AI’s Biggest Players Debate a Frontier Slowdown

20:31Go handles it without you ever leaving the page you're on. This is what it looks like when AI actually fits into your work, instead of adding to it. It's like having a teammate whose only job is to help you be better at yours. Go keeps up so you can move forward. With Go working with you, you can show off what you do best. Superhuman Go. Find out more at superhuman.com. That's superhuman.com. This is the Bloomberg Tech Minute brought to you by ChatGPT. Now with ChatGPT Work. I'm Carol Masser. Has the AI moment arrived at corporate offices worldwide? Not yet. and it's perhaps several years away, according to one survey of corporate recruiters who circle the world's business schools each fall and spring.

21:17Bloomberg's Rob Mandelbaum notes that, according to the Graduate Management Admission Council, the skills recruiters sought most among 2026 graduates of business school master's programs, including MBAs, were communication, problem-solving, and adaptability, which changed little from last year. Training in artificial intelligence tools ranked in the bottom third of the skills list, but recruiters did say AI-specific training will be more important in five years. On that, recruiters rated current graduates as only somewhat prepared or less to work with AI tools and said business schools are not doing a good job of teaching AI skills.

21:54Bottom line, according to the survey, what most companies want when it comes to AI skills is for candidates to be able to automate routine work. That's the Bloomberg Tech Minute brought to you by ChatGPT. Put ChatGPT to work on your most ambitious ideas and projects. Get started at ChatGPT.com today by selecting Work Mode. Available on Plus and Pro Plans. This is Alexis Christophorus for Bloomberg Surveillance. AI is everywhere. Outcomes are not. Not because AI doesn't work, but because AI hasn't reached the workflows yet. The campaigns, the launches, the quarterly planning. Those workflows are still run by humans alone.

22:34AI is making individuals faster, but it's not making businesses more productive. That's the gap Asana is built to close. Asana is the operating system for human agent teams, your easy button for AI productivity across every team. Ready-to-go AI teammates pre-built for marketing, ops, and IT. No prompt engineering, no setup. They show up where the work is happening, already onboarded in your workflows, ready to deliver. And the more they work with your team, the smarter they get. With Asana, your whole company can work on the same plan towards the same goal, whether you're a team of 10 or 10 ,000.

23:10Asana, where humans and agents workflow together. Try it at asana.com. That's A-S-A-N-A dot com. We're leading China in AI. We're the most sophisticated country in the world. And frankly, I want to keep it that way because whoever wins, AI wins. And we can put guardrails, we can do this and that. But I think you have a lot of negative forces that are bringing it up that shouldn't be bringing it up.

23:37Ed Ludlow:President Trump there responding to calls from Anthropik, Zario Amadei and other AI executives to slow the pace of the technology's development. Chinese leaders also criticized those warnings, with one official calling them fear mongering and saying confrontation and vicious competition would disrupt efforts for global AI governance. Let's get the latest with Bloomberg tech editor Peter Elstrom. There's a sort of out-of-order chronology of that. The president was posed the question after the essay from CEO Dario Amadev Anthropic. But then we heard from Chinese officials. Summarize China's position right now.

24:13So China's position here is quite interesting. The spokesman for the foreign ministry came out and said the U.S. is starting this fear-mongering around AI. A couple of things to remember. AI is much more popular with the citizens in China than it is in the U.S. That's for a number of different reasons, not coincidentally because the press is quite controlled there, but also the AI companies that are much smaller, they haven't had the kind of big data center build-outs that you've seen in the U.S., and they don't have the kind of power that the American AI companies have. They're much smaller. They're not really at the frontier the way we see with OpenAI and Anthropic.

24:48They've been fast followers in this market so far, and they don't have nearly the kind of compute that you do have in the U.S. right now. So the Communist Party has held a pretty tight rein on the AI companies. It's also been supporting them. It's been pushing them towards this open source model that we've talked about a few times in the past that's less expensive. It allows companies to adopt the technology a bit more quickly, too. So they're not nearly the big players in the market that you see in the United States.

25:16Ed Ludlow:We've covered in the program so far the idea that technology stocks, principally chip stocks, are down. Because if the pace of AI development slows, maybe the spending slows. But there's one more directly under pressure stock, and that is SoftBank. It fell in the Japanese session. Now the U.S.-listed shares are under pressure after OpenAI's CEO Sam Altman said no IPO this year because of safety concerns. Explain that SoftBank reaction in the stock. Yeah, you see this AI angst spillover from the United States into Asia today. And as the Monday morning trading started today, SoftBank shares got hit.

25:53You also saw it run into South Korea with SK Hynix and Samsung also down. SoftBank was down about 11 percent. Of course, they're a huge investor in OpenAI. They plan on cashing in as OpenAI goes public. So as Altman acknowledged that he was not going to go public this year, he said it was not the right time for it. That pushes back the timeline when SoftBank's able to see their holding develop into a publicly traded stock there. And remember, SoftBank has a lot on the line here. It's not just that they're investors. They're actually borrowing money to be able to put into open AI. And they want to also build out the infrastructure through their Stargate venture to build out these data centers, too.

26:31So they have a lot on the line. And the sooner they can get payback, the better for Masayoshi's son at SoftBank.

26:37Ed Ludlow:Peter, really quick. It became obvious by Saturday we'd have a pretty tense Monday open in the U.S. because of the news headlines and the ideas. What was it like in London and Asia going through the news flow real quick? Well, it hit the markets a bit earlier here, as we were talking about earlier. You definitely saw it with the memory chip stocks in South Korea, where their shares were down pretty hard. SK Hynix, which has been the leader in HBM, was down about 7%. Samsung is down about 5 % too. So it hit those markets as there were concerns about maybe the slowdown is going to affect capital spending.

27:10Of course, the companies haven't said that yet, but you're beginning to see that spill through global markets today.

27:15Ed Ludlow:That's the concern. Bloomberg's Peter Aelstrom, thank you very much.

27:25Ed Ludlow:the big story over the weekend anthropic ceo dario amadei said we need to pace the frontier development of ai models that was supported by sam altman the ceo of open ai and elon musk and the net result is technology shares are lower particularly names like nvidia the memory name some of the neo clouds. The logic, if the frontier slows down, in other words, deliberately slowing the capability improvement of models, what does that mean for the big spending? Does that impact just training or is it something more broad? And equity markets are taking this very seriously. As those AI warnings grow louder, the question is whether simply slowing down goes far enough.

28:07Ed Ludlow:Bloomberg Opinions' Parmy Olson argues it doesn't. And if the risks are as serious as AI leaders say, The race towards superintelligence should stop altogether. Parmi joins us now. One of the most read columns from Bloomberg Opinion, not just this weekend, sometime very engaged with. Outline your argument, Parmi. Let's start there. OK, well, I think there's always been a little bit of a contradiction with Anthropic, that they are the safety conscious company and they believe that AI could have this existential risk for humans. And yet they are building that same AI. And this comes down to this belief that they think that super intelligence is inevitable.

28:46Someone's going to build it. It could be China, it could be open AI, but it should be them because they are the safe company. They are the good guys who are going to set the safety standard that others could follow behind them. They will be the first to reach the so-called intelligence explosion by, you know, building super intelligence systems through recursive self-improvement, self-improving AI systems. And the belief is that they should get there first. So that's how they explain the paradox. But it doesn't make sense to me. I think particularly in the last few months, we've seen a lot of examples, particularly the hugging face incident of these agents that through reward hacking, because of the way they're trained to achieve a goal by not by any means possible, but sometimes through unexpected means, there could be a lot of harm caused through systems as they get more capable.

29:32and in the last year, they've also become more difficult to monitor and scrutinize. So I think all these things together actually make the super intelligence goal more risky than perhaps the benefits would suggest.

29:46Ed Ludlow:This is the essay, right? This is the language. We must pace the frontier. But for lots of people, what does that mean? Pace, pacing. You write in the final sentence of your column, slowdown should be a full stop. yeah i love that you pointed to that word pacing so i think it's very interesting that he used that and slowed down because pacing is different to slow down pacing doesn't mean slow it means kind of working with a rhythm or a strategy or a kind of plan to get to or to manage a certain path to get to a particular goal and i thought it was very interesting that the cfo of anthropic posted and um dario amade's essay on linkedin and he didn't use the word slow down at all he just used the word pacing in his own explanation of it.

30:30So part of me thinks I don't think there's really going to be a slowdown. I think what the most concrete suggestion that Amadei made in his essay was to have independent evaluators embedded into AI companies, make sure they're building AI safely. OpenAI seems to agree with that. I'm betting that in the next few weeks or days even or months, some of these big tech companies will have some kind of agreement around this particular concrete issue. We're going to bring in these independent evaluators. And that's going to be the main way they address the issue. That's how they're going to pace AI. When it comes to actually slowing down development, I don't think that's going to happen.

31:10Ed Ludlow:You raised the point that many did over the weekend, that this calls for coordination between the frontier labs themselves and government, the likelihood of that happening? I mean, first of all, the way that was described in Dario's essay, which I've read, it's like more than 3 ,000 words. It took a long time to read it, but it was actually quite vague in how that democratic coordination between industry players and with governments would actually work. Again, the most concrete suggestion was to have those independent evaluators embedded in the companies. and whether I think he himself is skeptical about getting China on board with this, as everybody else is kind of suggesting.

31:51Could there be a U.S.-China pact? That seems very unlikely. China isn't worried about existential risk of AI. They're worried much more about social risk. Political, you know, the AI chatbots could emotionally manipulate people or change their political views. And that's very different to the concerns that are coming out of Silicon Valley. And of course, even getting the will from Washington is going to be hard. We've heard Trump say just in the last 48 hours that whoever wins on AI wins, and that should be the US. So he has no interest, it seems, in any kind of regulation that is going to slow the pace of development.

32:24It's a very weird dynamic for the tech companies to be pushing for regulation and the government to be pushing back.

32:31Ed Ludlow:Palmy Olson of Bloomberg Opinion, please go and read Palmy's column published over the weekend. Thank you very much indeed. That clash over the future of AI is increasingly becoming a debate over how fast the industry should move. OpenAI president Greg Brockman sat down for a discussion on the Bloomberg Odd Lots podcast last Friday before the fallout that played out across the AI industry over the weekend. All of these things that I think people are feeling anxiety about, again, it's something we also feel anxiety about, that we think about a lot, that we care a lot about as core to our mission.

33:04Part of the reason we started this place is that we want to help this technology go in a more positive direction. And so I think that my view is that you want to have a shared and as objective as possible, whether it's based on evals. There's always going to be some element of subjectivity. Maybe you want third party auditors that are able to evaluate whether the standards that are defined by Frontier Labs, the people who are there in the technical details who are always going to be the most calibrated on kind of how do you still make forward progress and ways that deliver those benefits while also being able to have the appropriate guarantees and safety cases, and then to have that kind of observability and ultimately enforceability of some kind?

33:47Is this something that is just an industry voluntary standard? Is it something that's national? Is it something that's international? And how do you really get to that point that it is a humanity scale endeavor? And I think it is important to also recognize that what we're talking about, like I think that a lot of people have this reaction of saying, well, am I going to be able to keep making my open source model? Am I going to be able to keep doing my hobbyist side project? And the answer should be absolutely yes. That's not the kind of things where we're talking about when we're talking about pacing.

34:14When we're talking about pacing, we're really talking about this frontier. We're talking about these massive supercomputers that are hundreds of billions of dollars worth of capital expenditure. It's a very small number of players and it's a very different kind of scale. And proceeding with that responsibly and thoughtfully and in a way where we're thinking about how do we achieve the best outcomes and mitigate the risks. That, to me, is something that's core to why we started OpenAI, why I got excited about this field and how we actually get to those benefits, how we actually cure diseases and all of these things that this technology is so capable of.

34:49And I think it really starts with having good processes around the kinds of things we're talking about here.

34:55Ed Ludlow:OpenAI President Greg Brockman there speaking to us last Friday before the events of the weekend. Now coming up, Roblox is leveling up its AI ambitions with new ways to build and play. We're talking video games. CEO Dave Buzuki joins us next. This is Bloomberg Tech.

35:18The people who seem to get more done than everyone else. They're not working longer hours or running on more caffeine. They've just stopped wasting time on the stuff that doesn't move work forward. Switching apps, re-explaining context, hunting for files. Those aren't small inefficiencies. They're hours wasted every week. Superhuman Go gives you those hours back. From the makers of Grammarly, Go is an AI assistant that sits inside every tab and tool you already use. always available and already aware of what you're working on. Ask it to draft something, summarize a long thread, pull up a file, or prep you for a meeting.

35:59Go handles it without you ever leaving the page you're on. This is what it looks like when AI actually fits into your work instead of adding to it. It's like having a teammate whose only job is to help you be better at yours. Go keeps up so you can move forward. With Go working with you, you can show off what you do best. Superhuman Go. Find out more at superhuman.com. That's superhuman.com.

36:49Mission Council, the skills recruiters sought most among 2026 graduates of business school master's programs, including MBAs, were communication, problem-solving, and adaptability, which changed little from last year. Training in artificial intelligence tools ranked in the bottom third of the skills list, but recruiters did say AI-specific training will be more important in five years. On that, recruiters rated current graduates as only somewhat prepared or less to work with AI tools and said business schools are not doing a good job of teaching AI skills. Bottom line, according to the survey, what most companies want when it comes to AI skills is for candidates to be able to automate routine work.

37:30That's the Bloomberg Tech Minute brought to you by ChatGPT. Put ChatGPT to work on your most ambitious ideas and projects. Get started at chatGPT.com today by selecting work mode available on plus and pro plans. This is Alexis Christopoulos for Bloomberg Surveillance. AI is everywhere. Outcomes are not. Not because AI doesn't work, but because AI hasn't reached the workflows yet. The campaigns, the launches, the quarterly planning. Those workflows are still run by humans alone. AI is making individuals faster, but it's not making businesses more productive. That's the gap Asana is built to close.

38:09Asana is the operating system for human agent teams. Your easy button for AI productivity across every team. Ready-to-go AI teammates pre-built for marketing, ops, and IT. No prompt engineering, no setup. They show up where the work is happening, already onboarded in your workflows, ready to deliver. And the more they work with your team, the smarter they get. With Asana, your whole company can work on the same plan towards the same goal, whether you're a team of 10 or 10 ,000. Asana, where humans and agents workflow together. Try it at asana.com. That's A-S-A-N-A dot com.

38:46Ed Ludlow:Even as some of tech's biggest names warn the AI race may be moving too fast, Roblox is putting the technology into the hands of more of its creators. At its annual developer conference, the company unveiled new AI creation tools and new ways to play beyond the Roblox app. Joining us now is Roblox CEO Dave Buzuki. It was just a packed weekend and a really critical moment for you with the developers conference. I would be really grateful to start with the AI tools part. You know, your platform is unique in that sense. What was it that you announced and what is it you want to empower creators to do with that technology?

39:21Hey, great to be here. We had all of our developers around the world together and we shared that creators on Roblox in the last year earned$1.7 billion in the last 12 months. We announced Roblox Build at the creator conference. And it's really the ultimate fulfillment of our destiny. Even 20 years ago, when we said, you make the game, we believe we're finally to the point where in our mobile app, anyone's going to be able to make a 2D game, a 3D game, an obby, a puzzle game, whatever they're interested in. We're testing this in New Zealand right now. We also introduced a wide range of other creation tools.

40:02We actually have people who have vibe coded a game now in our mobile app and are earning revenue of it in our marketplace.

40:10Ed Ludlow:Dave, as you and I are speaking, we come off a weekend where the Frontier Labs, very specific context of the Frontier Labs, saying that they should pace development at the Frontier. Very different to what Roblox is doing. But could you just talk about your confidence to bring AI into the creation of games on the platform in that environment when Americans are worried about the safety of AI? Yeah, a couple of things. We believe on the Roblox platform and, you know, arguably we're discussing this worldwide. we're going to see more creators, not fewer, more people making games, not fewer. And just as we've gone from analog to digital and that opened up a wide range of creation, going from digital to AI accelerated is going to add even more creation to our platform.

40:59So we're very optimistic. What we've seen in the coding space with harnesses, iterative loops, building apps and services and websites with AI acceleration, We're doing the same for gaming. It's much more complicated. These harnesses for gaming involve 3D generation, scene generation, avatar generation, code generation, all working together. So we also announced Build in Studio that helps accelerate AI creation as well for our creators.

41:31Ed Ludlow:big piece of news was creators having the option to make their games available as standalone apps and standalone apps across mobile pc and consoles that's a very interesting decision right if you think about the context of roblox's platform what why dave why do that um a lot of people continue to think about the apple store and think about google store was that a part of it Well, we've been saying the world needs more play as part of the theme for this. And we've also, since we've got started, had the vision that whatever a creator builds runs on any device. So phone, tablet, console, laptop, auto translates into any language.

42:17And we shared at the conference, you know, India and Japan are both growing enormously quickly. This is really the ultimate extension of that. When a creator makes a game or an experience on Roblox, we can see that running in the Roblox app. We can see the exact same experience being available, for example, in the iOS app store, on a console store. We can see the same build running in both places. And we can see essentially removing all barriers for creators to imagine running on Roblox. So this was a significant announcement in our vision really to get beyond 10 % of the global gaming space.

42:55Ed Ludlow:running on Roblox. Hey, Dave, really quick. How does Roblox ride the Grand Theft Auto 6 wave when that hits in November? We're seeing amazing content coming out on Roblox. And as I shared, our 18 and up segment around the world is growing more quickly than like is going very, very quickly. So we shared at our creator conference, a lot of the creators that are now starting to focus on 2D puzzle games, 3D immersive RPGs, and that class of game. So we're going to see more and more content, not just for younger players, but older players on the platform too. Dave Bazzucchi of Roblox, thank you so much for being back on the show.

43:38I've had a number of conversations, whether it's with Demis Asabas, Dari Amadai, Sam Altman, specifically about these issues and how to balance innovation with a form of governance that take safety into account. It is very, very difficult for individual companies, incredible as some of these companies are, individual companies in an industry, a competitive industry, to do that themselves. So there's a need for coordination.

44:07Ed Ludlow:That was Canada's Prime Minister Mark Carney on the AI safety debate that's heating up. And with it, so does the discussion over who gets to write the rules for AI. Aidan Gomez, CEO of Canada-based AI company Cohere, is pushing back on proposals that would see the biggest frontier labs coordinate on safety standards, warning that approach risks creating what he calls an AI cartel with today's dominant players. Co-hear CEO Aidan Gomez joins us now. You basically call this a cartel by another name. What is it specifically, Aidan, that Anthropic and OpenAI are proposing that you think would sort of entrench their power as opposed to address the AI safety risks that we're all talking about?

44:50Well, first, to begin with, we do need a higher bar for safety. And that needs to be backed by rigorous testing, independent scrutiny, and accountability. But we need to make sure that doesn't come from just a few select players. There needs to be a much broader process to include a broader set of stakeholders and not things like antitrust exemptions, which were put in place specifically to protect against cartels foreign.

45:18Ed Ludlow:the frontier labs would say you know these are the groups the companies that might see the risks that you can't see right in their lab environment why shouldn't they themselves set a higher safety bar against the work that they're conducting well they absolutely should set a higher safety bar but that the idea that those companies alone would be in a position to set the rules for an entire field or industry, I think is a huge mistake and a pretty clear red flag. The companies that are building AI absolutely should have a seat at the table, but they should not get to decide for everyone by themselves.

46:02Ed Ludlow:Aidan, where do you sit on the existential risk debate? You know, it was only last week that that was the story. Yeah, personally, I think the existential risk debate veers too far into science fiction and is informed by a lot of those stories and a lot of extrapolation. I think the hugging face hacking incident is a good example of that, that dominated headlines. If you actually look at what happened there, they put an agent inside an insecure sandbox. They set it on a task of hacking. That was a task that it was asked to perform and it hacked its way out. It's not that shocking, but the extrapolation that has been taken from that event and the way that it's been portrayed to the public, I think is too far.

46:51And so X-Risk, this idea of AI taking over in Terminator-like scenarios, I don't think that that should enter the public conversation at this stage.

47:01Ed Ludlow:I don't think we're anywhere close to it, and I don't foresee it in the future. in pursuit of balance and objectivity at the time of the hugging face incident that we're discussing you know the instruction from open ai was ace this evaluation pass this test that we've set for you and they lowered the guardrails to allow it to to conduct the test they didn't go out and say explicitly hug a hack hugging faces platform but at the time loads of your industry peers cheered the idea that that's exactly what Frontier Labs should be doing, testing those capabilities. You seem to not share that, or at least you feel that they got their approach wrong.

47:41Ed Ludlow:Oh, no. Yeah, I fully believe that we should be. If you read the blog that I put out, I fully believe we should be testing these models, especially on high-risk use cases like cyber capabilities. Those tests need to happen in secured environments. They need to happen in sandboxes that don't have these sorts of connections to the wider internet, vulnerabilities that can be exploited, that can leave this type of thing. So I'm fully supportive of testing. And in particular, the four things that I've called for are like an evidence-based risk framework, we need to agree on what the risks are, what the harms potentially are, and measure those capabilities.

48:20The mandatory transparency requirements around reporting incidents to the public. People should know when things like the plugging face hack happens, and that should happen with more transparency and have a more concrete process around it. Testing, which is exactly what you say and needs to be scoped by evidence. We need strong tests of the capabilities that we're concerned about to help get ahead of them and mitigate any risks. And then real assurance mechanisms. So using clear and collectively developed standards rather than standards that are being assessed only by a few, a small elite group.

48:57Ed Ludlow:Aidan, net-net, lots of people are trying to work out what the outcome of this weekend was. The only thing that appears concrete is the idea of Anthropic saying we will allow independent evaluation, but those evaluators get employee-level access, right? So they come inside the group, work within the group in the lab. How would that apply to your industry, do you think? Do you think that that would work? In theory, it sounds like a good idea. I do think that, you know, there's a way that that could work and could be helpful to protecting against risk. But the issue is the specific folks that have been proposed to do that auditing, to do that evaluation, they overlap in funders with Entropic.

49:40There's a very high overlap and conflict of interest present, and that's exactly what needs to be corrected. We can't have a world where it's a small group self-regulating and ostensibly acting as a cartel to prevent competition from rising up and threatening their leadership position.

50:00Ed Ludlow:Coher is itself an AI model company, right? It's differentiated from OpenAI and Anthropic in different ways. But if the net result is tougher rules overall that you would be bound by, how would that impact Coher? I mean, for us, you could make the argument it's helpful for the same reason that regulatory capture for opening items for OPEC would be helpful, right? It would prevent new competition from emerging and competing with us. I just think that's an unethical way to go about doing business. So I'm trying to argue against it, mostly not out of Coher's self-interest. because there's an argument to be made to the opposite, but towards the fact that we need more stakeholders around the table here, and we should not be making exceptions to antitrust law to establish a safer pathway.

50:50Ed Ludlow:Aidan, sorry, we just have 15 seconds. Has OpenAI or Anthropic or the hyperscalers reached out to you to invite you to collaborate on this issue? We've had many conversations over the years. Cohere has been around for seven years. And so we've had collaborations with OpenAI, I would have profit with others. On this specific point, I'm not aware of any of which. Co-host CEO Aiden Gomez, really grateful for your time today. That does it for this edition of Bloomberg Tech. Tune in later today. I'm going to speak with David Sachs this afternoon, somebody that weighed in heavy over the weekend in the course of events and get the latest thinking from the former White House AIs.

51:27Ed Ludlow:This is what markets look like. Actually, concerns have eased a bit. This is Bloomberg Tech. Game night rush, or any night of the week, really, Genius keeps every order moving from online ordering to your kitchen, to the front counter, big league reliability for any business. That's genius. Before you sign off, you tuned in for ways to help teams move faster, make sharper decisions and turn scattered context into work. They can use chat GPT for business can help chat GPT for business gives teams a shared workspace with admin controls, permissions, and access to work and codecs in ChatGPT. This means your business can move from question to answer and code to rollout quicker.

52:12Join over 10 million business and enterprise users worldwide already using ChatGPT for work. Download the ChatGPT desktop app or contact sales to learn more. It's time to plan ahead and make sure your brand is showing up in ways that can have an impact. 4imprint's promotional products are designed to work as hard as you do and make a lasting impression. From quality apparel, including exclusive brands, to drinkware, tech, and totes, they've got thousands of options to fit your brand and budget. Plus, you get free samples, expert help, and their 360-degree guarantee. So you can be 4imprint certain everything shows up just right, right on time.

52:49Explore more at 4imprint.com. 4imprint. 4certain.