OpenAI Unveils Always-On AI Agent Dots, New $500 Paid Tier

0:00Bloomberg Businessweek Daily is brought to you by HPE, bringing you the self-driving network, a network that's self-optimizing, self-healing, and self-protecting, and only continues to get smarter. Learn more at hpe.com slash networking. Some people treat ChatGPT like some kind of smart search engine, and some use it to get work done. ChatGPT Work is a new way of working in ChatGPT that can take action across your apps and files, stay with a project for hours if needed, and turn a goal into finished work. It's designed to help you move from a chaotic starting point to a reviewable first version.

0:36So all the source materials, briefs, and scattered information that you have to grind through to turn into something useful can just become something useful. Put ChatGPT to work on your most ambitious ideas and projects. Get started at ChatGPT.com by selecting Work Mode, available on Plus and Pro plans. This is Robert Smith from Business History. If you're listening to this, there's a good chance you're a small business owner. And like every small business owner, you started with a dream to do what you love and watch it grow. What you probably didn't dream about? Keeping up with cyber threats.

1:10That's where MasterCard can help. with access to tools that help identify cyber threats to better protect your business. Building your dream business? Priceless. For cybersecurity in a changing world, there's MasterCard. Learn more at MasterCard.com slash smallbusiness. Bloomberg Audio Studios. Podcasts. Radio. News. This is Bloomberg Business Week Daily. reporting from the magazine that helps global leaders stay ahead with insight on the people, companies, and trends shaping today's complex economy, plus global business, finance, and tech news as it happens. The Bloomberg Businessweek Daily Podcast with Carol Masser and Tim Stenevec on Bloomberg Radio.

1:58In just a moment, our own Ed Ludlow sitting down with OpenAI's Sam Allman. We're going to take you to the West Coast as soon as that begins. This is happening at OpenAI's Dev Day, Developers Day, a company targeting$30 billion in funding at a$1.4 trillion valuation. And that is what their goal was, to get to that valuation. Yeah. That, though, is like, it's not overshadowing, but that's not, you know, Dev Day is supposed to be about dots, right? This new news challenger to Meta. So, yeah. Okay. Let's go now to San Francisco and to Ed Ludlow, who's standing by with Sam Altman, the CEO of OpenAI.

2:36Sam, do you know what's interesting in the presentation is how internally OpenAI staff talk to dots? Do you get what I mean? I think I do, but tell me. It isn't treating it like AI, like, hey, chat GPT, question. It's a conversation. Yeah, yeah, yeah. It's not. I don't think it's like you have to pre-morphize. they don't we don't quite talk to it like a person but there is a different style to it than traditional chatubuti the reason i wanted to get to that first is who's dots for really it's starting as a premium product we know it's you know we're putting our best model astro in this we run it a lot prosumer sorry to interrupt or it's an enterprise product more like more like for getting work done which i think is largely enterprise but there's a lot of people that are going to build their own startups that are sort of in the traditional prosumer category but this is like a very capable model and it's a ton of usage so we're starting for something at the top end of the market and you will see us over time make a mass market consumer version how much do you and the team various people at openai benchmark against the agents that are already out there so how much do you use a grok bot a muse and instinct from the product perspective because they've been out there a while and people are actually using them i've tried them i don't use them regularly I have not found a sort of place for them in my life, but I try to try every major new AI product.

3:59This is a totally different category. Like the amount of work, the level of capability, the level of intelligence this can take on. I think this is just a completely new thing. Why is it different? Like literally, are we talking about its capabilities or is this an economic discussion of how you've priced it or where you've made it available in existing tiers? Well, I mean, it is priced higher than other things in the market. But the reason for that is because of the capability and the amount of compute we're putting into this, which is why I think it's useful in a new kind of way at a new level.

4:31You've launched GPT 6.1 Sol at like Astra capability. At a fifth of the price. But the price is so different. And I've been trying to understand what is the big picture story. It seems like we're moving towards cheaper inference at a time where everyone's much more willing to pay for it. how did you arrive at the tiers that you've done? I mean, we've talked for a long time about our goal is to make increasingly better intelligence and to drive the price down and down. And the reason that I think both those are true is the better you make AI, the more people want to use it. So people are clearly willing to spend more on AI than ever before.

5:09But they want to do so much with it. So our belief is that if we continue to drive the quality up and the price down and really lead the parade of frontier at every point on that curve. People will use our tools in tremendous ways. And just even here today, like hearing from developers about what they're building with this and what they can do if we can continue to drive the price down is really quite remarkable. What is OpenAI's business momentum right now? How did you end the third quarter? What does that look like? I don't think we put our specific numbers out. There's been lots of numbers in the newspapers this morning.

5:42It is our, I mean, it's like a very steep growth period for us. I think what people love are models. We have really been able to make new products and tools for developers, for enterprises. ChatGPT is growing phenomenally well. So across the board, consumers, developers, enterprises, all this new stuff that we're launching at DevDay, we see a lot of business momentum. The reported number was ARR 70 billion, I think as a 3Q number. I don't know. Is that meaningful to you? I don't know where these reported leaks or reported numbers come from, but I am obviously not going to comment either way on that, but I will say I'm very pleased with the growth.

6:25The interesting thing about DOT as a product and agent, but also 6.1 Sol, is you made the decision to release those. this week you held back the release of astra 6.1 and last week you paused training on a separate model yeah explain that how you come to that decision so soul 6.1 okay we held back astra 6.1 um we've been talking about this for a few months now but our we're clearly getting into a very steep part of the capabilities curve. And given that our whole mission is to build this for people and to empower people, and part of that safety, part of that's why they're distributing the benefits, when we need to make a decision about slowing down our training or our model release so that we can make more progress and put more of our attention into safety, alignment, monitoring, security, we will happily do that.

7:25Now, of course, we're going to going to make progress, but the world should always have confidence in our safety cases, in our safety claims, and that we are going to be responsible with this technology. And we don't want to do this thing where like, oh, we need regulation or we need our competitors to do X, Y, or Z, or we won't be safe. We are going to do this responsibly. We want the world to have that confidence. Now, there are still things that we can very reasonably release. So like a faster version of Astra or a cheaper version of Astra 6, we can totally do these wonderful new products. But, you know, I think like I understand why some people are just like full YOLO pedal to the metal go.

8:05But we're going to we're going to have to be responsible as we push towards these new levels of capabilities. There are some difficult questions posed. I'm very conscious of the developer audience and also like the users of the technology. You said that the two main things that they want are speed and performance. and you released new mechanisms to achieve that with existing generation of model. Why is that better as a technique than releasing a more capable model? Well, I mean, they want that too. Yes. They'll also do that. So I think all of those things are good. You know, speed, price, and capability are certainly three of the things people ask for the most often.

8:41And I think the answer is people want all of them in different use cases. That said, ultra-fast, I think, is a really special thing. I did not appreciate how much I was going to like it and how hard it is for me to go back. I should have known because, like, we all want everything to be faster, but it is awesome. Is it an economic proposition for OpenAI? Are the economics a bit good for you? Well, it's much more expensive to run, but we do charge more for it, too. So we tried to make it something that, you know, it's like a... We can talk to Sarah about the balance between... Yeah, I mean, it's like, you know, I think the market will decide, but I think people will be excited about it at this price, and it's reasonable for us.

9:17You were just asked again about an IPO. And what you said is you want to be public one day. Yes. But essentially, the duties of being a public company are incompatible with a dedication or focus on safety. Is that? Not quite. To be a newly public company and for us to understand how to live life as a public company with, you know, pressure from investors and sort of just all of the additional weight that comes with that. while we are going through this period of adjusting to this new level of capability and the new safety requirements for that. We just want to get our feet under us, make sure we understand how to operate in this new way, be able to make some of these decisions in front of us without the pressure of being a newly public company.

10:02And then I think we'll be public at some point. There will be more new levels that come, but it won't be all at the same time. Do you feel support from the existing cap table to just hold off 2028? Does it matter? our investors seem like very happy with us and very patient. Let me ask you this to conclude. It's very rare people have said for you and Dario Amodei and Elon Musk to agree on something, this idea of pacing the frontier. And OpenAI has disclosed quite clearly internal frameworks of what you plan to do. I think people are still trying to understand what the collective effort looks like.

10:37If you collaborate with your peers or you work with the administration, What does that actually result in on an AI safety plan? I think one thing that would be good would be shared standards for safety cases about when, you know, what you need at each level to proceed without undue risk. And then some sort of shared methods for ensuring that everybody is doing a good job in implementing those. And that could be independent evaluators. That could be government review. That could be companies checking each other's work. There have been many proposals there. But I think those two things would be good.

11:10How do you feel about the disclosure of the risk? You talked to the United Nations about it doesn't matter the percentage risk, it's still too great. The news cycle had in Anthropics prospectus, the existential threat to humanity disclosed in black and white. Is that incumbent on you as a frontier lab? I appreciate you filed confidently, but do you just need to put it out there? I think this has nothing to do with going public. This is just, I think, our responsibility for our mission and what we're working on. To do it anyway. To explain to the world the risks and benefits we see in this sort of shifting landscape as things evolve.

11:44Sam Altman, OpenAI CEO, having us here at Dev Day. Thank you very much. Thank you for coming. Indeed. Romain, back to you. All right. There you have it. OpenAI CEO Sam Altman out there in San Francisco. It's Dev Day for OpenAI. That's Sam Altman sitting down with Bloomberg Tech's Ed Ludlow. All right. We want to continue with this. There were a few headlines out of that, building off of some of the Bloomberg reporting earlier today off of the keynote and more. Dave Lee is back with us, columnist at Bloomberg Opinion. He also has a column out, can Sam Altman and OpenAI catch up while slowing down?

12:16It's a great question. What jumped out for you there in that conversation? There's so much swirling around Altman speaking at this moment in time. Yeah, what I thought was particularly interesting was right at the beginning of the interview, Ed, as he often really does very well, he sort of said, who is this product for? And he was describing Dots, this new sort of friendier looking way to interact with open AI agents, feels to me, and it certainly looks like a response to some of the things Meta has been doing with its Muse bot, also what XAI has done with Grokbot. But when he said, who's this for?

12:51I don't think Sam Maltwin was too convincing. Because for one thing, the cheapest you can get this new bot is going to be$100 a month. That's the plan that it arrives on. And I wonder, you know, are the people using OpenAI for$100 a month, do they really need a cuddly version of their agents? I'm not sure. I think, you know, not to sort of downplay what this new agent can do, but the branding does feel a little bit messed. You know, I was just looking at this in the last hour and actually pulled up Reddit to see what people on Reddit were saying about this. And a common criticism, it seems, is that question of who this stuff is for.

13:29One person, I don't have it in front of me, and people don't use their real names on Reddit, but said something that I think was very interesting. And you spent years in Silicon Valley. You get this. the idea that this is being developed for people who who live in some sort of different world of a couple square miles in the bay area who interact with this technology each and every day not necessarily the people who are are out there actually living lives right i think there is a tendency not to say those people don't live their lives in san francisco and silicon valley but the idea of like okay who's the normal person right and a lot of those people's lives are taken up by following AI news, seeing what the latest models do, et cetera, et cetera.

14:11I do think OpenAI does have a communication problem there because often when it does these presentations, and granted, this was a dev day, right? So it was for developers. But when Apple does a dev day, there are moments in that presentation that you know are designed from the mainstream audience. This is how our product's going to work. This is why you might find it useful. And the use cases are things like, I'm planning a barbecue with my friends or typical things like that. When OpenAI does it, they seem to focus on, oh, we're trying to migrate from this API to the new. And I'm thinking, wow, what Sam Altman needs to perhaps realize is that not everybody is building a startup, right?

14:46That's not what regular people typically do, even if they might want to someday. And that's something OpenAI needs to address, really, because when you visit ChatGPT, particularly today, what you're presented with is a sort of list of all these different options and customizations. Unless you're following this day-to-day, and I try to and find this difficult, unless you're following so closely, you don't really know what they all do. Here's somebody weighing in. If these assistant dots are accurate, follow protocol, don't hallucinate or make mistakes, then sure, but imagine going on an international trip and booking the wrong flight or a hotel in the wrong part of town.

15:17The downside is that it's been dotted. With small errors. But I mean, I don't know. Does it, is it kind of, think about when you're planning a trip, sometimes you can spend a long time doing searches to figure everything out. What if it kind of maps it out and then you have to go back and check it? Maybe you do save some time. I think there's a question of, you know, is it quicker to do something yourself and know it's right or to fix something somebody else has done? I think some of these things mirror to me the sort of early days of the Internet. We weren't quite sure whether things could be trusted, particularly when it came to paying for things and so forth.

15:48I think that's a problem that AI will solve just by sort of repeatedly getting it right. You know, I've been using particularly Meta's assistant quite frequently. And I found sure it does some things wrong and there are questions still about what data it's using. but by and large most of the things i ask it to do it does pretty competently well and so i think that trust does does does come over time but there's this there's this adoption um you know i think a lot about my dad in the 1980s when he lived in new york and worked in advertising uh yeah there were no computers on the desks uh somebody came by each morning with all the newspapers uh they weren't checking their email first thing seems great doesn't it it does seem great but the workflow workflow was completely different obviously the pc revolution changed that the mobile revolution changed that we've adopted workflow a different way are we going to use these devices these muses these dots um are we going to build those into our workflows i think it's quite possible you do one of the things that is going to be a really interesting discussion with this is to what degree these bots will be interoperable right because right now if i'm amazon for example i'm going to say well our thing is e-commerce and we're not going to let an external bot use our system which is what they've done they've been blocking these bots right if i'm meta i'm going to say well guess what your bot's not going to be allowed to run on my social network because that's that's our product um google don't get on google search don't get on google maps if none of these individual bots can do it all then they might as well do nothing right because because they need to do a number of basic things, because then that will be the one bot you want to use.

17:27But of course, all of these companies want to be the company that produces the one bot, which is what the race is, what we're seeing right now. When it comes to corporations, though, and they have different softwares from different places, don't they have to, isn't that part of, they've got to figure out how to work, and they do ultimately, right? They have to figure out how to work. The question is, you know, to what depth that working is. So for example, Apple is more than happy for you to have an app on its phone. Delighted for that to happen. It makes them money. it can make the company's money as well.

17:53What Apple doesn't want you to do is read a person's messages because that's Apple's. They don't want you to read their emails and so forth. And so Apple can retain that for itself and I think have a very, very competitive consumer AI. I think in fact Apple, because of the iPhone, is in the driving seat even though many people think it's behind. Interesting. We only have 30 seconds. Did Sam Altman assuage our fears about the AI apocalypse? I am no more discontent than I was before. Okay. But do you think an IPO could be pushed out even further? 2028, I'd said, and he's like, well, his investors are happy right now.

18:28I think Sam Altman is in no rush to do it, put it that way. If they can keep on raising money, which it seems like they're confident they will do. Yeah. What's the latest valuation? Well, in talks for a$1.4 trillion valuation with a$30 billion funding round. Yeah, and that's just now. It's just unbelievable for a private company. So much going on. We know you'll be back. Dave, thank you so much. Dave Lee, thanks for listening in and giving us some analysis there. Dave Lee, columnist for Bloomberg Opinion, covers technology right here in our Bloomberg Interactive Brokers Studio. Stay with us. More from Bloomberg Businessweek Daily coming up after this.

19:07Bloomberg Businessweek Daily is brought to you by HPE. bringing you the self-driving network, a network that's self-optimizing, self-healing, and self-protecting, and only continues to get smarter. Learn more at hpe.com slash networking. Some people treat ChatGPT like some kind of smart search engine, and some use it to get work done. ChatGPT Work is a new way of working in ChatGPT that can take action across your apps and files, stay with a project for hours if needed, and turn a goal into finished work. It's designed to help you move from a chaotic starting point to a reviewable first version.

19:44So all the source materials, briefs, and scattered information that you have to grind through to turn into something useful can just become something useful. Put ChatGPT to work on your most ambitious ideas and projects. Get started at ChatGPT.com by selecting Work Mode, available on Plus and Pro plans. This is Robert Smith from Business History. If you're listening to this, there's a good chance you're a small business owner. And like every small business owner, you started with a dream to do what you love and watch it grow. What you probably didn't dream about? Keeping up with cyber threats.

20:17That's where MasterCard can help with access to tools that help identify cyber threats to better protect your business. Building your dream business? Priceless. For cybersecurity in a changing world, there's MasterCard. Learn more at MasterCard.com slash small business. You're listening to the Bloomberg Business Week Daily Podcast. Catch us live weekday afternoons from 2 to 5 Eastern. Listen on Apple CarPlay and Android Auto with the Bloomberg Business app. Or watch us live on YouTube. All right, TikTok, everybody. We've got about 18, 19 minutes to go until we wrap up the trade on this Tuesday.

20:55Carol Master, Tim Stavik live here in our Bloomberg Interactive Brokers studio on a day where we are seeing kind of little moves when it comes to the equity averages. I'm just pulling it up here. Just down about 10 points on the S &P 500. Most names, though, in the S &P more than 300 are actually lower on the day. So again, more of a risk-off trade. NASDAQ 100, though, bucking the downward trend, it's up about 62 points. Tim, that's good for a gain of about two-tenths of a percentage point. Who cares about the higher yields? Who cares? Tech stocks don't. I think it is surprising. I don't know.

21:33I'm trying to make sense. Oil prices down up? I thought they pulled back a little bit. Oil prices are down ever so slightly today. So like around, you know, one or two a barrel for Brent. Yeah, for Brent. But it's not like all of a sudden the equity trade is saying, okay, we're not worried about inflation, you know, and kind of thinking that the Fed will do less than more when it comes to increases. Look at the long end. I know. 30-year yields, highest level going back to 2002. Yeah. So, okay. Okay. Tomorrow's going to be important, the inflation print. And then we've got jobs on Friday, and then we start to get into earnings season.

22:03Curious what Ann Maletti has to say about all of it. She is head of equity investments at Allspring Global, and she joins us. They've got about$642 billion in assets under management. She's back with us from Milwaukee, Wisconsin. And you know market environments. Once again, it's like, where are we? Trying to make sense of it. What's your read on this Tuesday in September as we get ready to kick off the fourth quarter? It's a great question. It does feel like we're on a plane that's still accelerating in takeoff because earnings growth has been so strong. In fact, we expect that we're going to enter Q3 reporting season with the third consecutive quarter in a row with accelerated earnings growth.

22:45So that's pretty spectacular. At the same time, we're kind of getting hit with a headwind or a crosswind, which is creating a little drag. And that's definitely the yields that we're seeing in the fixed income market. And that is bringing a little bit of, I guess, you know, creating some unsettled feelings for the equity market. But equities don't seem that unsettled. You know, Tim, I was listening to your comments when you started. Like, it does feel like that on the surface. Like, you're kind of surprised that the overall pullback has been less than 5%, but there has been more disruption underneath.

23:21And if you look at the S &P 500, more than half of the index is trading 15 % below its 52-week high. And so we are seeing greater disruption. And I think that, again, points to what industries are tolerating this new headwind in a stronger way than others. And that's kind of what we're looking at. There are some underlying factors that we need to assess and address as equity investors. And if I look at the major industry groups in the S &P 500, no surprise to see energy at the top of that list. There are 11 of them, right? It's up 37%. Infotech is up 27%. Healthcare is up 10%. I mean, most of these industry groups, eight of them are higher.

24:08At the bottom are financials, consumer discretionary, and utilities. When you look at that churn, consumer staples are up almost 6%. Materials are up almost 8.5%. I mean, what picture does that paint for you, that there has been a broadening of the trade? A broadening of the trade, but also a realization that we're likely in a higher rate environment for longer than any of us probably expected. Again, it's an interesting time because the fundamentals remain quite strong, but inflation is not where it needs to be. The Fed's very, very focused, as well as the bond market, on that. And that's what's creating the drag.

24:49And so it feels to me like we're in not a typical higher for longer rate environment, but still seeing some things that rhyme. So you have sectors like technology, you know, against selective areas of technology where growth looks like it's pretty sustainable, at least over the next couple of quarters, to be working really well. But then also investors looking for other places. There's a lot of concentration in the market and places like healthcare is a place where investors really haven't paid a lot of attention. Fundamentals look strong there to us as well. Earnings growth may not be as great as technology, but valuations are near decade lows as well.

25:29So I think it's diversifying your holdings is really, really important instead of focusing on concentration, which I think the market was over the last couple of years, at least. Well, speaking of technology, some of the biggest names in technology were at the White House for a lunch today. Ostensibly to talk about AI and understand this stuff. I mean, some of the leaders from these companies, including Dario Amadei, was there. The president, we're getting some headlines, saying that he signed an AI constitution with firms, that AI firms know they need to self-police and that he's thinking about a committee of 10 people to watch over AI.

26:05Yeah, exactly. He says the AI accord, it's morally binding. This AI document, morally binding. He says bad actors will get caught. That's why we have the Department of Justice, the FBI. He says AI firms know they need to self-police. So if the AI model not used for good, we will nab them. So again, this continues to evolve. He's speaking alongside Jensen Wong live right now at the White House. Yeah. And I'm not sure if we can grab that live feed for everybody. We can't grab it. But nonetheless, we're watching and monitoring as these headlines come out. And the AI trade and narrative continues to evolve.