benchmark. >> Yeah.
9:06
And so the the the demos of like I built this game, I did this thing with it.
9:11
Uh and then also just the trusted voices of people who you know use a bunch of these models and they kind of give you the breakdown of what they like, what they don't.
9:17
Uh that has been where people lean a lot more.
9:20
But the artificial intelligence uh the artificial analysis intelligence index, this bar chart that you see has been a good way to kind of compress down a bunch of benchmarks into one meta benchmark. So uh Fable 5.
9:30
1 got the highest result ever on the index.
9:33
The score is also ahead of Opus 5 which got 63 and Fable which got a 62.
9:37
Anthropic says, "Bable 5.
9:39
1 is also cheaper and more efficient, made possible by an improved caching system.
9:43
Should make ordinary workloads 25% cheaper and long horizon agentic jobs 45% cheaper."
9:48
The company says that's good news.
9:51
Uh, interestingly, as uh and as Ben Thompson pointed out, Anthropic is also sort of dropping its no zero data retention policy.
10:00
Um, which there was a whole news cycle around a few weeks ago.
10:03
People were saying, you know, why is Fable not taking off in adoption?
10:07
It's a really a great model.
10:08
And there were a bunch of different explanations.
10:10
One of them was companies demand zero data retention.
10:16
They don't want closed source AI labs to be hoovering up their private information. >> Yeah.
10:21
And I think they said functionality that will allow for >> data retention, but it's on servers and infrastructure that the company owns. Yeah.
10:30
>> So, you do keep some of the data.
10:30
you still are monitored for hostile usage but it's not going straight into anthropics databases.
10:38
So Alex Karp and Sautinella both warned against this idea that models uh that uh companies should have data sovereignty.
10:46
So the policy is going to be replaced the no zero data retention no ZDR is going to be replaced with something called EFS enterprise frontier safeguards.
10:56
uh and uh and that may have attributed may may have contributed to lower fable adoption among enterprises and it sounds like it was a direct response to user feedback.
11:07
So good news that you know the people spoke and the companies listened.
11:09
So uh over in Google world Gemini 3.
11:12
8 Flash is the company's third flash release in six weeks.
11:18
They are flashing out these flash releases. It scored 73.
11:20
7% on deep sui just behind opus 5 and competitive with models that cost several times more.
11:27
Uh independent testing giving gave it a 59 intelligence score which isn't the absolute frontier but it's a great result for a model generating roughly 300 tokens per second.
11:38
So very quick and very good at coding at least on this particular benchmark deepu we'll see what adoption looks like uh and and where enterprise spend goes.
11:48
Uh Ara Karazzian has some very interesting uh data from the ramp economics lab you can go check out.
11:53
He also has a new post that's very interesting that we can talk about in a second.
11:58
But uh last model Meta Muse Spark 1.
12:00
3 did very well on benchmarks scoring 75. 4% higher than Gemini 3.
12:06
8 Flash on deep sweep beating both Opus 5 and GPT 5. 6 Soul.
12:11
It didn't sweep the board.
12:11
Opus still beats it on several professional work and computer use evaluations but it got six2 on the intelligence index which is only behind the newest claude models.
12:22
Uh tons of stuff to think about and discuss here.
12:23
So um the interesting post from Razian and I don't know if we have it in the timeline if we can pull it up but he was saying that there's a lot of concentration in the enterprise uh AI revenues right now.
12:35
uh OpenAI and Anthropic 80% of their enterprise revenue comes from just 1% of the companies.
12:41
And I was like 1% that seems crazy.
12:43
And and and he notes that this is uncommon for software categories.
12:46
Like if you look at CRM, if you look at databases, if you look at all sorts of different software spend, typically you don't see as much concentration, you don't see 1% driving 80% of the spend.
13:01
And I was wondering about this.
13:01
And I was wondering about this. And so I started looking up like what where else do we see this type of inequality if you can call it that this distribution this power law power laws are everywhere but where else does this exist and you might
13:14
go to hiring like is AI a drop in replacement for hiring is it going to be proportional to hiring and in fact 1% the top 1% of biggest companies in America they do hire a ton of people the top 1% of American businesses employ 65% of the total workspace work workforce uh not 80%. But interestingly,
13:33
But interestingly, >> the concentration risk there, John, 65% of jobs are tied to just 1% of companies. >> There is. Yeah, there is.
13:42
And and and I mean, yeah, you definitely see that.
13:44
Uh although the top 1% companies tend to be pretty lindy, you're talking about >> No, no, I know. I'm joking. >> Yeah.
13:52
The government and whatnot, but um >> the interesting 1% 80% uh correlation comes from sales.
13:59
So the top 1% of American companies by sales generate 80% of total revenue total revenue.
14:06
And so there's this weird dynamic where I don't know exactly how correlated it is, how causal it is, but uh there is an interesting dynamic there where it feels like if you look at the total AI spend, it's around uh 150 billion a year something like that.
14:23
Uh and then you look at total revenue for all US businesses.
14:28
AI is roughly a quarter of a percent of total US business revenue and it tracks fairly closely to the revenues of those individual firms.
14:40
So you see that that 1% of the top businesses generate 80% of the revenue. They also spend 80%.
14:48
They also generate any 80% of the AI revenue.
14:51
And so there's this interesting dynamic where because enterprise AI particularly, you're not going to be on the $20 plan.
14:57
You're not going to be on the $200 plan.
14:59
You're going to be consumptionbased and you're going to look at it a lot more like a marketing line item that's proportional to your revenue potentially.
15:07
That's at least one interpretation of this.
15:08
Uh uh another uh fellow over at RAMP said that this is a roar shock test for how you feel about AI.
15:16
Either you look at this and you're like it's great or you look at this like it's over.
15:20
Um, but fun fun fun uh fun chart to dig into. Anything else on this?
15:26
You guys read any of this? No.
15:29
>> Uh, >> let me tell you about codeex.
15:31
>> I just was realizing >> Codex is a powerful workspace for getting work done with AI agents.
15:35
Whether you're writing code, analyzing data, creating content, or automating business workflows.
15:38
Codeex helps you move projects forward from start to finish.
15:43
>> Uh, our guest today we forgot to cover.
15:45
Uh, Pablo Tore joining at 11:30. 11:30.
15:50
[laughter] uh to talk about the Clippers uh and Kawhi Leonard.
15:57
>> I mean, so these are people that go they watch live streams and they clip them and they put them out on social media. >> Yeah.
16:02
Why I think that's what the team why they named the team that. >> Yeah. Yeah.
16:05
Kind of an homage because there's a lot of clipping that happens in LA.
16:08
Tik Tok clips, Instagram clips. So they call >> Yeah.
16:11
This whole Balmer Kawhi Leonard thing.
16:14
Uh, you had an interesting pronunciation of Kawaii's uh, name earlier because I don't think you'd ever heard of it.
16:21
>> Well, I was calling him uh, Steve Bal.
16:21
I I I dropped the R cuz I thought it was French. Oh, nice. No. >> Yeah.
16:27
Uh, and then we got a bunch of others.
16:29
Somewhat of a lightning round.
16:30
We have Moheit from Siphon.
16:30
He's got a big funding round from Altimter.
16:34
We have a Shay from Pocket.
16:37
Sold over 200,000 devices.
16:41
Uh we were talking to Jimmy yesterday.
16:43
He was saying, "Why can't uh big companies do hardware while Pocket's doing it? They're making it work."
16:48
Jimmy >> did say that he sold he was like, "Oh, Meta sold 2 million pairs of glasses.
16:53
I sold 2 million pairs of headphones in Brooklyn, >> which was a great line."
16:59
Um and a bunch of other great uh uh teams joining.
17:05
>> And we'll cap it off with the CEO of Snowflake coming off a great quarter.
17:09
Yeah, stock is way way up. Uh, very exciting.
17:12
Let me tell you about MongoDB.
17:14
What's the only thing faster than the AI market?
17:15
Your business on MongoDB.
17:15
Don't just build AI, own the data platform that powers it.
17:20
So, uh, the other big story in the news, the opensource community is stronger than ever.
17:25
We saw uh, three, four, who knows, five closed source releases.
17:31
All the all the big labs are duking it out.
17:33
Meanwhile, Nvidia is going even bigger on open source with the 13 billion dollar acquisition for hugging face.
17:39
Uh, all over the timeline.
17:42
This was leaked a couple weeks ago, I feel like. Rumored.
17:46
What are you laughing at?
17:47
>> No, just every single day I would see a headline about Nvidia hugging face and think, "Okay, now it's official." >> Yeah. Now it's official. >> Today is the day. They can talk about it. They can explain it.
17:56
Uh, and there's a lot that makes sense.
17:58
There's not too many questions.
18:00
It's just a great outcome generally.
18:01
But it's an interesting story because it's a true 10-year overnight success.
18:06
Um, and it and they're having fun with the acquisition price.
18:10
Uh, Nvidia agreed to pay 12.
18:12
93 billion for HuggingFace, which just happens to be the exact decimal code for the hugging face emoji, which of course is the icon used by the company.
18:24
Uh, and then also if you take that number and you turn it into a color code, I think you get a green that sort of hints at Nvidia.
18:33
So, they're having fun both ways.
18:35
Like, ah, this was always in the plan. Uh, symbolism.
18:37
Uh, and it's just funny to be having fun with a price this big.
18:43
I remember the Instagram acquisition and the idea of a billion dollar outcome as being insane during the social media boom.
18:51
And now uh we're seeing like you decacorn liquidity events every couple weeks.
18:56
Uh I'm of course thinking of open router and >> it's honestly an incredible time to be investing in AI. Seven years ago. [laughter] >> Yes. Yes. Best data plan.
19:08
>> Three to seven years ago was an amazing time to be investing in AI.
19:11
>> It was um and so obviously this is this shouldn't come as that much of a surprise because Jensen has been the probably the loudest voice on open source.
19:20
He put out that open letter that everyone signed on to uh and uh he wants to maintain Nvidia's dominant position in the AI ecosystem uh both selling chips to closed source labs uh who might wind up making their own chips as well.
19:35
And there's a whole tug of war there but for open-source uh AI development uh he wants to be the place where developers and companies go to pick their models and then hopefully rack them on GP on Nvidia GPUs.
19:46
Uh the simple distillation is just Hugging Face is the GitHub of AI.
19:51
It's a little more complicated than than the Microsoft GitHub deal.
19:53
Um but it still makes a lot of sense.
19:55
So uh Hugging Face doesn't own the smartest models.
20:00
They don't even try to build them.
20:01
And there's some interesting financial dynamics there about how capital efficient they were because of that decision.
20:06
Um but they created this nexus for people to upload, discover, test, modify models.
20:12
And the numbers are good.
20:16
uh they have over 18 million developers, 200,000 companies using the product, three million models and over half a million data sets.
20:23
And so uh they have certainly created this this vortex of activity that's really valuable.
20:28
Uh how much how strong is the network effect?
20:32
It it it's it's you know it's certainly cooking and it's certainly driving a lot of value here.
20:36
[snorts] So um the interesting thing about hugging face is that it did not start as an AI GitHub for AI.
20:42
It started as a completely different idea.
20:45
The founder worked at a French computer vision startup called Moodtocks that was eventually acquired by Google and in 2016 he teamed up with two co-founders.
20:54
One who was a mathematician and the other one who was a scientist who had worked in patent law apparently uh and they started building an AI that could basically talk about everything.
21:01
So, this was post Siri, post Alexa, but instead of focusing on like tell me the weather and be be an assistant, uh, you know, set a timer, he wanted just to be able to talk to you and >> still not solved, by the way. >> Wait, which one? >> Siri. [laughter] >> Yeah. Yeah. Yeah.
21:20
So, yes, it's pretty good at setting timers.
21:22
Um, but so the goal was to build something like a Tamagotchi, something very cute.
21:26
Hence the hugging face icon.
21:28
uh a funny, emotional digital friend targeted at teenagers.
21:33
Uh the app let their name let users name the bot, text it, send selfies, trade emojis.
21:38
Uh it was explicitly marketed as an AI best friend.
21:43
Um for bored teenagers and they scaled it.
21:46
I think this is pretty significant.
21:48
It was doing a million messages a day.
21:50
They had more than 100 million messages in total by 2018.
21:52
That seems pretty significant.
21:55
That doesn't seem like, you know, you're languishing in the app store with no downloads.
22:00
Because how many messages a day can a bored teenager possibly put up with a with an AI agent?
22:05
Even if it's like a thousand, you still have, I guess, a thousand users, power users.
22:08
I don't know, probably the average user is doing 20 messages a day.
22:13
So, you're you're seeing pretty significant adoption.
22:15
Um, and so they were able to raise a series of financing rounds.
22:19
The big one that grabbed headlines was Kevin Durant uh was in the $1. 2 $2 million.
22:23
I think it was a preede round.
22:25
Uh Beta Works and SB Angel were also in there. That was 2017.
22:30
Then they did a proper $4 million round uh with Ron led by Ronnie Conway's A Capital in 2018.
22:35
Uh the company had raised money and the technology worked well enough to feel sort of magical, but it was still 2018. This is preGPT3.
22:46
Um and it didn't become a durable consumer business.
22:48
So in 2018, Google released BERT, which was sort of the first language model, very primitive, but uh people were really excited about it, but it wasn't delivered just as weights that you could download on the internet.
23:00
It was it was delivered as a paper from Google.
23:02
Um and uh and the paper was implemented in Google's TensorFlow framework and people like PyTorch.
23:08
So the hugging face team converted BERT from uh from TensorFlow to PyTorch and released the conversion for free.
23:16
And so developers really liked that and that became sort of like the initial go to market flywheel for developer adoption.
23:22
Uh and eventually they added more and more models eventually thousands.
23:26
Now I think they have millions of models which is sort of crazy but when you think about all the forks and fine-tunes it makes sense.
23:31
Uh eventually the team stopped trying to build this one application and focused on building tools became the picks and shovels trade.
23:39
So instead of trying to pick a winner you just host every model.
23:43
They became the Switzerland of AI to some degree.
23:45
Um and the and so the flywheel started compounding more models, more developers, more model creators, more companies.
23:50
Uh and it was the ba same basic network effect as uh GitHub.
23:55
The you know GitHub was the default home for open source software.
23:59
Hugging face very quickly became the default home for AI models.
24:01
Over time, HuggingFace grew from a code library to a place where developers could publish models, version them, attach data sets, uh discuss changes, and they even allowed them to build uh demos.
24:12
Hugging face eventually launched a spaces product where you could demo these uh the different models.
24:18
Um companies could maintain private repositories. Same GitHub strategy.
24:21
Um so 2019 Lux comes in with $15 million.
24:25
Then then yeah Lux got in early series A 15 mil.
24:29
Uh they also came back for the series C in 2022.
24:32
That was hundred million at a $2 billion valuation.
24:35
Sequoa and CO2 were in that round.
24:37
There was also a series B in 2021. Uh that was 40 mil.
24:39
And then the big step up was in August of 2023.
24:44
Hugging Face raised $235 million at a $4. 5 billion valuation.
24:47
And it's a murderer's row of potential acquirers.
24:53
You got Salesforce, Google, Amazon, Nvidia, AMD, Intel, Qualcomm, and IBM.
24:58
So you're, you know, it's not like they were doing a road show to sell the company, but it's very much like we want to be the Switzerland of AI.
25:04
We want good partnerships with everything.
25:06
We're going to be chip agnostic.
25:07
So yes, we have Nvidia on our cap table, but we also have AMD and Intel and Qualcomm.
25:13
So, you know, you you can count on HuggingFace as being like an independent place.
25:17
We're not we're not purely Nvidia backed, which is maybe one of the things that they'll have to deal with now, but they're not purely Nvidia backed at that time.
25:25
So, it's very much like, oh yeah, we'll host a model that runs well on AMD, we'll host a model that runs well on Nvidia, we'll host models that are from Google, from Amazon, etc.
25:33
Uh and so uh it looked like this like peace treaty moment from the major AI infrastructure companies.
25:40
You get everyone around the table, everyone's aligned with the mission.
25:44
Uh and hugging face becomes this neutral territory where uh it supported competing clouds, chips, frameworks, models, no simple no single company could control the platform.
25:52
And so even though they did a number of rounds, Hugging Face I'm going to say only raised under 400 million which is a lot of money but not at a12 billion outcome.
26:03
uh and it's pretty pretty small considering the outcome and it was very capital efficient because they weren't actually buying chips or serving models directly.
26:10
Uh and they had this flywheel that sort of spurred growth through the network effect naturally.
26:15
So not a lot of cost in the business.
26:17
They became profitable in 2025.
26:19
Still had half the money that they raised.
26:21
So Nvidia came in to offer 500 million late 2025 at a seven billion valuation, but they turned it down. >> Whoa.
26:29
turned it down because we said, "Hey, if we're going to go deeper with one particular area, it's got to be the whole shebang."
26:35
And so, uh, that's what what wound up happening.
26:38
So, pretty high revenue multiple.
26:41
>> My question My question is, um, I wonder what Jensen's vision for Hugging Face is.
26:46
Uh, they they do offer model routing. >> Yeah.
26:50
>> Uh, and and you know, they they rank a bunch of inference providers.
26:52
Is this something that could we see hugging face and open router and ramps router competing more and more? Yeah. >> Um, >> sure. I think it's two things.
27:00
I think I think one is if like closed source is already its own business line, sell chips to the labs, but the labs are building AS6.
27:09
They're doing a lot of stuff and there's this whole back and forth tugof-war there.
27:12
But on the flip side, you have open source which is continuing to grow.
27:16
And if you can be sort of the front door to that and then say hey you're you're you found your best model your framework on hugging face and now you're ready to go and buy chips or buy inference or buy compute and Nvidia's right there.
27:28
and Nvidia's right there. that is a very logical flow and anything that they can do to make open-source powerful exciting a place where where you can build a career have a great outcome I think that's beneficial to uh to Nvidia because a lot of people will see this and say yeah like you can go and build a
27:50
great company and have a fantastic outcome I mean the retention packages are apparently a billion dollars for a pretty small team I think it's in the hundreds still and So, uh, if you're if if Nvidia is just trying to send a massive signal to the world that you can make it at in the open-source world, that's a really good signal to send. And
28:09
And it feels like it's landing loud and clear, especially today.
28:13
Um, what else is on your mind regarding hugging face and Nvidia?
28:17
And while you think about it, I'm going to tell everyone about Console.
28:20
Console builds AI agents that automates 70% of IT, HR, and finance support, giving employees instant resolution for access requests and password resets.
28:30
What do we got on the timeline?
28:34
>> Uh people are still waiting for an official launch from uh of Astra.
28:37
Lisan Algib is sharing some Astra benchmarks. Arc AGI 3 98. 6.
28:47
He had uh they had previously said, "Are you ready for a nuke to hit Arc AGI 3?"
28:53
>> Uh so almost fully saturated.
28:53
Frontier math tier 4 v2 gets a 97. 6. Deep SW 74. 1.
28:59
Exploit bench also uh saturated at 100%. So >> wow.
29:06
>> Uh seems pretty good, but still no official announcement. >> Okay.
29:10
Well, we will keep monitoring it.
29:13
Let's go through what else is in the timeline.
29:14
What is John Palmer saying these days?
29:17
He says, "I actually think Snapchat for work might be a good idea in today's big companies.
29:21
Work platforms, work communication platforms have always followed what teenagers were doing 10 years ago.
29:27
I did use IRC when I was a teenager."
29:30
Unst Tyler has to look it up.
29:33
He doesn't know what internet relay chat is. >> Wow.
29:39
>> I for what it's worth, I didn't use IRC either. >> No. What did you use? AOL. >> You didn't use AIM. AOL instant messenger. Wow.
29:47
Y >> What was your first communication platform on the internet? >> Email. >> Email.
29:54
>> I remember I remember being inundated with emails.
29:56
I remember there was like a summer there was a summer.
30:00
>> He's laughing at me cuz I had a hot mail. Wow.
30:03
>> Well, I I just I remember spending a summer as like >> not even a teenager yet, just being super stressed about my inbox cuz like every kid had just started using email. Yeah.
30:12
>> And so they were just sending these super long emails and I would and I would be think I'd just be like playing outside in the grass and thinking, man, I gota I got to check my email.
30:20
>> Do you think there's anything actually to this?
30:21
It sounds like he's being serious.
30:22
In the age of AI slop, the most efficient form of communication is just short videos of yourself speaking. What do you think?
30:29
>> I would love to use Snapchat for work.
30:30
>> So if we if we said, "Hey, as as a team, we're going to communicate through short selfie videos." >> Yeah. With the filters.
30:38
with the you gotta have the dog filter or whatever the hot dog filter, whatever those were. Uh he says he's serious. A two-minute demo video.
30:46
Yeah, I guess I guess in terms of actually just taking a video of your screen showing people what what what what you're working on.
30:53
You know, there's a lot of different things you can do.
30:57
Uh there's a town that's for sale.
30:59
6 million bucks gets you an entire city >> in California, no less. Not a place to live.
31:06
>> You know what I'm thinking? What are you thinking?
31:07
[clears throat] Data center developers [laughter] have been having some issues, right?
31:09
Towns, they they they try to move into a town, start building the data center.
31:13
The town says absolutely not. Well, here you go. Buy the whole town. >> Buy the whole town.
31:18
>> Who's going to tell you no if you are the king of the castle? >> The town next door.
31:21
Maybe they might be annoyed. I don't know.
31:22
Anyway, >> king in the castle.
31:24
King of the [laughter] castle.
31:26
>> There's a there's a video here on Good Morning America breaking down the details.
31:29
An entire town went on sale for >> 2. 6 is TBPN city.
31:32
I did get excited about the potential of just welcome to TBPN California, but >> yeah, >> I'm not super eager to move.
31:43
>> This has Riley Wall's project written all over it.
31:47
>> Started with the street.
31:49
>> He's got to do Who who bought the street by the way? >> Notion. >> Notion. The Notion way.
31:53
I think that was what what it's called. So Notion, California. Maybe it's right there. Uh it's official. Hot bot summer is over.
32:04
Super Grock is saying goodbye to companions.
32:07
I remember we were debating, you know, there's obviously a lot of people that had ethical concerns about about AI, romantic companions, but we were more discussing just like is there actually a business here once you break the seal of like, okay, I'm we're we're doing it. He did it.
32:22
Um would it actually be successful?
32:25
Because Replica has seemed to get to scale.
32:27
There's been other products that have played in this world and and seemingly reached adoption and scale, but uh it's very I I this was this is the same thing as the Sora discourse where everyone was caught up in Sora is is either going to be the most powerful thing ever and we're not going to be able to stop watching it or it's going to be good and so fun and and it's going to be amazing and dominant.
32:55
But no one was counting just like, oh, it might just go away in six months.
32:56
And is the same thing here? >> I don't know.
32:59
I >> I think I was >> saying it was >> going to go away just because I thought it was going to be a tool. >> Yeah.
33:06
And I think I think we benchmarked the market for this like if you if you look at other romantic stuff first said, "Okay, I'm going all in on on adult entertainment."
33:17
I did think it was a potential path for for Grock to get into the single digit billions.
33:22
I probably overestimated overestimated uh >> the market there.
33:32
>> Um but it felt like it felt like a a a >> one of the plays that he had to get back in the game at the time. Yeah.
33:39
Um but again this was last year >> and back then if you could get into the single digit billions you were doing pretty well and then the game obviously you know really shifted to >> yeah it makes sense to wind it down focus on enterprise software.
33:53
That was what was in the SpaceX S1 was the the what what was it $13 trillion market he was going after something like that.
34:01
It was in the it it was a shocking shocking number maybe 20 trillion or something.
34:07
Absolutely huge huge numbers and it makes sense.
34:10
There's a lot of value in the enterprise much less in this controversial topic. Uh well fish.
34:13
audio ran a billboard campaign. We love out of home.
34:21
This one sort of confused people.
34:24
Uh, Heshi Brody says, "$100 for anyone that can explain what this company does without looking it up."
34:30
This is on the New York City subway today and we'll read it to you and you can take a guess. Uh, fish.
34:37
audio says, "We put voice AI on a silent sign. You see the problem."
34:41
Is this one of those jokes though?
34:45
Because >> I think it's a joke.
34:47
I think this is an 11 Labs competitor. We even know.
34:49
But but is this a real ad from a real company or is this a prankster making fun of tech ads?
34:56
Because there was that prankster who put up a bunch of uh billboards like fake billboards.
35:04
Remember like it really makes you think.
35:06
It really makes you think.
35:08
Uh it made me go to the website. >> Okay.
35:11
They make uh text to speech, speech to text, audio separation, voice changers, translation, and they partner with global innovators, John.
35:19
They're working with uh Hey Gen. >> Okay. >> Uh retail. >> Real company. Yeah.
35:24
>> A bunch of games companies.
35:24
It looks like >> Clout Kitchen.
35:27
You ever heard of Clout Kitchen, John? >> No.
35:31
>> You ever been in the kitchen cooking clout? >> Okay.
35:33
So, isn't it like they put voice AI on a silent sign?
35:35
You can't like if you have a billboard, you can't listen to it.
35:39
But their product is audio.
35:42
>> You can't listen to the billboard. That's the problem. >> Yeah, >> right.
35:46
>> I guess I would expect >> your explanation is a problem.
35:49
>> I I would expect the application.
35:51
>> More importantly, we have our first take a picture and it reads it to you.
35:52
That's what I would expect.
35:54
Anyway, uh we do have our next guest in the waiting room.
35:58
Let's bring in Pablo Toré from Pablo Tore founds finds out.
36:00
He's a host, investigative journalist and he's on an absolute terror. What are you doing? How you doing, >> Pablo?
36:06
It's an honor to have you.
36:06
I was telling John before the show started, we podcast for a living.
36:10
I'm not a big uh I don't I don't follow very many sports very much, but every time there's a sports story, I go find your content.
36:19
And when I watch your content, you make me feel like an amateur because you're just so good >> at what you do.
36:23
So good at yapping, but then you're also also an elite investigative journalist.
36:27
And we wanted to take Victory Lap with you.
36:30
Feels like a >> My show is designed Thank you.
36:34
A B my show is designed for people who don't care about sports at all to find it even mildly entertaining and interesting.
36:41
And this is a story that has immersed me in the world of securities law and the funneling of money and billionaires and especially one particular billionaire who used to run a certain company called Microsoft.
36:55
And so all of this I think is actually in your guys' wheelhouse.
36:58
So happy to be on with you. >> Yeah.
37:00
And I remember I remember I watched some of your content around a year ago when you were originally breaking this story >> and then I'm surprised that the story like at least went away out of the out of the public eye for some time.
37:12
I'm sure you were still following it, but take us maybe from back a little bit to first kind of uncovering this and then we'll get all the way to the present. >> Yeah.
37:23
So, this was an investigation that started with a tip.
37:25
And the tip was there's a weird company out in LA called Aspiration.
37:30
It was a tree planting startup, aka a carbon credits company.
37:35
The ESG thing, as you guys may recall, during the pandemic, was a real thing.
37:39
We're going to be good guys, right?
37:41
[laughter] Everyone's going to be a good guy now. And, uh, endors. >> Hey, hey, hey. I love I love trees.
37:48
I love planting trees, too.
37:51
I love what they do to carbon and oxygen reportedly.
37:56
Um, so Aspiration has all these endorsers that are public.
37:58
There's Robert Downey Jr.
38:00
, there's Leonardo DiCaprio, there's Drake, >> there's Cindy Crawford, people who grew up in the 90s deal.
38:08
>> Toronto figures into the story in lots of ways, it turns out.
38:10
But the point is that there are all these A-list stars, Orlando Bloom, I didn't even mention, all these A-list stars that this Good Guy company is paying to tell everybody about what they are. >> Yeah.
38:23
>> And what I get a tip on is, hey, you should look into this company. It's kind of weird.
38:26
Um, they just signed a deal, $300 million to be the Jersey Patch sponsor of Los Angeles Clippers, >> and uh, you know, they have another endorser that you'd be interested in. >> Yeah.
38:37
And when I look into this, it's not immediately obvious like what they're talking about, but when this company goes bankrupt because, and this is going to be in your wheelhouse again, they were going to go public via spa. >> What a time, right? What a time.
38:49
Spaxs ESG, [laughter] we're all going to get rich being good guys and no one really needs to ask many questions.
38:57
Just trust us, it's going to be fine.
38:59
be fine. And what turns out is uh this company is co-founded by two people two big dem donors plastic good guys went to Harvard I mean I resemble the remark I want to say that like this has been also awkward for me Steve Balmer Harvard all
39:14
these guys all of this right pedigree >> all of it >> the thing that's interesting is that when you dig in to how this company fell apart and they go into bankruptcy you of course get to examine some public filings >> and one of their big creditors was a company called KL2 Aspire LLC. >> Mhm. >> Mhm.
39:33
>> And if you're a basketball fan who's aware of the Clippers, you may realize that KL >> Yeah.
39:39
>> sounds like Kawhi Leonard 2 sounds like his jersey number.
39:41
KL2 Aspire sounds like a vehicle you've invented to accept money from aspiration. [laughter] >> Yeah.
39:49
>> And he was owed $7 million outstanding.
39:51
And the weird part was that this dude had no public record of being associated with aspiration personally at all.
40:00
>> And so when you dig in and you begin to ask questions of people who used to work at a company that has now gone into bankruptcy in which the co-founder Joe Sandberg is now, by the way, serving 14 years in federal prison for fraud.
40:09
You get people who are interested in explaining how crazy their life has been.
40:16
>> And what they say and what they provide in the tonnage of all the reporting I did for months, seven months before we came out with part one of the series, one year ago today was documentation that attested to the fact that Kawhi Leonard was paid, according to this agreement, a total of $48 million, 20 in stock, 28 in cash to do nothing. >> Yeah.
40:36
>> For a deal that never got announced.
40:36
And the question was why? >> Yeah.
40:40
And so you follow the threads and you get to, oh wait, there's a massive salary caps or convention scheme in which the richest owner in American sports, Steve Balmer, is trying to use all of his wealth in ways he's not allowed to to get money to a guy that he he needed to take away from the Lakers and the Toronto Raptors in order to make his dream of owning a professional basketball team exactly the dream that he imagined.
41:02
>> So how strict are these rules?
41:04
because I can imagine that's [laughter] a funny way.
41:10
>> Yeah, we we so we covered we covered the whole like the AI talent wars last summer every single day.
41:14
It was crazy and we kept coming back to this idea.
41:18
Part of why the talent wars worked in in many ways and why is there's no salary cap, right?
41:23
So Zuck could just spend exactly as much money as as the players or the talent would accept.
41:28
But in this case, rules are there's actual laws. >> Yeah.
41:32
I I I guess I just mean like if I'm a if I'm a player and I know that the owner runs a hedge fund and I see the benefit of putting some of the money that I earn directly with their hedge fund, it gets me access or oh I know that the owner of the team is friends with the CEO of Nike and I might be able to get a shoe deal.
41:48
Like there's some level of just like doing business above board that's probably acceptable but is this a blurry line?
41:55
Like how how defined is this?
41:57
Like how clear is it to you?
42:00
>> So two things to know.
42:00
One is that this is a cardinal rule of sports in which you raise a very astute point.
42:06
>> If you are LeBron James, it's kind of insane.
42:08
No disrespect intended to the latest person who got bidded up between, you know, Meta and OpenAI and Anthropic or whatever. No disrespect.
42:16
LeBron James is probably thinking to himself, why can't I make $100 million a year?
42:22
>> Yeah, >> this salary is capped.
42:24
>> And if you have problems with that, I get it.
42:26
Sports is a fun mix of socialism and capitalism that is convenient.
42:31
typically for the owners of these teams who are wealthier every time you check the news. >> Yeah.
42:36
>> So point A well taken.
42:36
Point B though is that it's a cardinal rule of sports.
42:41
Meaning that sports is one of the few places where your spending power is not automatically >> supposed to let you just buy whatever you want. >> Right?
42:51
This is how big markets and small markets ostensibly >> get to some level of par.
42:54
This is how and this is debatable.
42:56
Of course, we're getting into more philosophy than I expected, but it's a fair point.
42:59
It gets into the question of like what's actually moral and not.
43:05
>> What Steve Balmer did, given that blurriness around the ethics of it, >> was violate and really um I would say blow past any semblance >> of plausible deniability that hold on, I thought this was something that we could do here.
43:20
And I say that because according to the NBA's investigation and my own, this was not merely a scheme that he ran to funnel money with one company, Aspiration. He did it with four.
43:31
>> He did it with the scoreboard manufacturer, multi-million dollar deal for Kawhi Leonard to do nothing.
43:36
He did it with the insurance company, Locked in Insurance, which was the insurance company on the buildout of the Intuit. >> Mhm.
43:44
>> He apparently did nothing for that.
43:44
And he did it with, and this is a fun one, Boingo Wireless.
43:48
Oh yeah, >> which you may know as the wireless provider >> of Los Angeles Clippers.
43:54
And the thing you say to yourself, do I really need to pay for a boy? >> Yeah.
43:58
I'm like, oh, [laughter] >> is my personal review of the product.
44:03
>> So when when your original story when your original story broke, you got a bunch of push back, even from bunch of people that that are pretty tapped in and knowledgeable.
44:13
And some of that push back was just that there's no way that Balmer would be this dumb, right?
44:19
So this like like this story couldn't possibly be true because who would ever do who would ever like risk it all >> in this way.
44:26
Is that is that somewhat accurate?
44:28
I >> I think that's even the most generous defense.
44:32
I think there are a lot of people who are just being told by the Clippers and Balmer's crisis PR handlers, this weird podcaster is clout chasing.
44:40
None of this makes sense.
44:40
Steve would never do this. He's Steve Bong. >> Yeah.
44:44
The less um I think uh surprising version though which you just articulated is a common one which is you can't be this dumb. >> Yeah.
44:54
>> And and I dare say that one of the things you learn when you investigate financial impropriety in our great country is that that is something that you should never assume. >> Sure.
45:07
Um, there are many phenomenally successful people who because they are desperate to get the thing they can't just buy do things that they think will never get found out.
45:20
>> And the problem in this case across these four companies and across these employees and across my reporting is that he trusted people that he should not have trusted and also he maybe shouldn't have tried to do it in the first place.
45:32
Isn't there one weird trick that would have made this all work?
45:35
Which is just Balmer goes to Boingo Wireless and says, "Yeah, we got this deal.
45:41
Uh, it's expensive, but you you will get some Instagram posts."
45:48
And then Leonard actually follows through on some of the promotions and would have been enough top market for CPM like it could work. >> Yeah.
45:57
If he if he had actually carried out the endorsements, would that have been enough or was that is that still crossing the line?
46:03
>> There's an So, I've debated on my show literally Mark Cuban about this whole story before it became obvious what this was.
46:08
And I appreciate the apology he gave me on Twitter yesterday >> genuinely. >> Cool.
46:12
>> But there's a Shark Tank episode we could do just like pitching ideas for how to circumvent the salary cap that would have worked. >> Yeah.
46:18
[laughter] >> And and I think one of them, by the way, just to play that game briefly, it's like >> telling Kawhi Leonard about crypto. Yeah.
46:26
There are untraceable flows of money that I am told are reliable at this point in the calendar that you could use. Right?
46:34
But what they did was actually almost too clever by half.
46:38
What they really did was they created fake jobs for Kawhi Leonard and fake consulting agreements, consulting fees for the companies.
46:44
So they were trying to create separation, >> okay, >> a familiar term in the investigation of financial propriety, but they're trying to separate where the money was going from A to B to C.
46:54
The problem was that everybody in the course of doing it, people they trusted, left a trail of paper as well as testimony, as well as a lack of rational explanation as to why this existed if you never were to announce the deals.
47:06
And Kawhai, and this is the great comedy that is sports in which salaries are capped, but egos are not.
47:14
Kawi Leonard, >> Kawhi Leonard said to everybody, "I'm not doing a single thing for these off the books payments.
47:19
I am not doing work for them, but I want them."
47:23
And Steve Balmer said, "Got you. We'll make that work."
47:27
And because these deals could not be explained as endorsements because he literally never endorsed them. >> Yeah.
47:35
>> It became this comedy of errors.
47:38
>> That's very, very odd.
47:39
>> All right, let's get to the punishment.
47:41
What What's been your reaction to Kai got a $700,000 fine?
47:44
doesn't feel very significant, at least, you know, compared to the the actual payments and what was received.
47:53
And then, uh, you have a one-year ban, uh, for for Balmer.
47:59
And then, but but what was your reaction to the to the to the verdict from the NBA? >> Yeah.
48:06
Um, this was the biggest punishment for an owner in the history of, I think, American professional sports.
48:11
And the NBA took five first round picks, which is the most you could imagine.
48:17
They took $30 million, which is a rounding error, of course, for Balmer.
48:23
They made him pay the $50 million legal fee to walk to Lipton, the NBA's outside counsel, which is also a rounding error, but still, >> yeah, >> pretty annoying, I would imagine.
48:31
Um, but the real thing that I thought was meaningful was that they banned Steve Balmer from his own building for a year. He loves basketball. He sits courtside.
48:44
He measured literally the toilets at the Intuit dome when he was designing the whole building.
48:47
He was a micromanager who knew everything about every piece of the building.
48:51
And they banned him from being inside his own building which he funded for $2 billion privately to his credit.
48:59
And they also banned his president of business for a year.
49:01
And they banned his GM, his president of basketball for 6 months.
49:04
And so Kawaii got effectively a slap on the wrist.
49:07
And that is wild on the merits.
49:12
>> The reason they did it briefly is because when Balmer now is rattling his saber because he is the former CEO of Microsoft who when presented with antitrust findings by the literal federal government also fought right this is unsurprising if you know the trajectory of his life.
49:28
What he has threatened is litigation and what he has not been able to pursue is arbitration which is interesting because arbitration briefly in the world of sports court which is another TV show we should pilot together.
49:44
[laughter] In sports court in sports court you have an arbitrator assistant arbitrator but that is an arrangement between the player Kawhi Leonard and his union as well as the league. Mhm.
49:55
>> And when Kawhi Leonard got a slap on the wrist, only $700,000, no suspension, no voiding of the contract, no nothing, he agreed to a settlement that was quite favorable.
50:04
And the question was why?
50:06
>> Well, Kawhi Leonard's settlement meant that the arbitration option was legally removed.
50:13
>> So >> now they boxed in Balmer, >> which is again no small thing.
50:16
My reaction was, it's no small thing to be at war with a guy with $145 billion depending on Microsoft stock >> and to say you got one move >> and it's to enter yet more discovery when you know >> he doesn't want to.
50:31
It sounds like Yeah, it sounds like a nightmare situation for Balmer at this point to >> get more of the details out like he kind of I I >> there's more >> there's more.
50:44
So, so your your read on it is he just has to basically accept the consequences and and you know get to watch as a fan for the next year.
50:51
Um but how much does this set back the the Clippers?
50:57
Uh like how impactful, you know, in many ways like he tried to cheat his way, you know, a couple steps forward, but in the process of that set the team back.
51:08
like basically took the team out of probably contention for I want to say a decade.
51:13
Like what what what player would want to go to this organization now knowing that they're not going to get any first round picks >> for the next five years?
51:23
Like it just feels like a really tough situation.
51:27
>> The next Clippers firstrounder that they will be able to select hasn't even sniffed puberty yet.
51:35
[laughter] like we're just in a timeline that is a nuclear winter kind of feeling if you're a Clipper fan.
51:42
But I want to say that at a certain point perhaps things get so bad that sort of things almost need to reset.
51:51
And so actually the thing I'm wondering aloud with you guys is as we contemplate what is Balmer going to do next?
51:57
He has now threatened personally with litigation Adam Silver by name in a letter through his lawyers.
52:02
he is signaling it is wartime and we just gave you some of the stakes around what that would entail practically.
52:08
But historically, if you're going to go to war with your sport and with the commissioner and with other owners, by the way, that's the sort of backstage >> politics of this.
52:19
Adam doesn't do this as a crusade.
52:21
He does this because he's read the room of other billionaires who own these 29 other teams.
52:26
If you're going to go to war with the other 29 ownership groups and the commissioner and the league and journalism and everything, >> typically it stops being as fun for you. >> Yeah. >> Typically.
52:40
>> And so, does Steve Balmer even want to consider perhaps >> selling and taking a punishment that is a humiliation on one level, but also you know what the Lakers just went for, $12. 5 billion.
52:54
What could you make on the back end if money was a thing you cared about? >> Yeah. >> So, >> yeah.
52:59
What how what does this do to the value?
53:01
Like, there's no way you could value the the Clippers at least within 10% of the Lakers, right? >> No.
53:12
Doing back of the envelope math, they're not they're not getting doubledigit.
53:15
They're not getting eight figures.
53:17
Um, you know, one of Sorry, excuse me.
53:20
They're not getting 10 figures.
53:21
Um, what I think is interesting though is they're going to make more than people I think are prepared to intuit it based on the fact that into it is a pun not even intended based on how horrific this franchise has been.
53:39
We're just at a place in sports where the scarcity is so clear that these are detached valuations from revenue.
53:48
You know, again, there are a couple places in this world where that is true.
53:52
You guys cover a lot of them.
53:55
Uh, sports though is on its own crazy ride that is getting divorced from the reality of things. >> I don't know.
54:02
I think sports sports and the private markets are pretty nei [laughter] but uh I take your point. >> No.
54:12
And by and by the way, that's the story of the Lakers and Mark Walter, which we were also investigating.
54:15
My point being that they're not going to get 12 and a half, they're not going to get 10.
54:19
But for Balmer, it's a trade-off between what do you really want?
54:22
You have all of the money. >> What do you want? Do you want to fight?
54:27
Do you want to surrender?
54:27
Do you want to walk away with something like your, you know, ego or dignity? Whatever.
54:32
I don't want to be too over the top here, but like he has a calculation that is not a normal calculation to make because right now it's signaling that he's going to fight. >> Yeah. >> Yeah.
54:43
Fighting fighting seems like a rough going to war with the with with you seems seems pretty rough.
54:48
Um what uh do do you feel like do you feel like this has probably been happening?
54:54
Do you think it's possible that that Balmer knew like felt confident in doing something like this because it was happening at a larger scale?
55:02
And if you actually look at a bunch of other teams, it's it's really much more widespread and and if you start looking at a bunch of these endorsement deals, >> there's there's definitely a a better way to go about, you know, >> you you said you could have a whole show about getting around the salary cap, and I and I I just have to imagine there's so much money involved. there's so much ego.
55:25
If you own a team, you want to you want to compete, you want to win.
55:28
Uh and I I would I would guess that there's a bunch of other owners and GMs and things like that that are kind of a little bit nervous now uh because they may have done something similar or even at a smaller scale.
55:43
>> I think two things are true.
55:43
One is by degree lots of other examples of salary cap circumvention funneling money to players have happened will happen are happening. Mhm.
55:55
>> But the other thing that's true is that in terms of the scale and ambition and the paper trail and the sort of almost just like subplots, nothing is quite like a story in which a guy who doesn't want to do anything, who has zero discernable endorsement value, even when he does something, has leverage over the richest donor in
56:15
sports, who is willing to spare no expense almost literally to funnel money through four different companies, one of whom, by the way, is a public company, Dtactronics, that is now acknowledged in yesterday the earnings call that their CFO according to him is now dealing with an SEC investigation. So there there's
56:28
So there there's also a civil suit happening in LA court against E Balmer personally by 11 aspiration investors about fraud.
56:35
My point is there are lots of other examples about how to circumvent the cap that are I'm sure getting no attention and maybe we'll get to some of them.
56:45
But in terms of what this one is, the reason it's the biggest punishment is because they found stuff that is singular in terms of all of the webs that it entails.
56:58
And that is >> well, I just going, >> but now the the the people that the person that is most incentivized to uncover all these other salary cap circumvention, you know, potential deals that have been happening is actually Steve Balmer.
57:14
Because you imagine in his war the only thing that he he's he's like he's got his back against the wall.
57:20
He's sitting there thinking the only thing that I could possibly do to kind of clear my reputation is make it come out that >> you think you think he's going to start feeding Pablo tips. >> Yeah.
57:30
>> Be your number one tipster.
57:30
[laughter] >> I I I would gladly welcome >> back checkable information from any source.
57:36
I also welcome the idea that maybe the biggest consequence of my reporting is that Steve Balmer becomes himself a podcaster >> and that would be something that I welcome as well. >> That'd be fantastic.
57:49
>> Final job once you have everything.
57:51
>> Well, he's got a lot to learn.
57:52
>> It's only you and the mic.
57:52
[laughter] >> I will say based on what I've learned from your show, that trajectory is also an increasingly common thing for a billionaire to do.
58:00
So perhaps we're not so different, you and I. >> Yeah.
58:03
>> Well, thank you so much for coming on the show. This was fantastic.
58:05
the show. This was fantastic. honor to have you and um congratulations I'm guess congrat I mean it's it's a it's it's a strange thing to say congratulations for a story that is ultimately quite sad and and unfortunate
58:19
but um yeah credit credit you on display here >> look as always um comedy is when bad things happen to other people um so there are Laker fans who think this is the funniest story in the world I do shed a tear >> for the Clippers fans who are like, >> "Yeah, >> really >> again. >> This again,
58:40
>> This again, >> another owner.
58:41
Another owner that we might not." Anyway, >> yeah. >> Uh, last question.
58:45
Is it possible that he sells the Clippers and overpays for another team >> just to get right back in the game?
58:54
>> Look, what he wanted from the beginning, what he was interested in, of course, because of Microsoft and geography, was a Seattle team back again.
58:59
Now, I will advise this just as a podcaster to a potential future podcaster.
59:05
Um, I am told that one way to be approved to buy a new team is to not go to war with the 29 other ownership groups that will need to approve that.
59:15
>> So, just a consideration as he continues to examine contingencies. >> That makes sense.
59:20
Well, thank you so much for coming on the show.
59:22
>> Yeah, thank you so much. This was a week.
59:24
We really appreciate >> and congratulations on on your overall overall >> run everything. >> It's fantastic.
59:29
>> We'll talk to you soon. Anytime. Thank you. Have a good one. Goodbye.
59:31
Let me tell you about the New York Stock Exchange.
59:34
Want to change the world?
59:36
Raise capital at the New York Stock Exchange.
59:38
Up next, we have Moit Aron from Sipin coming out of stealth live. How are you? Welcome to the show.
59:48
>> Hey, thank you for having me here.
59:50
>> Thanks so much for hopping on.
59:50
Um, I would love to start with a little bit of your background.
59:55
This is uh your latest company, but not your first.
59:57
Can you give us the uh the career highlights thus far? >> Yeah, absolutely.
1:00:04
Uh you know, my uh my last company was called Cohesidity U and the one before that was called Nutanix.
1:00:13
So I'm the founder of two previous companies.
1:00:16
Both of them fortunately are doing north of a billion dollars in AR.
1:00:19
Uh the first company is already public. >> Yeah.
1:00:24
>> And and this one is um uh the Kohhis is yet to go public.
1:00:29
>> So >> So you had a tough time [laughter] raising new company, I'm assuming. >> Sure. It was quick.
1:00:35
>> You really had to, you know, do all the laps up and down Sandill, scrape scrape together pennies to get this done, I'm sure.
1:00:41
[laughter] >> No, the round came together, I'm sure, pretty smoothly.
1:00:44
But uh what were you laying out?
1:00:46
I imagine it was like I'm I've I've gotten to a billion of ARR twice.
1:00:50
Here's the plan to get to a billion of ARR really really quickly. What What is the plan? What was the pitch? >> Yeah.
1:00:57
The pitch was that look in my first company uh I brought together or converged infrastructure. >> Yeah.
1:01:04
>> Uh in the second company I converged data. >> Mhm.
1:01:07
>> I'm here to converge context for go to market team. >> Okay.
1:01:13
[applause] Somebody had to do it. So >> yeah, to do it.
1:01:17
>> Uh I imagine with the connections that you've built through your previous companies, uh you can you can start up market with your go to market. Is that the plan?
1:01:27
Are you are you thinking uh work with bigger companies, bigger deals, or do you still want to uh scratch the the bottoms up itch, the the the the smaller, more agile, early adopter crowd?
1:01:41
What's the go to market here?
1:01:43
you know, I I'm very passionate about serving customers.
1:01:46
Uh I think the problem is all over the stack, both small companies and big ones.
1:01:51
Um um my sweet point is obvious obviously going to be the u the big enterprise.
1:01:56
U but I don't want to leave the smaller companies behind.
1:02:00
So it's going to be all of them.
1:02:03
>> How do you think about the the actual value that the product will bring to these teams?
1:02:08
these teams? because uh there's been there was a little mini boom cycle for like AI SDRs which feels adjacent and a lot of people were saying hey it's way too early to let an AI any AI even if you're wrapping a frontier model I don't
1:02:25
want it emailing my customers who I might have been golfing with this weekend and you don't have that context and so I don't want you embarrassing me but it feels like that's where we're going how do you see the the the process developing over time. >> Yeah. The uh I focus on the problem and >> Yeah.
1:02:39
The uh I focus on the problem and the problem is uh something that we describe as the context gap. >> Mhm.
1:02:47
>> And I've seen this problem, you know, running my previous two companies uh even preai, right?
1:02:51
We execute our go to market operations without enough context. >> Sure.
1:02:58
>> The context is all over the place.
1:02:58
It's sitting in multiple tools, people's heads, and it's not even accurate.
1:03:03
It doesn't reflect reality. >> Mhm.
1:03:07
And we are trying to make decisions without it. Right?
1:03:09
And the way we try to do stuff is we maybe do a lot of meetings.
1:03:14
>> 80% of the time spent in meetings is to gather the context, not change the news, right?
1:03:19
Gather the news, not change the news.
1:03:20
And so this is a unique opportunity especially with AI to actually solve that problem. >> So how much Yeah.
1:03:26
I because I feel like if if if you were able to create an eight an eighth day in the week and sit down the whole team and say everyone dump everything into some voice notes and some some shared text files.
1:03:39
Uh the the frontier models can summarize and and and put all of that together, but it's really a data acquisition problem.
1:03:50
And so uh how much of this is about enabling teams to do change management the way things work versus autonomously going and gathering data both what's the actual solution >> yeah the actual solution is all of that so uh automatically gathering the data
1:04:06
making sure that humans are in the loop so that AI is not just hallucinating and you know uh putting garbage uh it is also in helping people in real time uh real time coaching and stuff uh where they can ask any question or know anything they want from the context. So
1:04:21
So the solution includes all of that. >> Well, it's Oh, sorry. J please.
1:04:27
>> Key lesson from the first company that you brought to the second company and then key lesson from the second company that you're bringing to this company.
1:04:34
>> Uh key lesson from the actually both my companies I would say is it's all about the people.
1:04:39
>> Uh it's all about the people inside the company.
1:04:41
It's all about the the people outside the company which is our customers.
1:04:45
And so you focus on that and the right things will happen. bearish on AI. [laughter] Just kidding.
1:04:50
Uh, it's your third company, but it's your first gong hit here on TVPN. How much did you raise?
1:04:58
>> I think it will be the last.
1:05:00
>> How much did you raise? >> How much did I raise? $44 million.
1:05:07
>> Well, thank you so much for coming on the show and breaking it down for us.
1:05:10
>> Let's do it again soon. Great.
1:05:11
>> Let's do it again soon.
1:05:11
I'm sure there'll be many milestones. We'll talk to you soon. Cheers.
1:05:14
Have a good rest of your day.
1:05:16
Let me tell everyone about Figma. Agents meet the canvas.
1:05:17
Your AI agents can now create and modify your Figma files with design system context.
1:05:25
You don't get a lot of CEOs, a lot of lot of successful founders that say, you know what, it's really it's it's not about the people at all.
1:05:30
It's it's actually just the code. It's just the AI. It's just the AI agents. That's all that matters. No.
1:05:35
Uh it of course is the people.
1:05:37
And we have a great person next.
1:05:39
>> Often the best advice is the most simple straightforward. >> Yeah, it's true.
1:05:43
Well, uh, we have Ashai from Pocket.
1:05:45
He has sold over 200,000 of these devices.
1:05:49
Show you what they look like on the screen.
1:05:52
[music] Welcome to the show, Ashai. How you doing? >> Hey, John. Hey, Jordy. How are you guys doing?
1:05:58
Uh, it's great to be in TBPN.
1:06:00
>> I'm glad to have you here.
1:06:02
>> No, we we we started hearing and learn learning about your company probably a year ago at this point.
1:06:06
We're like, what's [snorts] this company that's just selling a lot of hardware somewhat somewhat under the radar?
1:06:13
I'm sure it doesn't feel like that to you, but you guys haven't really chased the spotlight as much as other companies that sell hardware that haven't sold any hardware. >> Yeah.
1:06:22
I mean, the the the the phrase in Silicon Valley, hardware is hard. Don't go after it.
1:06:26
It feels like AI has unlocked new opportunities both in there's demand for new devices and then there's also AI that you can ask to help you procure things and help you design things.
1:06:39
But what has the road to actually building the first device then scaling to 200,000 devices been like for you? >> Yeah.
1:06:46
So interestingly we we actually built an app before we actually built the hardware. Yeah.
1:06:50
Uh and the simple thing was that hey it's the most obvious thing to do.
1:06:55
it's the most easiest thing to scale, so let's just go build an app.
1:06:59
And we built a meeting notetaker app.
1:07:01
Um, and we we released it to a bunch of people and uh there was not a lot of response.
1:07:06
And uh being surprised, we we thought to ourselves, hey, what about that device that we were all thinking about that we should go do?
1:07:14
Uh that'll be like a necklace.
1:07:17
And uh that's kind of what we did u with with open source work at OMI.
1:07:22
uh we we went and did bunch of uh Kickstarters and uh we we printed a bunch of these devices and gave it to people and uh we we actually saw that people were were using the device 10 times more than the app.
1:07:36
Uh there were a couple of other challenges regarding uh being being always on and being a variable.
1:07:44
So when I started pocket my first immediate thought was this should not be always on.
1:07:49
Uh people are still not comfortable.
1:07:52
there's still dialogue that needs to happen before obvious on wearable becomes a thing and we said let's go build something that's a wearable for a phone so there so the people on Wall Street are happy to use it and it doesn't augment their style of their suits and coats that they wear and that's kind of where Pocket came around. >> That's very cool.
1:08:13
So yeah, walk me through the current product specs, price point, uh what the favorite features are, sort of like how you think about where you are now and then where you want to go.
1:08:25
>> So currently this this is Pocket and uh this retails for $129.
1:08:31
Uh the hardware comes with a premium subscription.
1:08:33
So you kind of get to do unlimited summaries, unlimited transcripts.
1:08:39
uh you get three messages you can ask about your conversations per day.
1:08:43
Uh if you would want more like speaker tagging, advanced speaker recognition, uh all these kinds of things and advanced uh AI models for reasoning uh and summary agents that can make better details out of your summaries and have access to 300 plus professionally sourced templates like soap notes for doctors, case notes for lawyers, and CRM entries for salespeople.
1:09:07
uh you can get on our pro plan for $20 a month or $200 per year and uh that's kind of the the whole product basically >> for health.
1:09:16
So, so, so the thing that stands out to me is like you you launch the initial app, you're not seeing the kind of growth that that you wanted to see.
1:09:24
And so then you're thinking, okay, we should la we the solution is to launch a hardware device.
1:09:29
>> Let's turn up to try to difficulty level, which normally normally you'd think like, okay, the product's not working.
1:09:34
Add a hardware thing that functionally does a lot of the same stuff.
1:09:38
It's probably not necessarily gonna I don't know.
1:09:41
My my intuition would be that it wouldn't solve the problem, but I I keep coming like what is why why does it make sense for this to be a separate device from the phone?
1:09:52
Because I'm sure you know if you met with investors, a lot of investors would be like I have a device that's a hardware device. It has battery. It has a microphone. It has a screen. It has all these things.
1:10:00
So why are your 200,000 plus customers saying like no, I want this.
1:10:05
I want this separate device because like clearly the customers are right. >> Yeah, absolutely. Uh, it's 300,000 now.
1:10:13
And, [laughter] uh, we got to hit back up. >> Founder, >> right?
1:10:22
U, so, so when we actually went and asked these people like, hey, why are you using this 10 times more more than more than the app that we had?
1:10:28
Uh we actually had one of the user who was actually using a different uh phone he just bought for recording conversations and he also had a battery attached to the phone on top.
1:10:42
>> So that's kind of what Pocket ended up being a product by just watching how people use these recording apps on the phone.
1:10:48
Uh there are a lot of limiting factors for recording on the phone.
1:10:51
Um a you cannot record your own Zoom meetings in Google Meet if you're already on the phone.
1:10:57
you cannot start right another recording.
1:11:01
>> Uh so there are a lot of people who have issue with that and there are people who get incoming phone calls all the time.
1:11:07
So if you're recording something and you get an incoming phone call it it it interrupts the recording and if you take the phone your recording is stopping essentially.
1:11:14
Uh apart from all of these technical reasons, the the real reason is actually that the people who take 10 15 meetings a day, these doctors, consultants who are like, you know, patient goes out, another patient comes in.
1:11:27
Uh the the client goes out, another client comes in and there's 10 to 15 backtoback meetings.
1:11:31
uh they they feel awkward in the first 8 seconds to take their phone and start a recording by unlocking their iPhone and setting up the recording and hoping so that the operating system doesn't kill background running apps for a long time.
1:11:44
running apps for a long time. So people just felt that you know this is much more reliable thing a quick access thing that that can immediately start recording and I don't need to feel awkward on placing it on a desk u so people don't take take their attention oh why is this guy looking at his phone
1:12:03
we just met >> so that's kind of what we came around that you know convenience beats intentions so if your users have intention to record and you don't make it convenient enough they will not record but if If you actually make it convenient enough to record and they want to record even a little bit, they'll 100% record. >> Interesting. >> Interesting.
1:12:23
Should more people be taking notes >> with and and recording their conversations? >> Oh, yeah.
1:12:29
>> Is I haven't recorded a conversation in years? >> Yeah.
1:12:34
>> Um, but what am I what am I missing? >> Yeah, absolutely.
1:12:37
I think like there >> I guess I've record recorded hundreds of hours of conversations through the show.
1:12:43
So uh you have to >> we have people recording you know tens of conversations almost daily uh and and most of these people are actually field workers uh you know consultants salespeople real estate agents were showing people's you know homes that that are noting down requirements uh places where note-taking is actually you know useful.
1:13:05
Uh so for example, Door Dash is one of our customers.
1:13:08
They they use Pocket uh to go to restaurants for for the salespeople to record their pitches and then come back to their HQ and sync with everybody else in in a knowledge base of hey how did my pitch go with with this vendor with with that restaurant owner.
1:13:22
So I think like there's there's definitely like a use case for field note takingaking and that's kind of where Pocket comes and sits in.
1:13:30
Uh but but yeah, I think like in a broader sense uh Your AI needs a lot of context about you to help you. >> Yeah.
1:13:40
>> And every single time you essentially use pocket, you can connect your cloud MCP or OpenAI's API and your your chat GPD already knows about all your conversations.
1:13:51
>> Uh and and you can you can pretty much manipulate all your transcripts and and and and and understand how people are talking to you and and all these cues.
1:14:00
uh for example if you go to you know VC meeting uh the VC says 100 other things like it actually sounds so interested like I felt they almost invested in the startup but they never said the word invest. >> Yeah.
1:14:13
[laughter] >> So the found going back thinking oh I have a term sheet or something but the truth was it was just a high. >> Yeah.
1:14:20
No no I've I've seen that play out where someone comes away and they say oh yeah this VC said they're going to invest.
1:14:26
They tell that to another VC. They call each other. I didn't say that.
1:14:31
uh and it becomes a >> who said are you are you broadly bullish on new AI hardware because I think one of the reasons that people have been generally bearish is you've had you know you had the rabbit you've had friend you've had a bunch of these shots and
1:14:44
attempts that humane so there's been a there's been a handful that that haven't got the level of traction that you have had but given your experience so far do you think there's >> a lot more devices to build that are that are AI native um in in the real world I think so. And I think like you know
1:15:02
And I think like you know rabbit and humane died so you know pocket could live and we learned a lot from all these other devices that essentially died and they primarily died because they were trying to replace the whole smartphone game.
1:15:16
>> Doing too much for sure.
1:15:18
>> Too much too much is stuff. Yeah. >> Yeah. Exactly.
1:15:20
>> Yeah. Exactly. and and they were actually just going uh against going towards real world work you know the way that the workflows work for real people you know they they don't want all the other things like ordering you know over eats from from from AI all they want is
1:15:38
like that AI to fit in their existing workflows and what we did was like we we we looked at a lot of people >> had a lot of meetings and and we didn't go into things like personal note-taking for example All these others are like you know dead basically. Uh whoever went and did
1:15:53
Uh whoever went and did anything other than note takingaking and note-taking was essentially a product market fit.
1:15:59
It was almost like fitness for for Apple Watch kind of product market fit and everything else was basically like people didn't want them. >> Yeah. Yeah.
1:16:07
I when I saw Rabbit R1 I never got a chance to actually use it but I was thinking like it's such it's such a beautiful design.
1:16:14
And I thought it would be perfect for kids where if they're just going around taking some photos and learning a little bit about the world.
1:16:21
And I actually wound up buying a different product that's a circular device with a screen and a camera on it and it has little games where the kid can go and say it'll say like find something flat, find something round, and they'll go and find it.
1:16:32
And then uh that other company, Sticker Box, that we've had on the show, is doing really well in what I would call AI hardware, but you push the button, you say a prompt, and then it prints a cartoon for the kid, and it makes a little sticker.
1:16:46
And it's these like very focused things that have it's just do you want this at this price? It'll do this for you. It makes a lot of sense.
1:16:53
Um what are you uh doing or what do you need to do on the regulatory side?
1:16:58
because I imagine if it has a radio in it needs FCC approval.
1:17:00
But uh you mentioned the medical discipline and uh HIPPA does that is that relevant to you?
1:17:08
How how have you solved that problem? Yeah.
1:17:11
So, we're HIPPA compliant and we essentially have HIPPA compliant servers that we we run HIPPA workloads on specifically for medical use cases and uh and yeah, like from from day one, we we thought about being HIPPA compliant and and so to compliant for enterprise as well.
1:17:28
So, FCC is definitely like if you're if you're making a device with with Bluetooth radio, you pretty much need FCC.
1:17:34
So, from from day one, u you would need that as well.
1:17:37
So yeah, in terms of like uh you know consent is what comes up you know most of the time for for these kind of uh recording uh you know devices and uh what we did with pocket was we said we would like to bring the risk level down to almost a user action.
1:17:57
So everything happens with with a user action.
1:17:59
So you're supposed to like tap a button to start and stop recording.
1:18:04
uh although it might be a friction step to to to this but we would love for for the user action to be the risk level where where we are for pocket.
1:18:14
So uh everything starts with the consent of the user at least it starts with one one party uh one state consent um and then you know it's on the user to basically ask for you know other people's consent and and surprisingly not that people actually are very open when you actually ask them to to record. >> Sure.
1:18:33
and and they're actually happy that you would share the transcript and the meeting notes of the recording. >> Sure. Sure.
1:18:39
>> Uh do you think you'll ever launch something like a co-worker agent or or assistant functionality?
1:18:46
Is is that cuz because I imagine if you can give your device a bunch of contexts, hey, I want to do this thing.
1:18:51
Eventually, it could get to the point where you can send off tasks in the background um that the user can check on maybe on their mobile device. >> Yeah.
1:19:00
So, so that's actually our own road map in the next three months that we would actually like to build uh agents that would actually take context directly from from your conversations and actually act on those uh make make reports, make docs, PPTs, presentations, slides, all kinds of things from from your meetings.
1:19:19
So this this is like increasingly becoming like the regular workflow from all kinds of consultant sales uh pitches that people take the conversation and make slides out of them >> and and come back for the next meeting.
1:19:32
So we would love to make an interface for that.
1:19:34
So in the next few months, >> how do you think about picking different AI models for various pieces of the of the workflow?
1:19:42
Because I can imagine that uh you're you're selling something pretty valuable.
1:19:47
So cost might not be the the most important thing.
1:19:50
At the same time, you can just wait and things get cheaper often.
1:19:54
Switching from one model to another might inject like a different flavor or vibe sometimes.
1:19:59
So how have you thought about this?
1:20:02
Are you fine-tuning your own models using open source, bouncing back and forth between whatever frontier models are available?
1:20:10
What's your decision criteria right now?
1:20:13
So, so surprisingly we we we have the highest mar margins in in subscription industry for for AI.
1:20:17
I can imagine roughly 75% margin um in in software subscriptions essentially.
1:20:23
U so >> we we do this uh with with different kinds of routing.
1:20:30
Um >> oh yeah so we we have our own version of open router uh where we essentially route to different kinds of providers and models. Okay.
1:20:41
Uh we use a a section of open source for transcription. >> Sure.
1:20:46
>> Uh because people love to choose uh what model they would like to use for summarization. >> Oh, they do? Okay. Yeah. >> Yeah.
1:20:53
So, so, so transcription is pretty much left to us.
1:20:55
So, >> so we we we do it on ourselves.
1:20:57
So, that's kind of why our our margins are super high because we we get to use whatever we like.
1:21:03
Um and and we use our own models uh that are fine-tuned on top of OpenAI's whisper.
1:21:08
of OpenAI's whisper. uh because a lot of this online models were trained over YouTube data sets like walkab uh which are extremely clean data sets um and work really well for zoom meetings and meetings of online notetakers but when you come for offline there's like a
1:21:26
train going next to you there's there's bunch of things happening so they they actually go go crazy when when there's like an offline recording so we need to fine-tune for them to work for you know offline recording Essentially, >> that makes a lot of sense. Uh, thank you
1:21:41
Uh, thank you so much for coming on the show.
1:21:44
I forgot that I am a a featured customer on the Zapier blog from 2015 because I developed a workflow that would record all of our conference calls, send them to a human transcription service, and then upload the transcription uh into Google Drive.
1:22:06
And then you could search by keyword just using Google Drive search.
1:22:10
>> Uh, no, nowhere near the power level now, but >> interestingly, Y Combinator actually currently uses Pocket to record all their interviews. >> Oh, really? >> Cool.
1:22:21
>> And send them to their online systems through our APIs. >> There you go. >> So, yeah. >> Very cool. >> Something.
1:22:27
>> Well, congratulations on the progress.
1:22:29
Thank you so much for coming on and breaking it down. >> Love the approach.
1:22:31
>> Have a great rest of your day. We'll talk to you soon. Cheers.
1:22:34
>> Let me tell everyone about Railway.
1:22:36
Railway is the all-in-one intelligent cloud provider.
1:22:38
Use your favorite agent to deploy web app servers, databases, and more.
1:22:41
While Railway automatically takes care of scaling, monitoring, and security. >> Who we got next?
1:22:46
>> Our next guest is Hari from Aura.
1:22:46
He's the founder and CEO with a great update for us. How are you doing? Welcome to the show.
1:22:56
>> Hey guys, good to see you.
1:22:56
uh since it's your first time on the show would love a uh an overview of the company and yourself and then we can go into the actual news. >> Sure.
1:23:06
Um the the company Aura were uh in the consumer security, consumer safety uh space.
1:23:12
So basically we thought about um all the various different ways that a family could get uh impacted by using services online scam, spam, um transaction fraud, >> all the way from that all the way to making sure that um if you have teenage kids for example, if they're using devices too much, we have ability to kind of figure out u how best to sort of guide and navigate them around digital detox, etc.
1:23:35
Um so yeah, so that's that's the core of the product and uh I've been an entrepreneur my whole life.
1:23:40
This is my third third company now. So >> awesome. How big is the problem?
1:23:44
Because now that I'm thinking about it, I mean, we were just at uh Crowd Strikes Falcon on Tuesday and uh we heard that fishing open rates, fishing email open rates have jumped from 12% to something like 60% in the AI era.
1:23:57
And when I actually go back to the number of cyber security scams, I have just as many examples of uh friends who thought that their grandmother was asking for gift cards as I do in a corporate setting.
1:24:11
And so how how prevalent is this?
1:24:13
What are the what's the data that you monitor to understand the scope of the problem?
1:24:19
Look, I I I think that over the years, both the breadth of the problem and sort of the number of people it impacts has gotten like astronomically uh bigger basically where uh certain scams used to be, I don't know, a few thousand worth of scams.
1:24:32
Now we see that's gone up to like $25,000 per scam on average.
1:24:33
uh the incidence rate is quite high because now that there are a lot more AI based tools, the bad actors also use the same tools to do deep fakes to do um uh ID theft with a lot more data out on the dark web that ends up becoming a precursor to basically being able to use that to uh to uh uh do nefarious things to families.
1:24:56
And I think the the other big thing is um if you thought about kids where when I was growing up we weren't native to phones and native to um uh online services.
1:25:05
My kids I mean they are digital natives.
1:25:08
So they basically you know are on their devices all day long which means uh the incidence of them putting information out there is just much higher and so as a result there's just more and more and more of this happening which is uh pretty scary.
1:25:22
>> Uh talk about the recent hack leak.
1:25:25
there was some ID verification service that uh had a breach and I think uh from what I could see it seemed like >> half the people in the US were impacted something like that with with their now you know image of their ID out on the
1:25:39
web um but break down the the breach and then and then what people should do >> yeah look I think uh you know starting with even a couple years ago there's the NPD breach the national uh the the the national level of social security number etc. Now there's a lot more ID uh
1:25:52
Now there's a lot more ID uh digital image theft etc.
1:25:54
I think that the key thing I would say is the incidence of these is is so high now that when people hear about it you just start kind of immunized like you're like okay well whatever like you know I'm sure my data is already out there is what's going through people's minds.
1:26:06
The the trouble with it is this is just an early tell for the next set of things that are going to start happening.
1:26:12
that are going to start happening. So for example at Aura we have a fully trained foundation model that is looking at sequences of stuff right and you know hey like what's the first thing that happens and what's the second thing that happens what's the third thing that happens and how can you sort of um uh analyze risk with the with systems and
1:26:29
we see that uh with uh data breaches and information getting out in the dark web is a pretty early and good tell if it's happened multiple times uh you see that the incidence of actual identity theft uh is bad at the back end so the the the very tangible things I'd say, you know, monitor your credit, make sure that you're looking at your credit card n numbers and statements. Don't wait for
1:26:46
Don't wait for the whole statement to show up.
1:26:48
If you start seeing uh odd things on physical mail, like, you know, keep an eye out for that as well.
1:26:52
We've had even, you know, weird situations where people have stolen information >> um and you know, it it gets sold to folks uh for physical crimes as well.
1:27:02
So, we see a lot more intersection of physical and digital uh happening a lot more these days as well. So, >> interesting.
1:27:07
What what is the uh uh what is the industry side of the business look like?
1:27:13
Like I imagine that as you're protecting homes, schools, individuals, identity theft, there there's the flip side, which is if I'm a business, I don't want someone purchasing with a stolen identity.
1:27:26
And having a clear communication line between those two parties probably makes sense.
1:27:31
But what does that actually look like in practice?
1:27:35
practice? I mean that's that's a great question right because if you thought if you thought about your own life like as or or me the surfaces I spend time on are my home my workplace and a parent as a parent my my kids school so there's a lot of data that kind of goes across uh that stack and even more than just you
1:27:53
know are you using a nefarious card for you know for um purchasing something from me even more basic than that is a lot of the enterprise scams like the the the ones you were talking about even with Falcon for example many of those origin on the consumer side and many of the breaches that happen on the enterprise side is social engineering attacks anymore, right? So, it's
1:28:10
So, it's basically going through people who are employees and then kind of using that to get into the enterprise.
1:28:15
So, on both those pieces, I think getting the uh consumer to be more aware of these issues ends up becoming pretty critical.
1:28:25
we see a lot more overlap now where you know we we do sell our product through large enterprises as well and so we talked to a lot of sisters of big companies and the the the feedback we get is look the lines are very blurred between what's at home what's at work um
1:28:38
and so even even besides sort of the uh loss prevention piece of it which is hey you use a fake fake credit card uh to to buy product it's even more imminent that they want to make sure their enterprise are kept safe and sometimes uh you know keeping the employees safe is the best way to do that. Yeah. Yeah. Uh, how much Yeah. Yeah.
1:28:53
Uh, how much time do you spend working on streamlining onboarding?
1:28:59
When I think about these security products, we've talked to the CEO of One Password, fantastic product, but uh onboarding can be a uh like a stumbling block for people that you know want to secure their life, secure their identity, but the actual integration of getting deep into all the accounts, collecting everything, making sure everything's set up properly, that can be a hurdle.
1:29:20
I imagine you have low churn but the conversion to I heard about this to actually I'm I'm using it properly uh has got to be a hurdle.
1:29:31
How much how how much are you focused on that?
1:29:34
>> We spend a lot of time on that.
1:29:34
Time to value is a huge uh metric for us which is you know how how quickly can you get the customer to something of value basically. But I'll say two things.
1:29:42
One is on the way that the industry works today.
1:29:47
Um, if you thought about a product you bought, you know, for for either of you guys, right, it's going to look the same.
1:29:53
Like you buy uh something to kind of keep your identity safe or you uh buy something to keep your passwords locked up or even something for kids, for example.
1:30:00
The the solution doesn't really understand or know you.
1:30:04
It knows the problem, right?
1:30:04
And so one of the things we have done which is very different is we're probably the only product that puts the actual family at the center.
1:30:12
the center. So there's actually a large sort of family graph that maps out all of the relationships for people who are you you know who are the people that you sort of you know connect with and then all of the reasoning and the inference systems that we run on top of that have
1:30:25
good ability to have a view of you end to end which means that when you go into onboarding we can do a better job first getting you on board and then as we're watching and observing we want to make sure that uh that um you know you it's it's very personalized for you and your digital footprint. We have a a massive gong here. Any any
1:30:41
We have a a massive gong here.
1:30:41
Any any recent revenue milestones [laughter] or growth growth milestones you could share that we could uh use to hit this gong for you? >> Oh yeah, sure.
1:30:50
We uh we hit we're about 340 some odd million of AR. So >> thank you. Thank you. Not too shabby.
1:31:04
>> And thank you for coming on. >> Yeah.
1:31:08
Great to see you on the show.
1:31:08
Have a great rest of your day. We'll talk to you soon. >> Thank you, guys.
1:31:12
>> Let me tell you about public. com.
1:31:14
Investing for those that take it seriously.
1:31:15
We got stocks, options, bonds, crypto, treasuries, and more with great customer service.
1:31:18
Uh GPT6 Astra has landed. The blog post is up.
1:31:21
Uh you can go check it out on openai. com. Uh >> to the stars. >> To the stars.
1:31:29
And the headline, I believe, uh good performance on a bunch of things, but the RKGI 3 number is crazy. The score is 99. 9%.
1:31:38
So it feels like they just beat that. They just beat RKGI3.
1:31:45
So that's the video games, the ones that Tyler was briefly >> globally ranked. >> Globally ranked.
1:31:50
Uh you're out of a job, Tyler. Say say goodbye.
1:31:53
Hang up your Arc AGI V3 hat. Uh don't worry.
1:31:56
The team over at Arc AGI is working on V4.
1:32:01
They're we're gonna move the back we're gonna move the goalpost soon.
1:32:03
We're gonna move the goalpost soon.
1:32:04
But uh this is very impressive.
1:32:06
If you've played around with Arc AGI V3, uh it it it's it requires Yeah.
1:32:11
like some real creative thinking to actually learn how the games work, how to be efficient.
1:32:16
Um what else is sticking out?
1:32:18
Has anyone been able to to monitor the timeline at all? >> Very chaotic launch.
1:32:23
So, uh I I think it's hard to get a real reaction yet.
1:32:28
Tyler, what are you seeing?
1:32:30
>> Yeah, I mean there's still no actual post on Open Eyes like X account.
1:32:31
It's just the blog post that's now live. >> Okay.
1:32:35
Um, but I mean there's a bunch of benchmarks in there that are >> Yeah.
1:32:39
>> Terminal bench science.
1:32:39
We're going over into the world of science. GPT6 Astra scored 64.
1:32:43
6% on reasoning effort max cost 26 26 bucks.
1:32:50
Uh, I'm sure there will be a lot more.
1:32:52
Also, the uh exploit exploit gym honeypot lower is the better. GBT6 Astra 0%.
1:32:59
So, good performance there.
1:32:59
Uh, the world's best computer use model.
1:33:02
I'm putting that to the test ASAP.
1:33:04
Let's see if it can onev one me on Rust because if it's good at using a computer, should be able to no scope, right? That's the bar.
1:33:14
That's where my goalposts are.
1:33:14
Um, agents last exam, good performance.
1:33:17
So, you can go check it all out.
1:33:19
I'm sure >> is going to be on Bloomberg just a few minutes over with our with our buddy Ed Lllo. So, be fun.
1:33:26
Uh so we'll be digging into this and we have some special guests lined up to talk more once we've been able to digest, take it for a spin and have a lot a lot of fun with it.
1:33:34
Uh we have our next guest joining in just a second.
1:33:39
But first, let me tell you about Shopify.
1:33:42
Shopify is the commerce platform that grows with your business and lets you sell in seconds online, in store, on mobile, on social, on marketplaces, and now with AI agents. Um very uh exciting.
1:33:52
We have two guests joining at the same time.
1:33:55
Matt Caldwell and Jordy Lyser from the Minnesota Timberwolves and Lynx and Jump.
1:34:00
Uh they're both CEOs and we're going to have them both on at the same time.
1:34:04
Uh let's we need to check in on uh Ed Zitron, see what his reaction to Astra is.
1:34:13
Is it any better or is it just more of the same junk?
1:34:16
Because [laughter] he has he has uh he has been a bear for a long time.
1:34:21
Uh he's coming up on coming up on the three-year anniversary of the bear posting.
1:34:26
Uh Dan Lou has collected all of Ed Zitron's position.
1:34:29
Uh >> you think he has sort of a gamblers mindset?
1:34:32
Like 80% of gamblers quit right before they're about to hit it big. >> Yeah.
1:34:38
I I I think it's just uh it's an entertaining sticktick. It gets views.
1:34:43
Kevin Roose, uh, formerly of the New York Times, now independent hardfork co-host, uh, who's writing the AGI Chronicles on sale October 10th.
1:34:50
Uh, he says, "Anyone who has interviewed Ed Zitran or had him on their show in the name of AI skepticism has made their audience dumber and less prepared for what is actually coming."
1:35:03
Wow, that's a uh, crazy thing.
1:35:06
Eric Newcomer chimes in and says, "It's his audience. They follow him around."
1:35:10
Which is interesting because I believe Eric's best episode of all time was with Ed Zitrin and he was like the Ed Zitran fan base like stuck around for a couple more episodes and was like I don't like this guy at all.
1:35:23
I came here for [laughter] like I don't I don't want reasonable AI and and Eric of course is >> I don't want to see both. Eric skeptical.
1:35:34
Sometimes he's optimistic, but he uses these and he doesn't he doesn't believe that it's like a >> RSC says when you ask Ed Zitro where he keeps writing all his predictions are when all his predictions are wrong.
1:35:44
And there's a quote from Dion Waiters.
1:35:46
I'd rather go 0 for 30 than zero for nine.
1:35:52
Because you go zero for nine, that means you stop shooting.
1:35:53
That means you lost confidence.
1:35:55
[laughter] >> That's pretty good. I like that.
1:35:57
Well, speaking of basketball, we have the CEO of the Minnesota Timberwolves, Matt Caldwell, here with us and Jordy Lyser joining as well.
1:36:04
Welcome to the show, fellas. How are you doing? >> Doing great. >> Hey guys, see you. >> Great to have you.
1:36:11
>> Uh, thanks so much for hopping on.
1:36:11
Um, I would love to start with the news.
1:36:14
Can you both uh introduce uh how this all came together and the various roles that uh how you guys see yourselves working together?
1:36:25
>> Yeah, Jordy, you can kick it off. >> Yeah.
1:36:27
Well, um, great to great to meet you. Thanks for having us on.
1:36:29
Big fan, uh, what you guys have built and been doing here.
1:36:32
Um, basically I started this company, Jump, with Mark Lori and Alex Rodriguez.
1:36:37
I think you had Mark on actually. I saw >> talking.
1:36:40
You also had Alexa, one of our one of our boys with Jump, too. >> Yeah. Nice. Love him. >> Yeah.
1:36:46
So, um, >> the whole theory was that Mark and I were going to buy a team >> and if they were going to buy a team, these guys are not sort of in the business of doing things the old way.
1:36:56
And uh and so buying the team, actually I just saw the Shopify plug.
1:36:58
We're trying to do for sports teams what Shopify does for merchants, which is basically totally unlock the direct to consumer proposition for a sports team.
1:37:07
And so Matt was hired, you know, right when Mark and Alex took over the club.
1:37:12
And so we got to know each other and then we were able to launch our platform in the NBA with the Timberwolves, uh with Matt, who's who's an unbelievable leader in sports executive um under the ownership of Marks when they finally took control of the club.
1:37:24
Yeah, when you think back to sports ticketing generally, just the different eras of technology, the risks, what they unlocked, uh what do you draw on as you know, things that organizations got right?
1:37:39
Any organizations that jumped on waves early and saw success?
1:37:41
Like what what what are you pattern matching against when you think about where you want the team in the organization to go from here? >> Yeah, Matt. >> Yeah, sure.
1:37:52
Thanks again for having me having me on.
1:37:54
Um, >> you know, listen, traditionally teams outsourced ticketing, right?
1:37:58
If you if you look at it from a team perspective, like there's a lot of work to be done.
1:38:04
You're trying to win games, you're signing players, you're uh marketing the brand [clears throat] and, you know, to create, you know, your own ticketing system that's uh, you know, transacting and selling season tickets and single game tickets, you know, it's a lot of work.
1:38:18
Um, you know, ticket master is has built a great business on this with both concerts and and uh, >> you know, sporting events.
1:38:24
Um, and you [clears throat] know, I think, you know, the last 10 years or so, teams are saying, okay, um, we have a lot of season ticket holders, but we don't really have data on people that just come to like one game or, you know, pick up a couple games during the year.
1:38:38
And >> you know when all that data is outsourced to another provider, you know, it's hard to figure out the the behaviors of that fan and and how do you market to them?
1:38:48
How do you make them a season ticket holder?
1:38:49
You know, essentially just just grow with that that fan.
1:38:52
So, you know, what what [clears throat] Jordi has done with, you know, Mark and Alex, um, you know, built a ticketing system that's in-house, you know, that, you know, when someone goes on our app now, and, you [clears throat] know, traditionally, we've had like, you know, our schedule and some content highlights on the app, >> but now you can just scroll and buy tickets on the app, you don't even know that you're on Jump's platform.
1:39:11
It looks like the team's app.
1:39:14
So, you know, it's a way to have the data on [clears throat] your side of the table on your app.
1:39:19
uh and you're working with a great partner, you know, uh with with Jordy and John.
1:39:25
>> So, how do you think about actually uh scaling the marketing now that you have more data, more of a direct relationship?
1:39:31
I've thought about times when uh I've gotten invited to a sporting event and my friend might have bought all the tickets.
1:39:40
Occasionally they'll say, "Oh, to get in, you actually have to download an app and and and and claim the ticket, and then you'll have it on your phone."
1:39:47
And then I should be getting retargeted to return the favor.
1:39:51
Maybe I should buy them a round of tickets next time.
1:39:53
Uh, and I I I have probably been served a few ad. I don't know.
1:39:59
I in one ear out the other a lot of with this stuff.
1:40:01
But how do you think about building your marketing organization internally?
1:40:05
you know, Shopify has some buttons where once you're set up, you can just say, "Oh, yeah, run some Facebook ads."
1:40:10
You're at a much larger scale.
1:40:11
How do you think about the rest of the organization internally executing this strategy?
1:40:18
>> It's a great example and and traditionally, you know, the the experience you talked about, your friend transfers your ticket, you download Ticket Master, StubHub, Segeek, um >> you know, all that data, that that person's information, cell phone, email, all stayed with the third party.
1:40:33
Now we're getting access to it.
1:40:37
>> Uh so we can have our sales staff call them.
1:40:39
We can have automated emails going to them.
1:40:42
>> Uh we can go visit them in their seat.
1:40:44
You know, we know someone came in, we know their name.
1:40:46
>> Um but you know, something that's really helped for us, you know, going to the next level is, you know, we've gotten to the point with with Jump where, >> you know, even if someone puts like a ticket in the shopping cart, you know, but doesn't close it out, uh or even just browses over a game, >> uh you know, we get that information, we get an automated email to that person.
1:41:05
Hey, these two tickets are still available.
1:41:06
Or even better, hey, those two tickets were just sold, but there's two other ones that are similar.
1:41:10
Uh, so you just have just more engagement, especially with, you know, what I would call the casual fan, single game buyer.
1:41:18
Like our season ticket holders, we know them really well.
1:41:19
They sit courtside, they're in suites, you can visit them at the games.
1:41:23
We every season ticket holder has a uh their own individual like concier sales rep.
1:41:28
But it's the people that like yourself got transfer ticket, come to a game.
1:41:32
We we want to make them longtime fans and now we just have so many ways to to reach out to them.
1:41:36
Now >> I went last last time I went to a I went to an NFL game and I swear I had to download three apps to get the actual ticket and even by the end I'm like I hope I got all the apps I need to get in [laughter] to use my ticket.
1:41:50
>> The parking ticket hot dog app and >> and it was a free ticket too.
1:41:53
I was like it should shouldn't be this hard. Um uh how are you? >> It was a lot of work. [laughter] Lot of work. I was committed though.
1:42:01
Uh Jordy, how are you thinking about uh like using this to deepen the the like fan engagement?
1:42:08
Knowing Alexis, like he's, you know, obsessed with collectibles.
1:42:12
I imagine by having that direct relationship, you can unlock, you know, different products and things like that for fans based on their actual engagement with the team.
1:42:21
So that pe, you know, fans have to actually like earn certain things.
1:42:25
But where what does that look like? >> Yeah, you nailed it.
1:42:29
>> Yeah, you nailed it. And actually this hits the problem of you know the three apps and the new login and the Lakers do some deal with some startup and then you got to create a login for that thing to like you it's just the whole thing is just one Frankenstein tech stack you know mess and so again
1:42:46
Shopify is the best comp like it sort of consolidates that into one seamless experience then everything else plugs into you >> so in that same vein collectibles merch concessions add-ons parking you should have you know the Timberwolves login the Timberwolves account add to cart cool
1:43:03
jersey add to cart cool collectible oh hey I know that you you are a season ticket member and you came to you know KG's coming back for the Wolves this year it's a big part of the story so they should know that these people who were there during the KG era you know it's not just that they know the CRM like we know Bob Smith and we know where
1:43:20
he lives and we know it's like >> did Bob go to those games was Bob you know kind of there during the era and then you can nudge and you can you can sell and you can upsell and you can engage and So, it's content, it's merch, it's media, it's everything sort of in one system. >> Have are are you thinking about
1:43:33
>> Have are are you thinking about cross-selling across different sports?
1:43:36
When I think about the Shopify example, uh they've launched the shop app, you know, you oh, you bought this pair of boots, maybe you want this pair of pants.
1:43:45
Is there a world where you bought this ticket to a basketball game, here's, you know, it's the offseason, here's a football game or a a baseball game.
1:43:54
Is that something >> that's a little bit more That's a little more the market.
1:43:58
That's kind of the marketplace model that you know that's the there's a bunch of ticketing marketplaces and there's there's YC versions of that and there's stuff you know >> and I think that that's the thing that you know Jordan you were saying is just kind of the beginning of the mess of this whole thing where if I want to buy you know Delta Airlines like if I'm going to fly on Delta I got the Delta app.
1:44:18
They know everything about me.
1:44:18
They went direct to consumer 20 years ago and like that's that's it.
1:44:23
>> And so I think that's the missing piece in tier one sports at least.
1:44:25
But it's really like if if you're if you're in Minnesota, you've been to a Wolves game.
1:44:30
Hey, let's kind of like get to know you as a Wolves fan just like you were, you know, a customer of any other brand. >> Yeah.
1:44:36
>> How sharp are the elbows of the of the other players in ticketing?
1:44:38
I mean, I I can't imagine [laughter] they're thrilled that a uh a platform that's allowing people to go D to C popped up with a bunch of backing. >> What do you think?
1:44:49
Yeah, I [laughter] imagine they have got like little knives right here and they're just >> But yeah, it feels like something that's so aligned, so aligned to the team, so aligned to the fan, it's just misaligned with the legacy >> uh players that no one actually I think likes. >> Yeah.
1:45:07
>> Well, that's a great question.
1:45:07
If you're investing in startups, do you want to invest in a platform that is beloved by its customers and beloved by its customers customers but not beloved by the 50-year-old, you know, incumbent?
1:45:21
Like >> I kind of like being on that side of history and I feel like we're we're in a good spot.
1:45:25
But at the end of the day, I mean, I'm in Los Angeles. Think you guys are here. >> Sure.
1:45:29
>> This is a town that just And you're in this space, >> sports, entertainment.
1:45:32
I would say also in like the team owner space, you've you know the headlines are pretty pretty wild recently.
1:45:39
>> Yeah, >> there's just a lot there's a lot of going.
1:45:42
>> Yeah, I was going to say I don't I don't you know we're not super tapped into to the industry, but I was going to say if you do any endorsement deals with athletes, make sure they actually endorse [laughter] the company.
1:45:50
Uh just I I that's what I've learned.
1:45:53
We had Pablo on the show earlier uh and I that was one takeaway, you know, make sure they actually talk about Jump and and what they're doing.
1:46:02
Well, no, that's that's Matt's that's Matt's world.
1:46:05
[laughter] >> We No comment. No comment.
1:46:08
>> Uh, what what what else is happening in the world of technology and sports that you're tracking?
1:46:12
I mean, AI is getting slapped on everything.
1:46:14
And I love this story because it's it's just a very clear software enablement, like the actual relationship.
1:46:22
It gets to like the core of a business problem that you're trying to solve.
1:46:26
Uh, and then there's technology that is really powerful, but it's maybe a little bit further out.
1:46:31
But what else are you harnessing at the Timberwolves that you think is is interesting or you're starting to explore us?
1:46:39
Like talk to us about technologies role over the next year next year or decade. >> Yeah, Matt. Go. >> Yeah. Yeah.
1:46:47
I mean, [clears throat] the first thing that pops to mind, you know, for me is is the the pricing of our tickets.
1:46:54
If you think about, >> you know, we about 20, let's call it 20,000 seats in an arena.
1:46:59
>> You know, maybe half of them are season ticket holders, right?
1:47:01
So, you have 10,000 seats that you're selling, you know, at least 41 nights a year.
1:47:06
And if you have a manual process where you have different price codes and there's, you know, supply and demand, you know, and every couple hours or every day or so, you're updating, you know, uh, prices and and there's transactions happening.
1:47:21
um you're missing out on margin or or you're missing out on, you know, lowering pricing for a certain game and and getting more people in the building.
1:47:28
So, I mean, AI is going to explode sports just like it's doing for every other industry.
1:47:32
other industry. And you know to be with someone like Jordy that's you know very AI forward uh you know our pricing has gotten so much more better automated um you know pushing revenue and then and then we talk a lot about open distribution you know with all these
1:47:49
different ticketing systems you can imagine you know uh there's some tickets that are on Ticket Master Segeek StubHub uh they're all kind of spread out [clears throat] you're not getting the full marketplace looking at your tickets of all at once. So what we've done with
1:48:03
of all at once. So what we've done with with Jump is, you know, any ticket we put on the Jump platform, let's say, you know, a single game ticket in the upper level, you know, that same ticket is now being put on ticket masters, you know, TM plus, you know, uh, you know,
1:48:17
secondary market, seed geek, >> uh, yeah, [clears throat] stuff up all the secondary markets so that you're getting all that demand, all those eyeballs looking at the ticket and interesting, >> you know, you're hopefully getting, you know, hopefully driving up pricing if if the demand's there. So, and that's not
1:48:29
So, and that's not possible in the old way because obviously Ticket Master wants them on Ticket Master and Segeek wants them on Segeek.
1:48:36
And so our our model is like >> we you know if you got AirPods they they control the distribution.
1:48:40
Apple says I want some in Best Buy. I want some at Target. I want some Amazon.
1:48:43
I like we give the teams the total control to distribute.
1:48:47
And that's a that's a totally different model.
1:48:50
>> Uh our head producer Ben from Minnesota.
1:48:53
He just texted us and said, "I actually bought Wolves tickets last night using Jump. It was great." So >> all right. two thumbs. >> That was my plant. That was my plant. Yeah, I texted bad.
1:49:01
[laughter] >> Now, I saw in your feed there was some drop about a little bit of a little bit of a negative vibe about maybe the wolves won't get [laughter] all the way to the other side.
1:49:13
I I saw that recently, I didn't want to call it out, but since you mentioned it, like >> what's going on with that? >> No, no, no. Ben's a fan.
1:49:20
Ben >> Ben looks confused.
1:49:22
He's >> he's a he's a true fan. True fan. Thank you so much.
1:49:29
>> Uh yeah, great to meet you guys.
1:49:29
I love a business that just makes so much sense.
1:49:32
You guys were probably like feeling out this idea is like is there a reason this thing doesn't already exist?
1:49:37
Seems seems really obvious.
1:49:37
Seems really aligned to everyone involved. So very very cool. Great to meet you. >> Cool. Thanks guys. >> Thanks so much. Appreciate it.
1:49:44
>> We'll talk to you soon.
1:49:44
Let me tell you all about Cisco critical infrastructure for the AI era.
1:49:49
Unlock seamless real time experiences and new value with Cisco in death match. >> Uh up next. >> Yeah.
1:50:00
Wait, we're heading to the portal.
1:50:02
>> We got to get to the bottom of this.
1:50:02
Ben actually this is true.
1:50:03
Ben, you bought Wolves tickets last night using Jump? >> Yeah. Yeah. >> Okay. >> Yeah, I did. It was great.
1:50:09
>> Wait, are you a Wolves fan? >> Uh yeah. >> Okay.
1:50:12
Wait, why the hesitation? >> Not huge. Uh not huge.
1:50:15
>> It's a surprise for someone, so >> Oh, okay. Oh, sorry.
1:50:17
[laughter] whoever you watching. >> Okay, edit that out.
1:50:23
It's not like it's a live stream or anything.
1:50:24
[laughter] Anyway, I I believe we have our next guest already here.
1:50:27
So, let's bring in Jeff Thorn Thornberg from Portal Space Systems. >> It's portal time. >> How are you doing? [music] >> Fantastic.
1:50:38
>> Welcome to Thank you so much.
1:50:38
Uh let's uh let's start with the decision to start this company.
1:50:44
I mean, you you have experience from SpaceX.
1:50:46
uh what was the original thesis?
1:50:48
What was the the the the first uh prototype pitch deck?
1:50:51
How did you even uh begin this journey?
1:50:59
>> Uh so many things but I think the simple answer is uh after 30 over 30 years in the space business.
1:51:04
I had a lot of colleagues in the defense side >> that was really struggling with how do they combat adversarial advancement on orbit?
1:51:12
And uh we're running our spacecraft like hot air balloons and they're running theirs like fighter jets.
1:51:18
And so uh we really set out to build the fighter jets for orbit here at Portal while also allowing more space exploration for for NASA and the civil space community. >> Yeah.
1:51:28
What is the benefit of being able to maneuver in space generally?
1:51:31
I mean we've seen you know the trajectory of of traditional rockets, ballistic missiles.
1:51:38
Uh seems like a lot of stuff can happen in space, whether it's the Starlink constellation or the International Space Station.
1:51:45
Uh none of that required too much maneuverability. Why maneuverability? Why now?
1:51:51
>> Well, you've you've all seen the movies, right, where uh the the adversarial country runs out in the field and covers their stuff with tarps because they know the satellites coming overhead to take pictures and >> and so most things on orbit are very predictable.
1:52:04
And you know, to mix it up, we used to not have any real competition on orbit. And now, guess what?
1:52:08
Uh, China and Russia are challenging us um on the defense side.
1:52:14
And, you know, we don't want people throwing tarps over stuff.
1:52:17
We want to show up and do the things we need to do when we want to do them and not be constrained by the lack of fuel in the gas tank on orbit.
1:52:23
And a lot of the ways this has been looked at in the past is it's like you went to buy a new car and you throw that new car away as soon as you use the first tank of gas.
1:52:33
And you know, we really set out to build platforms where gas is not in the equation anymore. Use it, move it quickly.
1:52:39
You know, there's some generals out there that have this phrase I really love where, you know, we drive our spacecraft like we're going to church and our adversaries drive them like they stole them.
1:52:48
And and and [laughter] and we really want our customers to drive them like they stole them.
1:52:53
And and that's really what built the foundation of the business 5 years ago. >> So what Yeah.
1:52:59
What has progress been like over the last five years?
1:53:00
Uh is this are you at manufacturing stage?
1:53:03
Is this a prototype level?
1:53:05
Uh I imagine that you know you got to actually send stuff up to to at least low Earth orbit to really test the capabilities, but uh what's the what's the progression been like?
1:53:19
>> Yeah, I first have to really give a shout out to my team because none of what we've done would be possible without them.
1:53:24
And you know, we're about 60 people now here north of Seattle and the progress has been amazing.
1:53:28
I've got to work with some amazing people over the years.
1:53:32
The biggest compliment they've given to me is wanting to come back and work with me again.
1:53:35
And we were able to launch our first mission to orbit in March and successfully test a lot of our electronics hardware for the brains of our systems.
1:53:43
Uh and then we just finished our first uh small Starburst spacecraft that's about 600 lb that's going to go up and really demonstrate our ability to do some of the things we just talked about and that's going to launch October 19th.
1:53:55
So the the team has been grinding for the last several weeks to get ready to ship to the Cape and we're going to do that in the next few days and uh they couldn't be more excited or more fatigued all at the same time but they're doing a fantastic job getting us ready. >> Very cool.
1:54:10
uh space uh you know conflict in space is something that uh you kind of we hear about from from entrepreneurs and and various people that are in defense tech but it's not something that you necessarily read about in the newspaper.
1:54:27
It's happening uh way up above us.
1:54:30
Uh and and it's not something, you know, it involves private companies and governments, but no one is really incentivized to come out and and say like, "Hey, we we had this issue."
1:54:39
It seems like it just kind of gets talked about behind closed doors.
1:54:44
Do you think this is something that over time private companies will uh be like end up disclosing?
1:54:52
like when when do you think there's more of like a conversation around this or is it something that uh just kind of stays uh uh talked about behind closed doors?
1:55:03
>> I think there's changes happening right now in the conversation, but it's tough, right?
1:55:07
Because when national security is impacted, you don't necessarily want to be telegraphing all of the things that are going on.
1:55:13
So I think there'll be some element of protecting that information, but I think there's going to be more companies like Portal and you're seeing them now that are going to have a significant part of their business be deflated.
1:55:28
and the Department of Our and the Department of Defense in general are really trying to bust up the legacy compartmentalization of information so that we can support the defense department and and the agencies in ways that frankly they have to have.
1:55:43
And I think what's happened with a lot of the acquisition reform over the last few years is trying to get capability in the hands of the war fighter as quickly as possible.
1:55:51
Uh so there's a much more willingness to partner with industry.
1:55:55
A lot of the decision makers are saying, "Hey, bring it all. We want it all.
1:55:59
We want to maintain our number one position in the world and the western world with our allies in controlling the space domain."
1:56:06
Uh, so you're seeing a thaw in that, but I think there'll always be elements of national security where you will know there are things happening, but you won't necessarily know all the details just so the public can understand why it's important.
1:56:18
But I what I would point people to is just take a look at that 60 Minutes um report from a couple of years ago where they were with the Philippine Navy in the South China Sea and they were actually filming how China was engaging the Philippine Navy. It's all public domain.
1:56:34
It it all gives you a great idea of what's actually happening in other parts of the world with adversarial aggression and tactics.
1:56:40
And I think it's a great example that you know people could then kind of take forward.
1:56:46
And why is this important?
1:56:46
Um it's important because uh we protect our interests in land, air, and sea and now space.
1:56:54
And space has to become a domain that we defend for future commercial and commerce activity.
1:57:02
>> What's the biggest lesson that you're taking from SpaceX to Portal? >> Tenacity.
1:57:10
>> Not taking no for an answer.
1:57:10
M >> um every problem has a solution and with the right team you can get there and I think those are the biggest lessons I took away.
1:57:18
>> Do you have to show that to the 50 employees that you have on your team?
1:57:21
Do you is it enough to say it?
1:57:24
How do you actually display it if you're a CEO?
1:57:28
Because uh tenacity can look different in this particular seat.
1:57:33
>> Yeah, I think it's tenacity coupled with you know maintaining a cool head, >> working the problems as they come.
1:57:37
M >> um I joke with my group, you know, how many times do you see me running around with my hair on fire?
1:57:43
You know, that's that's not the right approach to inspire confidence in your team.
1:57:49
>> So, I I think the bottom line is uh you've got to project um a calm but aggressive attitude to work through these issues.
1:57:56
And I think just having experience in doing that in the past, which certainly all of us that came out of the SpaceX environment have an incredible amount of experience in in solving impossible problems and moving forward with a great team.
1:58:10
>> Well, Jeeoff, congrats on the progress.
1:58:12
Thank you for everything you're doing and thank you for coming on the show. >> I'm glad.
1:58:15
Well, I'll be driving it like we stole it.
1:58:18
>> I was about to say the same thing.
1:58:19
>> That's what I want to hear.
1:58:21
>> I want to hear about people in space, you know.
1:58:23
You know, they say in AI, they say you don't need to drive a Ferrari >> to to get groceries, but I hope we're going to be driving Ferraris like we stole them. >> Yes.
1:58:32
>> In space to do simple things like getting space groceries.
1:58:34
[laughter] >> Absolutely.
1:58:37
>> Kind of lost the plot.
1:58:39
>> Anyway, thank you so much.
1:58:41
>> Great to meet you, Jeff. Have a great day. We'll talk to you soon.
1:58:45
>> Next, we got groceries.
1:58:45
People are not talking about space groceries enough.
1:58:49
>> Really mixing metaphors there, I think. [laughter] Uh 10.
1:58:51
Uh the >> we got Charles from the head of AI model training. How you doing Charles? Welcome to the show. >> Hi. >> What's going on?
1:59:03
>> What's the best model you ever trained?
1:59:06
>> Best model I ever trained.
1:59:06
Um I can't talk about it.
1:59:08
>> Name every name every model. I didn't realize this.
1:59:11
I think every parameter has 5 million models trained. Uh is that enough? Do we need more? >> Um well, yeah.
1:59:19
I guess our bet is that, you know, we're going to end up in a world with some great models from the Frontier Close Source Labs. Sure.
1:59:26
>> But we're going to end up with, you know, not 5 million, but probably hundreds of millions of models.
1:59:29
Um, possibly even one model for each person, possibly, you know, tens of models for each person.
1:59:35
And like I guess that's the future that Base 10's betting on. >> Yeah.
1:59:38
So, what's the what is the framework or what is the the the the next model, the thesis behind model development that you want to execute on in this role? >> Yeah.
1:59:49
So I think like to be honest and the the closed lives will tell you differently but everyone is scaling the same recipe right now.
1:59:56
Like there is no difference between OpenAI's RL stack and you know the Chinese open source RL stack and the American open source RL stack.
2:00:05
>> We've got the same recipe.
2:00:05
People believe it scales.
2:00:08
>> We first scaled pre-training and model size. We're now scaling RL.
2:00:10
Um and you know people are going to keep doing that till the cows come home.
2:00:14
And the bet is that you know you hit higher and higher levels of intelligence and you'll be able to do more and more economically useful things.
2:00:20
I think to us like we care a lot for instance about continual learning.
2:00:24
So we announced space labs today which is going to be doing like less myopic longer term research around what models can actually do.
2:00:31
>> And continual learning looks very different for us as it does to the big labs.
2:00:34
>> Like the big labs already in this kind of continual learning loop of you know they train a model >> GPDN or or claude and they release it into the world.
2:00:41
They collect feedback on what it can do, what it can't do.
2:00:43
Then they build ton of Ral environments to like patch the holes in it >> and they go back and like from a from a high level view bird's eye point of view that is continual learning, right?
2:00:53
>> Um the continual learning that you or I might imagine is very different.
2:00:54
It's where you know you have an open source model that you specifically using either as a firm or a team within a firm or even an individual.
2:01:01
And that model is organically adapting to the information that it's learning.
2:01:06
it doesn't have to like you know write everything down in memory markdown files because it learns about you know your business and and so on.
2:01:12
Um and then we think there's like major paradigms to be unlocked within that regime and it's not necessarily going to get the focus from the big labs because that doesn't necessarily benefit them right like they want to serve one big model at scale and the pitch they've they've sold to their investors is that you know you do that at a large enough scale and you don't need this this kind of continual learning. >> Yeah.
2:01:30
How how do you how do you process their claim?
2:01:32
Because a lot of the big labs, there's this dance between, oh, well, there's a fine-tuned model.
2:01:36
It was really good at this one benchmark, then the next version of the big model that does everything is better than the fine tune.
2:01:43
And it feels like this horse race where I can totally see the cost argument and I can see even like the yeah, for 3 months this fine-tuned model was better.
2:01:54
Uh but but if we're talking just like raw capability, uh h how how do you how do you interrogate that claim that the big god model will you know always be better as long as you give it time? >> Yeah.
2:02:08
I this is where I think people are thinking about like intelligence capabilities in the wrong way. Okay.
2:02:13
Like people think about intelligence relativistically.
2:02:15
They say okay you know the open source gap is like 6 months behind closed source and you know GLM 5.
2:02:21
3 is at the point that opus 4. 8 and it was whatever.
2:02:24
I think the best way to think about what models can do for you and for the world is like absolutely.
2:02:27
So for any given task that you want to do with an LM, there is some intelligence threshold that below that you can't do the task and above that you have very diminishing marginal returns to more intelligence on the task.
2:02:38
And so when you think about it that way, the game of LLM over the last 5 years has been okay, we have these things we want to do with them.
2:02:46
them. close source hits it first which to be honest we think is a good thing for for many reasons which we can get into but you know open source eventually like six months later or nine months later or whatever it is can then do that task and then for many reasons like whether it's to control your own intelligence whether it's to improve at
2:03:01
that task specifically once you've had the base level of intelligence required to do it um you you probably do want to swap to open source and so it's not really about you know the god model being better like if I'm you know filing a tax return there is just a limit to like how much intelligence I need to do that particular thing. And so I think
2:03:17
And so I think the world is going to look like you know the frontier source labs are going to continue to push the frontier like we you do want to use the the most intelligent model.
2:03:26
You have very like inelastic demand for intelligence when you're doing like you know frontier science or frontier maths.
2:03:32
>> But for a lot of the economically valuable things it looks a lot like okay you know I'm a cursor or I'm one of these big companies who are now realizing like I can't just be a rapper anymore.
2:03:41
I've been through the life cycle of building a great product that people love and they tell me what they love and hate about it and I should be using that information to, you know, make my model better at the things that I care about and not at anything else.
2:03:52
And so that that that pattern was executed fairly well with composer.
2:03:54
Um it's obviously early days and the ecosystem isn't mature enough for anyone apart from you know the curses of the world and a few other big companies to to go about training but you know I think eventually we'll get there where it's where it's much more organic.
2:04:06
>> Can you talk about the the the goals of this project?
2:04:10
I mean I I I I understand the the research direction continual learning but is the goal more to uh to advance the research or produce a continual learning model product that is then sold.
2:04:25
Uh how are you balancing the discussion that needs to happen within the the ecosystem versus uh productizing something that could be a really meaningful break breakthrough?
2:04:36
>> Yeah, it's a good question.
2:04:36
I think like the founding kind of mandate of base labs is do whatever we can to make open- source models and probably bleeding into closed source models eventually as useful as possible.
2:04:49
>> And so it's very difficult to specify a priority what that looks like.
2:04:51
Like one obvious place to bet is yes this continual learning paradigm and doing research around that >> and like thinking about that in a different way to how you know the big labs think about continual learning on like you know millions of people at once. >> Yeah.
2:05:02
But like we're not hubistic enough to say that like that is the only way in which we're going to make models more useful.
2:05:08
useful. So for instance another thing we're thinking a lot about at the moment is everyone talks about aggregating comput for open source to keep up like you need a certain number of chips >> be able to train these things but not as many people are talking about
2:05:18
aggregating data like the closed source labs are spending billions a year on our environments and again like a is the next paradigm that everyone's scaling >> you know distillation and like you know cheap data and centralized data there's
2:05:30
a lot of discussions around these things about how the open source labs are currently keeping up but you know what keeps me up at night is I wonder if there's a point where it bifocates And you can't like get all the way there, you know, distillation and whatever data
2:05:42
you have access to like someone needs without a commercial incentive needs to be making these incredibly complex environments that you know normally cost billions of dollars um in the aggregate like and open sourcing them to the world. So a key part of I guess base
2:05:52
So a key part of I guess base labs is like >> can we make those environments and we've been doing this for the last few years like we've been making them for individual people why don't we just make them for everyone and release them.
2:06:00
So you know any open source or closed source provider can train on them and you kind of like cut the gap that way.
2:06:07
So aggregating data is another big bet that we want to make.
2:06:09
And there's there's a few other things you know like one one thing that we're starting to notice a lot is like as the capabilities have risen and you know even the open source models have kind of subsumed all the economically valuable tasks that like most people are going after and it's very difficult to tell the difference between an open source and a closed source model.
2:06:25
Then you start to think about what the values of these models are.
2:06:28
And you can't not have an opinion when you're training a model about what the values and ethics and morality of that model is.
2:06:32
like in your pre-training data and in your mid-training data in your post training and your your classifiers and your safety stack you run on top of it.
2:06:40
Everyone has an implicit or explicit opinion about what the values are.
2:06:41
It's a reason why so many American companies don't want to use Chinese open source is because we're not very clear what those those values are.
2:06:47
So how do you organically shape those values to what you want them to be both as a country, as a firm, and as an individual?
2:06:53
Like can we do post training to elicit the right values and models?
2:06:57
uh trust in benchmarks feels like it's at an all-time low and maybe is headed lower.
2:07:04
What what do you what what what other uh do you have any other ideas of how to communicate the sort of um maybe it's still that they they work well enough but uh uh any other ideas around how to communicate with certain new models >> bench maxing bench hacking and then also
2:07:23
just like inundation with so many that a lot of consumers and enterprise buyers they sort of just roll their eyes when they see a new one because they're like I got I've seen a hundred of these I I know that you can probably do well on one. Um, but yeah, what I'd love to hear
2:07:35
Um, but yeah, what I'd love to hear this your thinking.
2:07:39
>> I I think that let's think about like who makes benchmarks, right?
2:07:41
Like it's typically a bunch of like, you know, research fellows from Stanford or what have you who've pulled together and they're very smart people and they're like, okay, this is what, you know, this is this weird way in which we're going to trip this model up.
2:07:53
Like I think RKGI is like a very good example of this like weird pocket where the models like traditionally don't do well and they've obviously updated the benchmark more and more as the labs have have kind of caught up.
2:08:02
The common saying is like you know once you've benchmarked it you can RL on it.
2:08:05
So you know releasing these makes the models better.
2:08:07
I think the best way to do this is aggregate the economy and this is where we actually see us having a really big advantage compared to particularly the closed source labs like I do believe anthropic and open AI are generally good and like you know they don't look at user data if they say they don't and like they have this very like bird's eye view of like what the models are good at and what they're bad at.
2:08:25
It's why they go on Twitter and ask for feedback.
2:08:26
We have a real advantage in the sense that we have like you know thousands of people using them for different things and each of those individual companies has tried to build their own emails and benchmarks to test how good a particular model is at the task that they care about.
2:08:41
And so rather than you know like there's there's no like silver bullet here like there is no set of 10 benchmarks that is going to tell you you know exactly the core and the the jagged frontier of whatever model it is you're testing.
2:08:51
The best way to do this is to look at how everyone is using them in the economy for real things that people are paying for, not weird like niche things like ArcGI and aggregating those benchmarks.
2:09:03
And like I guess that's part of what we're trying to do with the RL environments and pumping them out at scale.
2:09:07
>> But yeah, like we we get a lot of really good signal from >> the record for the record here at TBPN, we had Tyler do ARKGIV3 tasks manually as a human.
2:09:16
He technically got paid for it. So you know what?
2:09:18
I he was he was he was globally ranked.
2:09:24
>> Globally ranked >> for for like a week. We were very early.
2:09:26
I think day day of launch he was like number seven.
2:09:29
>> Charlie, what does it mean to feel the whole elephant? >> Oh yeah. [laughter] >> Yeah. >> Great one.
2:09:35
>> Um I've I've gotten a lot of flack for this.
2:09:36
Um and people have interpreted it in quite inappropriate ways, including my CEO actually.
2:09:40
Um, but there's this old analogy of like, you know, blind men touching an elephant and like they don't really know what it is they're looking at and one's feeling the trunk, one's feeling the leg, one's feeling the tail.
2:09:51
I think it is a little bit like that with things like LMS and particularly like continual learning.
2:09:56
>> You know, some people say that continual learning is just like having a god model with, you know, a million tokens of context and it can search whatever it needs to and organize the information that way.
2:10:04
Other people really believe that no, every single token it should be updating, you know, its information.
2:10:07
um we know that that fries the model in different ways.
2:10:11
I think Thomas um wrote a lot about like the structure of scientific revolutions and I think we are pre-paradigm when it comes to things like continual learning like no one can agree on what the definitions of these things are.
2:10:21
things are. No one can agree on you know apart from the core recipe to produce the base LLM in the first place like what the right ways to be scaling this is you know the ecosystem around it like compaction is a great example of this
2:10:32
like open anthropic have gotten really good at compaction in in clawed code and codeex at the moment that's just summarizer models like there's probably like really really cool things that you can do if you train like you know neural
2:10:44
compactors if you have models themselves doing compaction in KB case space so I think there's a lot of work to like make this a science because so much of it is inside the closed source labs and I guess that's the the mission of base
2:10:55
labs is like how can we bring as much of this into the open as possible without a commercial agenda like even the open source labs have a commercial agenda they need to produce the best model this quarter and we don't have that pressure
2:11:06
and so like I guess we can pick the most interesting scientific problems and work on them for as long as we need to before we feel like we've made traction >> what is your >> very cool >> P brute force the probability that the
2:11:17
answer to continual learning is just brute force when I think about the what openthropic do for updating a model you know you mapped it out it's a it's a couple months of data collection building RL environments retraining the model that takes GPU hours but if you
2:11:33
get a 10x speed up in the amount of time it takes to train the actual model versus you can build the RL environments 10 times faster you can generate the data 10 times faster and you start and you start working at this over maybe
2:11:47
it's a decade is there a world where you could just do exactly what we're doing now, but every second and it feels continual because when I talk to anyone, they're continually learning, but they take a second. They say, "Oh, yeah,
2:11:59
They say, "Oh, yeah, okay. I'm updated. I I now know that fact." >> Yeah.
2:12:04
I I think there's a few ways to break this down.
2:12:06
From the perspective of RSI, like recursive self-improvement, >> I actually think that, you know, continual learning is like either unnecessary or it is brute forceable. Sure.
2:12:15
Sure. like you know we're going to asmtote towards models that have access to the whole you know open AR anthropic training stack they're going to make slight architectural improvements they're going to drop pre-training loss they're going to be able to pump out RL environments scale by themselves um and then that process will speed up and speed up and speed up as we get more and
2:12:30
more compute but for the everyday person and the everyday like you know AI native startup or enterprise or whatever continual learning doesn't look like that like at the scale that they're operating at you can't afford to like >> open anthropic they wash out all the noise and the gradients they take all the useful stuff and these big pre-training runs makes it work. >> But when you just focus on like, you
2:12:47
>> But when you just focus on like, you know, I have an agent which is a legal associate and I'm trying to fine-tune it on all these like complex relationships and all the things that the firm does and this implicit behavior that we want it to have.
2:12:57
>> Continual learning breaks down.
2:12:57
We don't have an answer to it. We can't SFT. It degrades the model.
2:13:00
We can't RL because it doesn't give knowledge acquisition in the right way.
2:13:03
You know, there's there's a lot of work to be done there.
2:13:05
So, I guess it depends on again which part of the elephant you're touching.
2:13:07
Like what definition of do you care about?
2:13:10
>> We got to get an elephant on board.
2:13:13
[laughter] Got horses part of the elephant you touching.
2:13:16
Thank you so much for coming on.
2:13:17
This is a fascinating discussion.
2:13:19
Congratulations on the project.
2:13:20
>> Yeah, great to meet you, Charlie.
2:13:21
>> Very excited for you to solve this. >> Come back soon. This was fun.
2:13:24
>> We'll talk to you soon. Have a great day. Goodbye.
2:13:28
>> We need an elephant >> snowflake. >> Bye tomorrow.
2:13:30
>> We got the CEO with us here live on TBPN. Straightart Ram Swami. Welcome back >> to TBPN. How are you doing?
2:13:36
Thank you so much for joining us and congratulations on the fantastic success.
2:13:40
[music] break it down for us. What's working?
2:13:42
Is there anything that's not working?
2:13:45
It seems like everything's going really well. >> How are you doing? >> Hi. Am I live? >> Yes, we're live. We're live. >> Welcome to TVPN.
2:13:53
Thank you so much for hopping on. >> Oh, thank you.
2:13:57
Um, we had a pretty amazing quarter. Uh, $1.
2:14:00
49 billion, 37% yearonear.
2:14:05
>> Our uh AI products, Coco and Co-work getting really broad uh adoption. >> Mhm.
2:14:10
We feel very good overall, but this is also a time of intense competition.
2:14:16
There's the hyperscalers, but also the foundation labs.
2:14:19
>> So, we think there's lots of business to be had, but a ton of change to keep up with both for our product teams and for our uh go to market teams.
2:14:28
>> Uh so, yeah, it's a day after earnings are done and uh it's it's back to work, pedal to the metal.
2:14:35
How much time are you spending trying to identify like a venture capitalist the next company that's going to be a major snowflake customer?
2:14:46
Because I I feel like the big labs, you identified them.
2:14:51
They're obviously huge consumers of data, huge businesses there.
2:14:53
Uh but it feels like every time we talk to somebody on the show, there's a new business that's getting venture funding.
2:15:01
They're getting users, they're getting revenue, and I can I can tell that they're just generating a ton of data.
2:15:06
>> Yeah, they're ramping ramping revenue >> way faster than any company, >> and they're probably ramping data two orders of magnitude faster than they're ramping revenue even.
2:15:14
So, what so what is what does that look like for you and for Snowflake?
2:15:20
>> Yeah, we spend a lot of time uh both with our own innovation.
2:15:23
What are great new ideas to be had at this moment because so much is possible.
2:15:30
uh AI has is industrializing software.
2:15:32
It's much easier to create it than it was before.
2:15:34
So I spent a lot of time with the product and engineering teams back to basics innovation.
2:15:40
>> We also meet a lot of startups and companies both established ones talking to them about the opportunity that is possible with AI talking to them about what we have done with our own sales teams uh but also brand new startups.
2:15:49
I met a startup I think it was day before yesterday barely 3 months out but uh they have dozens of customers all doing reinforcement learning on their platform.
2:16:01
We were talking about how we could better partner together.
2:16:02
So good amount of time spent on both sides of the cycle.
2:16:08
But what I think is unique and cool is how quickly ideas go um from it's just an idea to prototype to it's a product feature to getting adoption with a lot of customers. Yeah.
2:16:20
Uh, with all of the growth, I imagine that there's new challenges.
2:16:25
Where are you going to be, uh, focused on unbottlenecking the organization?
2:16:31
Are you going to be hiring, putting more AI, as you mentioned, in the hands of even an even broader swath of the organization, even though I imagine everyone has access to some level of token budget.
2:16:41
Uh but what does what what does the growth side of the business what are you trying to unlock in the next quarter or the next year?
2:16:50
>> For um the majority of the company the biggest mind shift u mindset shift that they need to go through >> is scale is not just about people any longer. >> Mhm.
2:17:04
>> What AI makes possible is stuff that needed individuals people to go and do >> can largely be automated.
2:17:11
Yeah, >> it's much more about the judgment.
2:17:14
What's the work that you're trying to get done?
2:17:15
What do you how do you want that?
2:17:18
Uh, you know, how do you want that done?
2:17:20
>> And then setting up a framework by which things can get stamped out.
2:17:24
>> U, but that's a pretty tough change for an org or a set of people even you and me >> that have traditionally looked at organization size as an important barometer of growth and that's something that we are stressing with the entirety of the snowflake team.
2:17:38
Yeah, >> I expect some functions to grow, but these are typically the ones that do things like interact with the external world.
2:17:46
>> So, account executives, I can see that growing for customers that are spending, but for a lot of other functions, including in engineering, there is so much leverage that you can get by being at the forefront of what is possible with agentic AI.
2:17:59
We continue to hire folks especially young folks who know of no other world because they bring a perspective um that is fresh and they know how to go from zero to 100 straight off the bat.
2:18:11
Um but a lot more focus is on how do we become effective as a company and a lot of it is also about how do we set up new efforts and give them space to experiment so that they can find that magic in a bottle. >> Yeah.
2:18:29
AI is not going to help us create great new products just like that.
2:18:33
There's an amount of experimentation to figure it and giving space to small teams to execute is among the biggest challenges that we need to make sure that we solve.
2:18:41
Koko for example, it's a breakout success >> but for much of its existence it never had more than five people and even now the team that works on Koko is tiny but that's the kind of impact that is possible today.
2:18:56
>> Are internal meetings underrated?
2:18:56
Do would you expect that in a world where uh AI can do more of the work but lacks some of the context and taste and decision-m and deciding what to do if you woke up and you said wow we're you know as a company Snowflake is having 10% 20% more internal meetings between various members of the organization.
2:19:22
Would that sit well with you?
2:19:22
Because there's been a long history of people oh internal meetings are such a waste of time.
2:19:27
Now it can be very cumbersome if everyone's calendar is just full.
2:19:29
That's just layers and layers of management.
2:19:33
But is there some world where that's actually the correct way to be running a business of this scale in this era?
2:19:41
>> I think bringing people along is an important function of every company.
2:19:48
>> I wish I could tell you that uh we've solved this information exchange problem perfectly.
2:19:54
There are a number of things that we are working on.
2:19:57
Uh we have an internal enterprise brain project like many other companies do in order to capture all of the relevant knowledge about the different things that are happening within the company so that the right piece of information gets to the right person.
2:20:14
>> So I do think of projects like that as great greatly enhancing internal commu communication.
2:20:19
Remember in a regular company people go to meetings so that they don't feel left out.
2:20:24
Um which is something that I really discourage people from doing.
2:20:28
My attitude is like if I have nothing to contribute to a meeting I should read the notes.
2:20:33
>> Um I really should not have to go to that meeting.
2:20:35
Uh but I would say it's a work in progress to make sure that we do a good job of transmitting internal information.
2:20:42
It also has more profound implications.
2:20:44
I think uh things like how many reports a manager should have needs to be dramatically different in the age of AI.
2:20:52
Of course, making sure that people feel happy and motivated about their work, that's always going to be the role of a good manager.
2:20:57
U but you know, information about what exactly get you done did you get done last week.
2:21:05
That's the kind of stuff that AI can facilitate a lot. It's a work in progress.
2:21:09
I truly hope, you know, I've not measured it recently.
2:21:10
You bring up a good point. We should go look at it.
2:21:12
Um I hope you're not spending more time with uh internal meetings.
2:21:16
Um but we should also not pretend that communication is a solved problem even in the world of AI. >> Yeah.
2:21:25
>> How are you thinking about your own internal AI spend and budgeting heading into next year?
2:21:30
I think uh a lot of you know companies planned uh spent a bunch of time planning next year and then there was a capability jump and they blew through their budgets more quickly.
2:21:41
I think as a leader you have to assume there's going to be capabilities jumps which may mean that you spend more but at the same time there's a lot of uh drive for efficiency cheaper models open source etc.
2:21:52
But how are you planning around uh token spend you know looking forward >> my take overall is that uh the money that we are spending on AI tokens is well worth the cost we continuously optimize absolutely we don't tell people to token max or do dumb things like that it is about driving real results impact as uh as it were and uh when our cost does go up.
2:22:22
We have a good team that focuses on optimization.
2:22:24
Everything from what's the default model uh to can you create task graphs where the simpler aspects of solving a problem are handled by models that are not quite as uh expensive.
2:22:39
>> And my take is that we gain a lot by having people embrace the technology and feel like it can make a real difference to their job.
2:22:47
I feel pretty good about optimizing.
2:22:49
We also practice what we preach uh where we have things like the AI gateway that's meant to make model access more efficient.
2:22:57
We are absolutely experimenting with openweight models because they can be a good vector for uh lowering costs because we also control the harness uh which is Coco a lot of our internal teams use.
2:23:09
There's a lot of instrumentation that we have for what exactly are people doing?
2:23:14
How can we come up with better ways for doing the same thing?
2:23:18
our support and our SR teams, the folks that keep Snowflake up, they spend a lot of money on tokens.
2:23:23
But on the other hand, if they are running through tens of thousands of alerts that are coming in every day, makes sense to go there and optimize because that's very leveraged work that can benefit everybody.
2:23:37
So, you know, yes, we are planning for it, but AI token cost is not the top thing on my mind.
2:23:42
creating great products, getting our customers to adopt them, having the framework by which these things become more self-correcting.
2:23:50
That's kind of how I think about it.
2:23:53
>> Based on all your experience in the enterprise, how do you think the model routing landscape will evolve?
2:23:58
This seems like something that uh a lot of big companies are excited about, you know, getting a slice of that market.
2:24:08
You saw Stripes deal with Open Router, focusing more on developers.
2:24:10
RAMP, our partner, has a product.
2:24:12
Everybody wants >> uh to be in the in the token flow, but how do you think that uh it'll evolve?
2:24:22
>> That's not the highest value creation point for Snowflake as a company. >> Mhm.
2:24:27
>> We have products like co-work >> that can literally get deployed to every employee in a company.
2:24:32
So we spend a lot of time thinking about what does it mean for an enterprise sales team to be AI build to be operating at the edge of what is possible with uh AI and we are seeing deployments of uh co-work go to thousands of users within within companies.
2:24:51
It's operating at a much higher level than simply model routing.
2:24:57
I think it's a good infrastructure capability, but I focus a lot more on things like how do we get every data engineer within a company to be using Koko, our builder product or how do we get large fractions of employees within companies to be using co-work.
2:25:10
what are high leverage, high value projects that we could be doing for customers because our customers have always trusted Snowflake with all of their most important data and focusing on how we can drive real business outcomes is um is where a lot of our energy is at. Absolutely.
2:25:27
The gateway is a good product and it's not just model routing.
2:25:31
We are also offering things like access to tools, MCP tools as it were that uh provide governed access to lots of different applications within uh an enterprise.
2:25:40
Um we are also actively experimenting with is there a security solution around agent trajectories to make sure that people are not misusing models.
2:25:49
So there's a a slew of these things that are there at the infrastructure layer but I think our bigger prize is in delivering value for our customers. >> Last question.
2:25:59
There's a lot of M&A news.
2:26:01
Obviously it's a very exciting time.
2:26:04
You spent 15 years at Google very inquisitive company.
2:26:06
How do you think about M&A?
2:26:09
What makes for a successful acquisition?
2:26:12
How is it unique at Snowflake when you consider it, when you don't?
2:26:20
>> Our strength is as a data platform. >> Mhm.
2:26:23
>> People trust us with their most important data.
2:26:24
People trust us with helping them get insights and drive actions from their most important data.
2:26:31
My primary lens is will a company being part of Snowflake accelerate that mission. Mhm.
2:26:39
>> We are thrilled that we bought NATO >> because MCP is increasingly really, really important for Snowflake and all of our customers because it provides the real-time context of everything that's happening in your my life whether it's Slack or email right in the harness and that was a great acquisition to make.
2:26:59
That's the kind of lens that we bring which is how does something being part of Snowflake accelerate both the company that we buy but also the larger mission of Snowflake as the AI and data platform for every enterprise that there is.
2:27:14
>> Thank you so much for coming on.
2:27:16
>> Yeah, great update, great quarter.
2:27:18
>> Congratulations on the quarter.
2:27:19
>> Yeah, congratulations.
2:27:19
Can't wait to talk to you again. >> Thanks for having me. >> Have a great day. Goodbye. >> Take care.
2:27:25
[applause] Didn't get a chance to hit the gong, but very gongworthy quarter.
2:27:29
>> Uh, huge update in artificial intelligence world.
2:27:32
Someone has created a product that speeds up Lex Freiedman when he asks questions on his podcast. You saw this? >> No way. >> Lex Speedman. com.
2:27:43
Mario Dion says, "I wanted to watch the Lex Freedman episode with DHH, but Lex talks way too slow for me, so I built Lex Speedman.
2:27:49
It speeds up Lex to 2x speed and keeps the guest at 1x, cutting up to 20% of the episode.
2:27:56
Wait, so he so Lex will talk for 40% of his episodes? That's not possible.
2:28:03
I feel like Lex is like he asks one question, he really lets the person talk.
2:28:06
That's kind of a benefit of >> he does. >> He talks for a while. >> Kind of.
2:28:11
I wonder we should pull some stats on like on like who who talks more because sometimes we have good back and forth and we're both we're all talking with the guest.
2:28:20
Sometimes it's just us short question.
2:28:22
They go on a rant for a couple minutes, you know. >> Yeah.
2:28:26
Sometimes we'll get some very angry very angry comments around >> about what?
2:28:30
>> Cutting each other off.
2:28:31
>> Would you please let the guests talk? >> Oh, they say that. Interesting.
2:28:34
I remember early on they were saying you guys got to stop cutting each other off.
2:28:39
And we do we ever was I cutting you off or were you cutting me off?
2:28:42
I feel like this is just the way we talk and like it never >> My real fans went to the rescue.
2:28:48
>> Yeah, we never really got to the bottom of it.
2:28:50
I think it was me cutting you off sometimes, but uh especially with the ads.
2:28:54
I love cutting you off with an ad. That's the best. >> That's the best.
2:28:57
>> But also no dead air, you know.
2:28:59
>> I love cutting you off with a flashbang. >> No, don't do it. Don't do it. Flashbang.
2:29:04
We're going to the stars. Astra GPT6 is out. Go check it out.
2:29:07
There's a crazy video by Matt Schumer.
2:29:10
Uh he oneshotted a game in Unreal Engine.
2:29:12
Now Unreal Engine has MCP so you can actually build the game of your dreams basically just by chatting with the model.
2:29:20
>> And Mike over at Arc is >> moving the goal. Moving the gold post. >> Moving the goal. They're moving. Thank you. Let's do it.
2:29:26
>> Thank you for not waiting to move the goalpost either.
2:29:28
You know, >> he said Astra is is the new state-of-the-art on ART AGI 3.
2:29:33
It's a qualitatively large leap towards AGI and the pace of progress is frankly surprising.
2:29:39
That said, we lack evidence to call this AGI yet.
2:29:42
[laughter] >> While we are still studying the human capability gaps, we believe open-ended invention is unsolved and this will form the new basis for ARC AGI4.
2:29:57
So now it's like you got to invent new physics, new science, you got to go to that.
2:30:01
who actually Astra has to actually go to the stars.
2:30:06
>> Yeah, he really said, "What have you done for me lately?" I love it. I love it.
2:30:10
It's It's AGI when Mike says it's AGI. Uh it's a good time.
2:30:12
Uh anyway, very interesting.
2:30:15
You can dig into his post.
2:30:17
He gives a lot more context there about uh Arc AGI V3 benchmarks, Astra.
2:30:24
You can go check it all out.
2:30:24
And of course, we'll be discussing it tomorrow.
2:30:27
We have a bunch of special guests. So, have a great day. We'll see you tomorrow.
2:30:31
or leave us five stars on Apple Podcast and Spotify.
2:30:32
Sign up for a newsletter at tvpn.
2:30:33
com and we will see you tomorrow.