Anthropic’s IPO Leak, OpenAI’s Dots vs. Meta’s Muse, Visual Turing Test

0:00 Anthropic's IPO numbers leak. Are they good enough to save the economy? OpenAI has a new always on agent called DOTS. And have we passed the visual touring test? That's coming up on a Big Technology Podcast Friday edition right after this. Welcome to Big Technology Podcast Friday edition where we break down the news in our traditional coolheaded and nuanced format. We have a great show for you today. The long- aaited anthropic preIPO numbers are out courtesy of a leaked report in Reuters.

0:28 We're going to go over the numbers and talk about whether they're good enough to keep this economy running. We also have a new always on agent from OpenAI called DOTS. I've tried it. I'm going to give my reflections and why I'm sending it a ton of work already. And then finally, we're going to talk about Tavis and the visual touring test. Yes, there are some bots now that can completely fool you into thinking they're human on a video call.

0:51 Joining us as always on Fridays to do it is Ron John Roy of Margins. Ran John, great to see you. >> Good to see you. We're getting so much closer to the real S1. We're getting these tidbits and we're going to dig into it. But I mean, did you see they're going to I think they announced investor day is supposed to be October 14th and that means we're on track for a pre-Thanksgiving anthropic IPO. >> Exactly. Yeah, it's coming in November.

1:17 I think there's no doubt it comes this November. And as you mentioned, we got a leak. So, Reuters did scoop basically the entire S1. They didn't print it out for us, but they gave us all the numbers, that Anthropic is taking to Wall Street. And we finally have some like binding numbers, not ARR. and I'm actually curious how you read them because there are some good numbers in there. And there are some numbers in there that make you just kind of shake your head and ask, how is this going to work? So, if if you will, let me give you a couple of them to start off with. So the good news is the company's revenue is for if you're anthropic the good news is the company's revenue is growing exponentially right 400 million in 2024 4.6 billion total in 2025 now at the end of the year they said they had a 10 billion ARR but like if you have a 1 billion ARR in January and 10 in December 4.6 is kind of where you would imagine things net out the crazy number to me is on that 4.6 6 billion in revenue that they reported they said they told investors they plan to spend 518 billion over the next decade on computing and infrastructure and of that 80% of that spend 80% of the spend is binding. So you're looking at a company that just basically made in the mo in the most recent year that it can report $4 billion that has committed $400 billion and has no way of getting out of it. These are the numbers. Your reaction? My reaction is I was a little disappointed not in the numbers which we will get into is that the leak doesn't have the H1 2026 the first half of the year financials when they're they're definitely going to need to be filing and actually publicly stating some amount of 2026 financials and revenue numbers and loss numbers which I'm sure they'll likely be. But so so we have a good look at what Anthropic looked like at the end of 2025. And again, 4.6 billion in revenue. I think what was the net loss was around for reference. There's this $42 billion number that's being thrown around as a net loss. And I say this someone who probably is going to be skeptical or dig into a lot of their numbers, but it's not actually 42 billion. $34 billion of it is it's a weird accounting treatment issue related to the company actually growing in value a lot, but the net loss is still around $8 billion. They're still losing around $2 to actually make every single dollar that they're making.

3:58 So to me, in 2025, they were losing a lot of money. What does it look like in 2026? There's all these like little leaks that they were supposedly profitable back in May and June. I don't know. I I want to see that in the documents. >> Yeah. And and also remember there was talk of a hundred billion dollar run rate or or hundred billion dollars in 2026 like revenue. That's where that ARR and all these kind of like reported numbers were going around. And when you think about it to get to that $2 trillion valuation, you would think maybe you do start to need to see something up in that direction. So, so I guess let's get into the compute commitments in just a bit, but I want to know where do you think their revenue is going to end up? What kind of numbers do you think they're going to show about the current state of the business as opposed to 2025? Cuz come on, who remembers 2025 so long ago?

4:54 >> No, it's very important because you know, if the that number doesn't look like it's continuing to go exponential, a lot of these big numbers become very scary. So, for instance, Anthropic recently, I I think it was Anthropic, you know, told the New York Times, I mean, we we obviously it's like so sources close to the matter, right? was it Anthropic? Was it one of their investors? Probably one or the other. Told the New York Times that it was going to be it was going to be at 100 billion annualized revenue at the end of the year. By the end of the year.

5:25 So, you know, you would assume that this year they're going to hit, you know, somewhere between 50 and 100 billion when all is said and done. you know, if they can do that, then spending 500 billion is less daunting than if they're at 10. but we don't know. And I think you're right. When we end up seeing those numbers, it's going to matter a lot for most of the year. Remember, they've been at something like 45 to 60 billion AR, at least that's what we've been told, right? So, I think they're definitely growing, but the rate that they're growing is going to matter a lot. And this is something that Daario's kind of addressed, right? He's talked about this with Dwarkash for instance. He talked about it with Andrew Ross Orcin. I think he and I had a bit of a conversation about it when I was there last year which is that the company has to make these projections of how much compute it's going to need years out. And there's one risk which is that you buy too little and you can't serve the use case and your competitor has a lot which is the story that's been been told you know recently because OpenAI seems to have more compute and is going yolo in the words of Daario and able to meet that need. But the other thing is if your revenue isn't able to grow to meet your commitments you know careful or not if you cannot meet these commitments and 80% of them are are not canceable your company goes bankrupt and that is the state of anthropic and you know I wrote about it today in big technology and I called it a leap of faith that anthropic is taking everybody on and it really is this leap of faith because the economy today is dependent on anthropic and Open AAI meeting these commitments and if they're not able to you know not only is this going to be bad news for Anthropic which will go bankrupt if they can't meet these obligations but there is going to be a cascade because the amount of money that they've pledged to companies like Alphabet like Amazon and like Microsoft not to mention Nvidia is massive and factors into these companies projections. I do agree it will these are massive commitments and like will they be able to pay them and like what does that actually mean for their financial future but I don't think it means they go bankrupt like like actually if we outline how this plays out even again 80% of that overall compute commit is non-cancellable but in reality if this starts to have some kind of cascading effect do you do do you not think the lawyers are going to get involved and everyone's going to realize and anthropic. This is again too big to fail type stuff relative to the Broadcom, the Alphabet, the Amazon, the Microsoft, the people who they have the com commitment to also have a very strong vested interest that this does not all collapse. So do you think they're they would like drive anthropic into bankruptcy just to that's the because that's the nature of the original contract.

8:15 >> Well, okay. So that was Dar Daario's words. Maybe I was, you know. >> Wait, did he say yolo? >> No. yo, he said yolo as AI is yoloing. Yes, >> he actually used yolo as a verb. >> He said there are others that are yoloing and we're trying to be careful, right? and but but he, you know, he used like basically these terms that yeah, we're going to go bankrupt if we can't because we're make signing binding obligations. Look, of course, if you're Amazon, which you have a big investment in anthropic, right? A lot of these, this is, by the way, it goes back to these discussions we've been having the entire time, the circular financing, right? The investments, the the compute deals, right? It's all tied in together. So, if Anthropic goes bad, yeah, I don't think Amazon's going to let them you know, go bankrupt. but I it will be irrevocably bad for Amazon and for Anthropic. So, listen to these numbers. the commitment the spending commitments that that Anthropic has to Alphabet which is a shareholder is $111 billion over the next decade. To Amazon it's $ 110 billion and to Microsoft it's 31.4 billion. You remember seeing like the these earnings numbers of these cloud businesses. now of course that's spread over time but we see these earnings numbers for these cloud businesses and a part of it is selling direct to their customers. But a lot of the the growth that we're seeing is coming from their dealings with the open AIS and anthropics and and the health of these companies. Like PreAI for instance, AWS was growing at like 17% consistently a year, which like on a big base is nice, but the company could not increase its growth rate. that has boomed. It's in like the 30 range, 30% range, which is just crazy given it scale. So, of course, if Amazon saw Anthropic doing poorly, you know, it would try to make a deal to save the company, save its, you know, save its ability to continue to sell its technology, save its investment in the company. Same with Alphabet. I don't doubt it. So, yeah, bankruptcy might not happen, but it might not take a bankruptcy to have a cascading issue in our economy. And this is what I really want to speak with you, Ron, John, because you've you've been on the trading floor. You've seen economic collapses. I don't think we're heading towards an economic cl economic collapse, but I think we have to talk a little bit about the dependence of the broader market on the financial performance of the anthropic IPO of the OpenAI IPO and the companies more broadly. So, I wrote about this in big technology today, Anthropics IPO and our collective leap of faith. today's public markets and much of the economy depend on AI's continued growth. The S&P 500 is currently experiencing its worst breadth since the dotcom bubble with the median stock trading 16% below its 52- week high even as the index is up around 13% this year. Nvidia has been holding it up adding more than a trillion in market cap in 2026. And half of the S&P 500's earning per share growth this year is coming from AI investment according to Goldman Sachs.

11:15 Steve Ratner put a chart up talking about this. you know, the S&P 500 is near a record, he says. but if you strip out the AI companies, it's down 5% since late August, right? So, this is kind of what it comes down to. Anthropic has made these big commitments and we finally got a chance to see the number with this S1 and it has to make good on them because we have this like crazy breadth in our market where many of the companies in the S&P 500 are down significantly off their highs, right?

11:47 you're not seeing any growth from them. The growth is concentrated again in big tech mostly in the AI trade in companies like Nvidia and in these you know big tech companies and should the anthropic story or the open AI story for that matter begin to wobble you could see a cascade here. >> Okay. So let's think about how this plays out because again I agree I think we're actually agreed that like these the magnitude of these commit compute commits to the Amazon's metas and hyperscalers actually present as much risk to the hyperscalers as they do to anthropic in the way that it's actually if anthropic can't pay that's really bad for Meta Google especially Broadcom Nvidia and these others so so I think they need to be worried about this. They need to kind of start trying to understand I don't know like there is so much energy behind this trade. there's so much vested interest in capital that you know everyone has this interest and so much of the energy around marketing AI and pushing AI and I say this as someone who strongly believes that it's going to transform everything over time but like all of it like has to be done by this especially kind of concentrated group of companies in order to keep this trade propped up like I think we should get into are there kind of like cascading effects is there actual kind like largecale economic turmoil. But I think to start, what does it actually look like in terms of anthropic? Do we think them having an S1 that does not excite the market is the trigger? Do we think it's 6 months from now growth slows, growth is flat, commit, everyone starts to question compute commits, which then has a follow-on worry around Amazon's cloud earnings, etc.

13:46 Like how do you see it playing out? >> You know what? >> Assuming Anthropic does not have >> just ridiculously good numbers that just saves everyone and we're all okay. >> I think it could be. I mean, here's the thing. SpaceX set the precedent. Now, of course, SpaceX is a very different company. It's not pure play AI. but it went out very strong, of course, and has kind of had some trouble since even though it is today still trading above its IPO price and hasn't led to some sort of collapse. But anthropic has some structural risk and the enthusiasm that investors showed around SpaceX. I don't know if you're going to see the same enthusiasm around anthropic especially given all the other risks that we know exist for anthropic which is basically it's frontier models are competing with standard models that are doing a fairly good job and companies are starting to say I'm going to mal route and I'm going to go off the frontier and be cheaper. it has a customer base right now that is trying to spend less on tokens. than they have in the past and there's all these government and like the worldending headwinds that like some might say are marketing but they're a variable and you don't want a variable when you go out but look the fundamental thing here among everything that I think investors are going to look at and they'll have to make an option for them.

15:02 they'll have to make a decision which is basically do you want a a stake in a company that may end up you know creating some worldchanging technology or do you look at the picture and the picture shows big losses big spending coming up you have spending that you're locked into again 80% of that 518 million is commitments that are uncancellable whatever that means right billions billion sorry billion and uncancellable commitments >> and while you can't cancel your customers And so these are two other things that came up in the risk factors.

15:35 Nearly a quarter of Anthropics revenue came from two customers last year. Yes. One quarter of the revenue came from two customers last year. And are Anthropics customers locked in in a world where everybody seems to be developing AI models? No. Again, from the risk factors, many of the company's largest clients were not locked into long-term contracts and could cut or stop spending with them basically at will. again, shout out to Reuters and Echo for, you know, scooping all this. These numbers are fascinating. But this this contrast of being locked into spending while your customers are not and having such concentration in a couple of customers.

16:18 >> I don't know if investors are going to run towards a company that that has those dynamics. >> It doesn't look good. I mean at that level obviously growth rate over the last couple of years is like unprecedented and like human history basically at that kind of level of growth in terms of like revenue but but I agree there's all these clear and present risks that are dramatic. Again that concentration of customers in any other kind of company would be terrifying. And it's funny because SpaceX, to Elon's credit, I mean, they were at $18.7 billion in revenue and losing money, not at the scale of anthropic, but still losing 4 billion on the 18. And they came out at over a trillion and got to two trillion for a bit. So maybe it's not unreasonable from that perspective, but but I think yeah, the market, everyone's mentality is in a very different place today. and this in early October versus where it was a few months ago. So I think people will look carefully at this S1. But I guess what would be the most dangerous red flag for you? The compute commits I guess I think I differ from you. I don't think that is as glaring and shocking an issue relative to the actual core state of the business. So the concentration of customers losses if we see slowing growth that is actually the greatest risk of all. So like those are the biggest risks. Compute commits I I still that plays out on such a longer time scale. I feel that ends up in kind of like corporate lawyers hashing it out and everyone figures that out like what what do you think is the biggest risk?

18:01 >> I I think don't you think the market the market is really tenuous right now? Of course, we've seen like the bond market, you know, show some worrying signs and you just have all these investors concentrated in AI and these AI adjacent companies and I've continued and I'm going to answer your question, but let me just say I've continued to see these tweets show up on my ex timeline that like make me think, huh, maybe that's a problem. This is something from Citadel. Microsoft, Nvidia, Apple, and Meta alone added about 300 points to the S&P in Q3, more than 200% of the index's total gain. The other 49 499ish stocks together subtracted about 150 points. This is from Peter Schiff. The S&P 500 is 7% below its high. Yet 430 of the stocks are 27.1% below its their highs. That means on average 86% of stocks are in a bare market. breadth has only been this bad twice in January 1973 and in 1992 2000 on both occasions. The S&P then crashed nearly 50%. You know, am I wrong in thinking that the This might be a hyperbole, but I'm going to throw it out there. the market is basically being held up by this AI story. It's a narrative, right? It's a narrative that and I think it's a strong narrative that companies like Anthropic and OpenAI are they're just on to something, right?

19:26 They have found this new technology. It is going to change the world. It's going to change the world fast and it's going to be very profitable in and maybe not too long of a time horizon. if that story crumbles, it might be what you said, which and I think that's as good of a reason as any, right? Or, you know, I was listening to some investors this week and they were like, "Listen, this idea that 25% is two customers." We know like there was that rumor earlier this year that one customer, you know, spent like what 500 >> token maxing in a day or whatever million in a day or something like that and then was like, "Oh, we should put controls." We know businesses everywhere are putting controls on. We probably won't see a momentum slow in whatever numbers they present. that would be bad. But like any little bit of wobbliness could start to I'm not going to say bring the stock market down because that's ridiculous but could make the stock market sort of revert to where I think it's been trying to point which is like down. You know S&P is up 13% this year because of these companies.

20:26 You might end up seeing it give up those gains and go down five or 10% because people stop believing in the story. >> Five or 10%. I think if this if this plays out, I'm not saying 50% but Oh, sorry. On the year. So, that would be like a 20% drop from where they are now. Not not I I know you don't like this, but I have to say it. It's not investment advice. We're just having a conversation here, but go ahead John.

20:51 >> Not in advice. I'm not trading off of this either because the irrationality of the market has gone on long enough. And also, as you said, there is a real there's a real story here, but there's like separating that out from the actual short-term narrative is important to me. And and actually that that that stat that 86% of the stocks in the S&P 500 are in a bare market, that they're actually like the vast majority of the market. And and if you you think about what that actually means from like an actual market mechanic standpoint, what has been happening and I'm sure many of our listeners feel this and I have felt this myself of like if everything you're if you were invested outside of this concentrated trade, whether it's on the actual semi side or whether it's the actual tech companies themselves, you're not just missing out, you're doing badly on any of those stocks. So everyone starts rotating into them more. So there's more momentum behind them and it that has definitely been happening that if you're not in Nvidia, why are you investing? if you're not in Broadcom or any of these other stocks and like versus trying to find what are good companies in other sectors of the market or even within technology itself and trying to actually understand are they good businesses and and again that is not the way the market's been trading and any kind of switch in that mentality I think will be pretty dramatic and I do think the S1 is like I mean I'm trying not to go full doomer because I still don't think this is like economic calamity or collapse. I think there's but I do think that simple narrative breaks suddenly rich valuations incredibly rich valuations for a very small group of companies starts to be questioned and then people actually start to either pull out or move it back to other companies that have been getting hammered over the last year. So I think like I think the S1 it better be good. It better be >> I mean that's we've just seen it. We I I growth rate give that we haven't seen it. We haven't seen it.

23:12 >> You think they're going to add before before they they go public they're going to add these first half numbers. >> They will definitely go to the market. Yeah. Again, I like this is the first thing and I kept like generally any kind of S1, you're going to have to show some relatively recent. Again, you're not supposed to make forward-looking projections, but you should give an accurate reflection of where the business is today. So, if you're going for a $2 trillion valuation, you can't throw out $4.5 billion as the last reflection of what the economics of your business are and then expect that the market, again, that's how things normally work. I don't know what Anthropic is going to be allowed to do. Maybe they will, but but they're going to come out with more.

23:56 It's going to be a lot more. We're going to need an emergency pod when that thing comes out cuz >> Okay. >> It's going to be it's going to be juicy. >> We had people in my mentions, I think yours, too, is saying, "Hey, do the do the emergency pod when the Reuters story came out." But I think let's wait till the actual document comes out and we can read through it for everybody live on YouTube. And by the way, you said, you know, you you hinted at it, but this is obviously why Trump was with all the AI CEOs in Washington this year.

24:24 He's basically just like >> think about the way >> Yeah. >> Yeah. Well, he's the he's the president. He wants the stock market to do well. And he's basically just like, if you guys do any more of this dumerism or slowdown stuff, the whole market goes down. And that reflects poorly on my presidency. Jensen, you're going to sit next to me. Daario, you're going to sit so far down the table. I'm not even going to be able to see you. No, no more doomerism. We're going to keep building because if we stop, you know, it's not just again like to put it to put it clearly, it's not just an open AAI or anthropic story. It's a market story more broadly. It is a market story. I mean, it definitely is going to be a larger market story. Do do you think we have to wait for OpenAI? I mean, if you saw I'm sure you saw they're raising another 30 billion now. So like they're definitely not going out in the next in the immediate vicinity but like do you think it do you think anthropic is going to give us clear direction on where this is all going?

25:25 >> Yes. I mean I've always thought as anthropic goes so goes the AI trade and the AI but because they have bet so strongly on pure play technology even though they have some products now I I've always thought as anthropic goes so goes AI. So that is that is I think where where we're going to get a lot of clarity around whether a the market wants this stuff, whether it believes in it, whether this entire narrative can continue to be held up and the momentum can be held up or not.

25:56 >> So So if you are Mark Zuckerberg or be Mark Zuckerberg for a moment, >> is >> cracking jokes with the president behind Daario's back as he tries to speak. Get back on the hydro press. Come on. Just get on the foil with the American flag. We'll all love it. No, just don't don't do that. But okay. >> throw it at me. >> Yeah. Okay. So, there's weird competitive dynamics here because Anthropic has spending commits to you.

26:27 Anthropic is kind of like the basis of the narrative. Again, Muse has been a smash hit. like Meta's back in the game more than any of us ever would have given them credit for a even a few months ago, much less a year ago. But are you trying to kneecap your competitors? They are, you know, like the open- source openweight conversation for a while. I mean, Meta's been kind of taking this approach, but like by going super aggressively at your competitors, like do you go in with men?

27:02 We're going to talk about Meta's enterprise launch. Do you just undercut price just massively in an effort to kind of just smash anthropic and or or open AI knowing that the following effects could actually severely impact your stock price in the short term? No, definitely not a reflection of like longerterm business strategy, but do you are you cold-blooded enough that you you you take that approach? I I mean I appreciate the thought experiment that like Mark Zuckerberg right now is the arbiter of who was going to win and lose in AI. I just don't >> He's one of them. All of Sundar is too.

27:40 >> He's been he's been trying. I mean he's been trying and we're going to get into Muse versus Dots in a moment. But Zuck has been trying for you know since the AI race really began to undercut these companies and it hasn't worked. they've only, you know, continued to do. I mean, Anthropic, if it goes out, you know, even at a disappointing valuation, is going to be a bigger company than Meta right off the bat. So, yeah, I think Zuckerberg will try. Remember, he's got compute, he has cash, so he'll try to do whatever he can to get people to his service versus theirs. But here's the difference, and I think we're gonna talk about this with DOT. when it comes to enterprise, like Muse is a good app for consumers. when it comes to enterprise, you really want the best intelligence. You want the frontier intelligence. So that's why I don't think this meta for enterprise thing, which Meta is introducing, right? So folks, they this week announced that they're going to be doing this Muse for enterprise, I don't see it being extremely competitive to anthropic and open AAI, though I don't give Meta a hard time for trying. Remember, they have good data. They have relationships with all these big companies. They understand what drives their business.

28:50 and it makes sense to go after them. >> See, I should we dig into Meta for Enterprise? >> Yes. But hold on. We should do We should do it in this order. We should do We should talk about Dots, >> then talk about the competition with Muse, and then talk about Meta for Enterprise. >> Yeah, let's do real time run of show here for listeners. >> So, we're going to do it after the break, but I need to do two things before the break. First of all, Ron, I got to tell you, this was a absolute fumble the ball week for me.

29:19 because today I just discovered that I was invited to the revealing of America.gov at the White House. and I missed the message. >> You could have been up there next to >> I could I would have been in the in the gaggle or something. >> No, no, no. You would have been right right next to Daario. Right next to Daario. >> You know, did you see how Daario went back and hid behind everybody and Trump still called him out? If I was anywhere close to there, I would be hiding even further behind than Dart did. Trust me.

29:52 Okay. Second thing I want to say is just a a mistake I want to acknowledge on my part. Last week I talked about how Muse should be rolled out to is being promoted to Meta's 3 billion users. Of course, it's not available to the three billion users. It's just available to a small segment of the population mostly in the United States. And so, you know, the looking at the Meta's Metamus's growth, I should have used that as the baseline, not all of Meta's users, regret the error. Muses grow nicely, but now it has real competition. So, thank you to those who've pointed that out, including podcast enjoyer and a bunch of others.

30:32 I always appreciate feedback from listeners. We will get into the Muse versus OpenAI dots competition when we come back right after this. And we're back here on big technology podcast Friday edition with Ranjan Roy of Margins. Ranjan big week this week in terms of a always on AI assistants. They can't just be assistants they have to always be on and that is with the introduction of dots from open AI. I'll just introduce it. This is from the New York Times. A new AI agent enters the marketplace. Open AI's dots. OpenAI now has artificial intelligence agents that act as personal digital assistants announced at the company's annual developer conference on Tuesday in San Francisco. The agents called Dots will compete with Meta's AI agent called Muse which was also released this month.

31:22 Google's Gemini Spark which came out in May and Instinct an AI agent startup that raised 1 billion in its latest fund funding round. Where's Grockbot in New York Times? Hm. And we should also say where's Grockbot in our discussions? Well, I've mentioned it folks. Grockbot belongs in that paragraph. Okay. Dots are OpenAI's most aggressive push to create a product out of AI agents which will do things like booking restaurant reservations or managing finance finances. Users can message their agent which they can name in chat GPT Slack or Microsoft Teams or through text and connect to thirdparty or through text and connect to thirdparty apps. They can also talk to the agent via phone call. The agents will be available starting Tuesday to subscribers of Chat GPT's pro plans which start at $100 a pop. shaky debut I would say for DOTS right they did not nail the demo as they were struggling with the demo. Meta's stock went you know shooting up an indication the market believes that these products are directly competitive and if OpenAI stumbles it's Meta's bounty. So what do you think about the release of DOT Ranjan? I think in terms of the stumble, did you also see Sarah Frier on CNBC accidentally called her own product Muse and had to correct herself to saying it's dots? Okay. I have not tried it yet. I believe you have. Is that correct?

32:46 >> Yes. So, I will give you my review now. >> Tell me my Tell me yours. >> Amazing. It's amazing. it is. So, I decided, you know what? I'm gonna go all the way in now and and give it full control of shouldn't even be saying this, but give it full control of my laptop and my browser and you know have it in chat GPT. and it's incredible. I mean, you can just keep messaging it and it just will pick up it will basically figure out what you want it to do and pick up your work and start sequencing it. And you know, especially when you have a lot of tasks piled up, and I'll admit or I will say it's perfect for me because like in my job in particular, I have so many things that pile up and I also want to make time for content creation which like requires like deep work and you know moving away from the inbox. So like things can like you know sort of just build up to the point where I have to just get things done very quickly. Lots of emails, lots of invoices, you know, trip planning, all this stuff. And so last night I just sat there and I told it well again like I I don't know how much of this I should be sharing but for the sake of transparency I'm gonna do >> for for the listeners. Come on.

33:59 >> I told her read read like as many of my recent emails and articles as you can and pick up my writing style. And then I started going through my email and you know you know you know you know when you open an an an email and you know it's like you know exactly what you want to say but it's like phrasing it in that like email way where it comes off as like direct but not offensive. You know what I'm saying? is takes like 90% of the energy. this thing got my voice right away and started like drafting email drafts and you know I so the action I had was like you know the intended action but it starts piling up these email drafts you know I'm just like write write one of these like write this write this and it comes it comes in the draft so I can go in my drafts alter it a bit and then send and just very quickly clear out the inbox. Then I said all right you know I need to send an invoice. So, it managed to like and I said, "My invoice is in drive. I need you to go into drive, you know, download the invoice, put it, in Gmail in a draft to the people I usually send it to and write my note. and it was able to do that fairly effectively.

35:10 Then I was like, okay, I know my are we debate the travel eval? >> just I'm I'm I'm in. I'm in. I give up. I give up. I was I was like I am just going to delegate everything to the DOT here and I have a weekend trip to El Salvador coming up and I was I was like these are the places we want to go. my hotel budget is $100 a night cumulatively. So you can go a little bit above and a little little bit beneath depending on the night. you know find me the rental car, find me the hotels, book the hotels, and then put together an itinerary that we could follow when we're there. You let it book and I well it got it basically got got into booking and needed me to just hit like the final step. So I chose the hotels that it chose. I hit the book to complete the thing. Then it realized >> Wait, wait, hold on. No, no. I want to push you. How many options did it give you or you literally narrowed it down to one and went with it?

36:11 >> okay. So, so two of the three I liked. >> okay. There's still human in the loop. Human. >> I got I got in that fully automated. >> I want I want you in El Salvador in a dot picture hotel. >> Listen to this. It realized that one of the hotels had some like off-roading before you get there. So, it's like, I think this will work, but I'm not a 100% sure. And it's like, should I just send emails to budget, like the rental car agency, and the hotel, seeing if the car can make it there? And I said yes. And then I just like looked in my sent messages and perfectly drafted in my voice emails have been sent to these folks. So, I just I really went ham on it. And I think because it's built for for enterprise, because it's built with Frontier Intelligence, it really can just bang out a lot of work for you. And I have to say, I'm impressed. And compared to that and Muse, you know, Muse is really good for some consumer things, but I almost thought like, man, I I should just delete Muse. I can't believe I'm saying it, but I felt like deleting it because I thought the experience was so good. Well, so this is where and again I think we should differentiate because like what you're describing is not enterprise in that sense like and again I can dig >> the invoicing is the emails are >> okay fair still it's work it's not enterprise it's work it's not large scale in terms of like privacy security >> no no enterprise business an enterprise business is selling for work clients versus consumer clients >> no no but but Wait, when any of these companies are talk, again, this is my world, like what I have to deal with every day. When these companies are saying enterprise, they're not talking about SMB across small and mediumsiz business across which I actually think meta should go after and not and we can get into the likelihood of them actually nailing the enterprise but like large complex organization security that invoice who does it have to get approved like that's the still the kind of stuff for enterprise versus individual kind of has full full say in management over some set of work that's being done and then that that's the end of the responsibility. So I'm ju just I'm gonna push back on that side because I I actually think it is important in terms of like where dots and this kind of work and these kind of agents are actually going to be have a lot of potential. But I do think to me the most interesting part of this week and I've not tried dots yet, but I actually went back I didn't even realize Manis who Meta bought in 2024 25 they're one of the earliest that had some kind of like natural language workflow generation for a consumer. They had to unwind that. And Manis is out with their own personal agent called Q, which is kind of like the hipster. I'm gonna say I was playing with that one this week and trying that out. They actually have like a video editor in line on the desktop, which is kind of cool. Main point being all these are already starting to feel interchangeable and it's literally like it's everyone is building the same thing in this type of like product but it I still think it's going to come down to who do you trust. I think it's interesting that you gave Open AI the keys to the car but you didn't give Muse the full keys to the car. Like I gave Instinct my credit card. I didn't give it to Muse, but then I gave Muse my MX login because I was curious for it to like check in on my platinum benefits and it did a good job on that. So, so I think do like how do you think you're going to differentiate as all these products start converging?

39:56 >> Yeah. I mean, I think you're right. It comes down to trust. And I don't really trust any of them, but there are some that I'm willing to tolerate and some that I'm not. And >> what do you mean what's tolerate though? What's tolerate? tolerate means like basically give my my details. I try to give actually I try to give Dots my credit card and it didn't let me which was >> but it actually you know it's it's weird like some of these things like for instance I've tried to get Claude Co-work to sign a contract for me recently. It wouldn't do it. And then I was like so I'm going to be going to to I shouldn't even say this again out loud. I'm going to be going to Websummit. There's like a speaker release that you sign for WebSummit that I've signed a thousand times and I just went to dot and I was like go into my email and sign that that speaker contract and it did.

40:38 >> Is that legally binding websummit? No. >> If I get if I get thrown out of websummit after after >> it'll actually make it'll make no it'll make for a good enough story that it'll be worth it that actually that's the headline right there. I had my agent sign the web release. I want to go to Lisbon. >> All right. All right. I think >> it would be the headline writes itself. My DOT signed a contract. Now I can't go to web I think I'll end up going. I think I'll end up going. I feel good about it. Okay. So, where do you like what is going to be the differentiator then as everyone and Tik Tok is probably going to release one? I've been heard hearing rumors like >> Oh, Tik Tok. Oh, there. I definitely want to give them my birth certificate.

41:22 >> I mean, hey, it's Americanowned now. It's all okay. >> but yeah, >> I generally No, social media companies I don't I do not want to trust with this stuff. Also, I I I do have a bone to pick with you. I think Frontier Intelligence matters. I think you get more if you have a better a smarter model. You don't think so? >> Oh, wait, wait. Sorry. Hold on. I didn't argue on that, but actually on that point.

41:45 >> Yeah. No, actually I I will push back on that because I think for a the majority of the type of tasks you describe do not require frontier intelligence. They don't like it's especially small amounts of do like a small corpus of information. our our any of our inboxes are actually relatively small like volumes of information to process work through creating a voice or a skill that is in line with like that's a pretty simple task for AI too that's that's like GPT4 five maybe 03 maybe we're still in 03 territory so so I think but actually frontier intelligence if they start airing on the side of that or kind of like biasing towards that this is going to get very expensive for these companies like the amount of work that's being done in the to go do especially this always on like nature of work to go be mining your inbox again instinct and I've been loving you for a while but it's like you don't have to try so hard about reminding me of some email is like oh hey by the way >> you need the engagement that's the thing >> I know no muse has been hammering nonstop Oh, >> non-stop. Non-stop.

43:02 >> Guess what you sent me this week. >> You sent me just a you know an a heads up that on a separate note, California AG Rob Bont issued an investigative subpoena to OpenAI yesterday over cyber security risks around its AI models just as a by the way. And I'm like, wait, did I ask you to send me AI news? And it's like, no, that was an overreach. >> Wait, what are you musing about? Are you Are you musing your news? Do you muse your news? I I don't use my news. It must have figured out, you know, that I'm a reporter and started sending me headlines.

43:34 >> Okay. Okay, that's fair. I haven't got any. >> This is its top competitor that it's trying. It's like, hey, by the way, you know, speaking of trust, open AI, cyber security problems, California AG, Muse, >> I mean, >> never a problem. Facebook may be a problem, but Muse is not Facebook. >> No, I I think that thirstiness is definitely going to start to be more of an issue. And actually I wonder are these I mean again Meta has been like you know from the rafters shouting daily active user numbers and I'm sure they're going to be pushed on this level of engagement. So that is a that's going to be an interesting user experiencing cuz like it's going to be such a sensitive balance between giving people something valuable honestly and I said this last week proactivity when it's valuable is like the most magical experience in AI I one of the most I've had to date of like oh by the way I just realized this thing and I think this will be really important to you and when it is that's a gamecher when it is, oh, by the way, like you'd look for a dinner reservation 2 days ago and actually, actually, this happened with with Muse. I'd look for a dinner reservation on a Friday and then on Sunday it went to resi and looked and found that it was now free, which that's nice. That's not what I was looking for.

44:58 So, yeah, as they kind of overreach, it's going to be it's going to be interesting. They got to play it cool. They got to have the Instinct's kind of cool. Instinct has like the way it actually writes and speaks is kind of laid-back. And I actually saw the founder Noah Shin talk about that though that like your agent should not be overly eager and stuff. So it's going to be a fun fun battle. >> I just want to talk about this intelligence thing for a minute. I mean I have found you know you you might say that frontier intelligence doesn't matter but I have found that you know especially with the latest models from open AI and anthropic compared to everything else there's just a level of depth and thoughtfulness that that far exceeds even the previous versions of their models and certainly their competitors.

45:46 >> What does that mean? What does that look like? >> It it means like they seem to understand things better. I think that's the best way to put it. And with writing, okay, like if I put a story draft in, right? It's like so often times I won't have AI write for me. I'm I'm against that. at least, you know, I won't write my news stories or my anything that I write on big technology. but like I will throw like often times I'll throw a story and interviews in and be like what did I miss or a story and different articles in and say what did I miss? what I found now is that it will do a much better job diagnosing what's going on in the story. it will do a much better job taking in different inputs like articles from around the web. clarifying what things mean like doing things that like oh that was really smart as opposed to like okay like that's helpful but still very basic. So I do think the frontier matters. So that's at least what I see on on my side from a personal level. But I'm actually I have a question for you on you know what the frontier can do from an enterprise standpoint. So like what are companies able to do with these next level models when they work on your your definition of enterprise problems that they couldn't do previously?

47:00 >> No, but I mean this is what we've seen firsthand and you see it in every ramp spending report that comes out. The enterprise is moving away from frontier models because from a cost perspective and from the actual jobs to be done or the tasks that need to be done, everyone is recognizing that actually again all those things that you started outlining that were magical to you about dots are not like frontier tasks. They're they're limited sets of information to do fairly I'm not trying to demean the magic of it. I'm saying like they're not they're not requiring we're not like sequencing DNA here. We're not I mean everyone seems to be like building world models on Astra or like you know 3D simulated environments but like email writing understanding again anything that's text understanding that has some finite number of steps of logic and reasoning and by finite I mean like you know some limited number but that still can be tens maybe even hundreds go find this much information pull it in process it that's not frontier anymore And that so that's why to me and again it's been in all the ramp data and it's been a major and it's going to be an anthropic risk to tie it all back together like enterprises are no longer just pouring money only into the frontier it's getting distributed across openw weight models people are building inhouse people >> I get it but I'm I'm asking you what is a frontier task that's what I'm trying to say what are the tasks that frontier intelligence can tackle or do better on >> these but that's where like what are the mo if if you got to sequence some DNA this weekend you got to go frontier if you want to again every like cool use case I've seen people kind of bragging about remember a few like a six months ago the kind of like YouTube hustle bro AI type of thing would be I just automated my life second brain I just sent the kind of things you're describing now everyone again I don't know did you Not everyone seemed to just be like building virtual worlds and video games and open world. No, no, but that's what like Astra is. Look what I just did. this super complex graphic with stuff going everywhere and it's that's all I've seen that like looks genuinely interesting like I versus something that can't be done with something that's close to a frontier model which is why I think open weight is such a massive threat in the sense of in the past they were open weight would always be 12 to 18 months behind and now it's six 3 to six months >> which brings us back to like this anthropic IPO question. Okay. and that's obviously something that we're going to be debating for a long time.

49:56 All right. So, we'll we'll see. Very interesting battle. Unsettled, continuing to get more competitive. Before we go, I want to talk to you about the visual touring test. So, there's a company called Tavis. I had Oh, I had never heard of him before. but they have been making AI video avatars. And I'll just read the story is from the rundown AI. Tavis previews Griffin, a lifelike AI that reacts on video calls. it's an AI model that renders a lielike person who can hear, see, speak, and react during a live video call. The start startup says nearly half of the participants in a short company test believed they were talking to a human. let's play it for a moment just so folks can see what this AI avatar actually looks like.

50:46 >> Have you thought any more about how we're introducing Griffin? I don't know. What are your ideas? >> I mean, I think we should stop trying to explain it and just, you know, show them what it looks like. >> All right, Rajan, that completely blew me away. What was your reaction to this? And of course, the study was like, it was a small company test, so it might have been, you know, true believers anyway. but has the video touring or the visual during touring test been passed?

51:12 >> I think it has. I think like >> I mean when you when you watch the video no when you watch video and actually what was interesting to me is it's all the latency like the actual pixel fidelity I feel has actually like not been fully solved but it's definitely been like can you create a human that kind of like adapts relatively quickly how long does it take to actually are you able to render that person but doing it where it's not cuz even I don't know have you ever seen this Tilly Norwood wood stuff.

51:46 >> It's like this AI actress. There's one she was on Piers Morgan. She >> Oh, I did. I did. That was terrible. >> Started. Yeah, that's what I mean. Like >> she was speaking Chinese, right? >> Yeah. Accidentally this thing started speaking Chinese. And it's funny cuz I feel that's one of those like weird like boomer meta Facebook things. so yeah, like that there was a latency. you could feel like there's this long pause while the model is running and processing and inference is inferencing and like they come up with the answer whereas this one is just natural. I mean, if it's I I got to try with everything in AI, I have to like Google's frontier model, which we didn't even talk about because they haven't even released it for all of us.

52:31 >> Pro. >> Yeah. no. Argon. Argon. >> >> yeah. Just release it soon, Dr. Congrats, guys. >> Yeah. >> Yeah. But, but I I it was very impressive. It was very very impressive and it's terrifying as well. >> Okay. Okay, the memes on this were just unbelievable. Autism Capital was just on a rampage. they wrote, "Your dad is about to give away your entire inheritance, you know, quoting this video of this person." and then, they also wrote, "This is going to hit the Facebook boomer community like crack in the 80s." And, you know, the jokes are funny, but they really do kind of make you wonder like, what is this for?

53:18 because it seems like if I'm a scammer, I am like, you know, jumping up and down and thrilled. maybe some customer service examples, but I don't get what it's for. What do you think it's for? >> What is the scam though? Like are do people have and this is a genuine question. Are people having video calls with like total strangers or like not >> if you guess text message the catfishing like those like hey how are you beautiful text that >> I don't know you get randomly >> I know no yeah yeah I I'll admit it I get them every now and again the pig butchering texts we've talked about it on this show for longtime listeners you know we've talked about it doesn't it get easier to do pig butchering when you can throw one of these things out there and 50% of people, you know, if you are to believe the survey, believe it's real. Scary.

54:11 >> Yeah. No, I mean I think it it's it's a problem. It is a genu this stuff is like I mean I don't know like occasionally I'll go on Facebook and the amount of just like random AI stuff on there anyways is kind of terrifying already. >> Just shrimp shrimp. We're so I mean this is the next evolution of shrimp Jesus. You can talk to Shrimp Jesus now. but actually I don't Can you Do you think you'll be able to Do you think they'll have a limited set of avatars or they'll be able to like you'll generate your avatar? This is how it goes in AI.

54:46 There's like one company that kind of does it properly and they have some guardrails and then other companies figure out what they're doing and they do it without guardrails. So yeah, I think there will be a shrimp Jesus avatar out there at some point. and it may may even begin a a new religion. >> All right. Well, Tavis, if you if you're not doing that startup idea for listeners everywhere, >> Shrimp Jesus AI avatar that actually has zero latency and is able to speak to people and actually make them think Shrimp Jesus is real. And maybe we didn't even get into anthropic and Catholicism and consciousness this week, but maybe next week we just it's all religion and AI.

55:30 >> Yeah. I mean, we should do let's do another show next week and we can keep talking about this stuff because >> man, just the the news doesn't end. So, I don't know about you, but I will be spending my weekend talking to Tavis, talking to this avatar, and I guess by the time our conversation is done, we'll be ready to record another episode of the show. >> Yep. Weekend plans. I'm going to have it read the next meter research report to me. He's going to just imagine that just reciting 80 pages of bots escaping being but the narrator is a bot.

56:03 >> Is a bot >> and that's our world. >> Yep. Really cool and normal everybody. Really cool and normal. >> All right. >> Wow. The show went by fast. How are we done already? >> Fast. I know. I know. >> All right man. Great talking as always. thank you Ran John. Thanks everybody for listening and watching and we'll see you next time on Big Technology Podcast.