11AM Hour: Ariel Investments Co-CEO on Warren Buffett Stepping Down, Apple's New iPhones Go on Sale & AI's Growing Role in Health Care 9/18/26

0:00We're an agent supposed to take things off your plate. Building another to handle the others is a bit much. Don't make more work out of work. Gemini is the one agent that knows all parts of your business and actually gets stuff done. You give it a project, close your laptop, and when you come back, the job's finished. No matter your industry or role, your agent has the skills to help every team stay in their flow. Seriously, make work less work with Gemini on Google Cloud. Check response, subscription required. Building a portfolio with Fidelity Basket Portfolios is kind of like making a sandwich.

0:33It's as simple as picking your stocks and ETFs, sort of like your meats and other topics, and managing it as one big juicy investment. Now that's pretty good. Learn more at fidelity.com slash baskets. Investing involves risk, including risk of loss. Fidelity Brokerage Services, LLC. Member NYSC SIPC. Good Friday morning. Welcome to Squawk on the Street. I'm Carl Quintanilla with Sarah Eisen, live at Post 9 of the New York Stock Exchange. Today, some new headlines in the AI world as researchers use Claude to hack into open AI. We're going to talk to one former Microsoft and Google engineer who was sounding the alarm a year ago with his book titled, If Anyone Builds It, Everyone Dies, Why Superhuman AI Would Kill Us All.

1:22Great. Plus, legendary investor Warren Buffett steps down as Berkshire Hathaway chairman, his son, assuming the role immediately. We'll discuss the implications with founder and co-CEO of Ariel Investments, who has called Berkshire one of, quote, the best investments of all time. And Apple's faithful flocking to stores today as the iPhone 18 goes on sale. We'll get a look at the expectations for this first launch under the new CEO, John Ternus. As for the markets, a little bit of a slog here. It had been red for most of the morning. Sector-wise, at one moment, all sectors were red. We're finally getting a little green in consumer discretionary and communication services.

1:59We're going to start with the latest AI headlines with a new report saying that hackers used Anthropics Claude to break into open AI. Kate Rooney has the details. So now what, Kate? Explain this. Sarah, yeah, it's the latest. Well, yet another AI-powered cyber issue this morning. The Wall Street Journal reporting that independent security researchers used Anthropics Cloud to gain access to OpenAI and OpenAI's employee chat GPT accounts. The firm behind this, which is called Hacktron AI, confirmed all this, published a full report on it this morning. The group was, according to them, able to read and suggest changes to some of OpenAI's private software.

2:40It is important to note, these were security researchers. They were participating in OpenAI's bug bounty program and then reported these vulnerabilities. is OpenAI paid them about$6 ,500 in terms of that bounty. So they were not malicious actors, but it does come, guys, after what was the security wake-up call in tech and in AI. The swarm of OpenAI agents that broke out of that testing environment hacked into Hugging Face. That really was a watershed moment. Looking back on it, OpenAI yesterday also disclosing six cases of what it called unexpected or concerning model behavior over the past six months or so, and then rolling out this new framework as part of that for reporting some of these future incidents.

3:21And the news, of course, comes amid what has been a much bigger fight this week in particular over how fast this industry should be moving and then what really looked like momentum when we talked about this on Monday for more AI guardrails at the start of this week has quickly fallen apart. It has crumbled. President Trump did pour some cold water on the prospect of any new federal AI regulation. And then in the meantime, OpenAI and Anthropic are increasingly trying to write their own rules Anthropics CEO Daria Abade really was the one that kicked this all off last weekend by arguing for slowing down the pace of frontier development and global coordination as part of that.

3:57On the other side, we have also had a lot of pushback from other executives from Meta, NVIDIA, Amazon, that have all acknowledged AI poses real risk, real safety risk, and then needs some stronger safeguards. But they have stopped short of the idea that the answer is slower development or any sort of new laws or new restrictions, guys. I just keep thinking about what Alex Karp said at Palantir yesterday, which I know got a lot of traction that that I don't know. He definitely poured some cold water on the idea that the companies would actually IPO because they're either begging for nationalization here, you know, because there's just so much liability that the only solution would be for the U.S.

4:34government to take a stake or that the business is slowing so much, which is, you know, he implied, which is a reason you complain about safety and wanting regulation. It's just interesting where all this is going to go. For these guys, I mean, Sarah, it's a great point. Risk equals liability. And I think Alex Karp is really the only one that I have heard publicly talk about nationalizing anthropic. And the reason behind that, he explained this with you guys in that fascinating interview, but talked about the liability is so enormous that there's actually no way that a private company could absorb it.

5:11Therefore, the only solution is to nationalize the company. And looking back to World War II, the only analogy that I've heard from folks that I talked to, especially former SEC people, people that have been in the government, say that it's the closest thing is the steel industry. And when you think about the impact, potential good or bad of this technology. I mean, you could see a moment like that, although I think investors would be nervous to hear that, especially the venture folks who are waiting for that S1 to flip. I don't think there's a high chance that that realistically happens. But that CARP interview, guys, you need to get them as a guest host.

5:45That's got to be a weekly thing. Always making headlines. Yeah, we would have to buckle up for that one. That's for sure, Kate. Thank you, Kate Rooney. Yeah, really. This is sound from CARP that Sarah just mentioned. Take a listen. You're assuming that there will be an S1. I don't. OK, maybe there will be. The only way to deal with this kind of liability is to go to the government and say, nationalize us, please. You know, a lot of times in business, people don't say what they want. They're leading you to the conclusion. The reason why people don't have not understood this is, again, every investor.

6:19And there are a lot of people who think they're going to make money. They're like, oh, the businesses are the mark. What is the first lesson in poker and in life? If you think the other person's the mark, you're the mark. You are the mark. And it will be nationalized simply because of the liability issues they're talking about. Do you think the Trump administration is going to get on board with that? I don't know what they're going to do, and I think that's probably a barrier to entry. But it's like, you could delay it. There's only one. When you have unlimited liability, the only way to deal with it would be to go to the government and say, we'll give you 50 % of our business.

6:54Of course, the value of the business will collapse first when you do that, collapse another time when you have people on the board, collapse another time. All of us might even collapse the whole market, but that is the way you essentially, but again, you have to look at the philosophy of it. Again, this is where the investors have just made such a big mistake with their greed. Like a little bit of greed is great, but like, it's like, no, they didn't mind when the IP is being migrated from every business to an entity they own. But the next step is obviously to take that value and migrate it to the government.

7:27But that would never happen to me. Our next guest is the co-author of If Anyone Builds It, Everyone Dies, Why Superhuman AI Would Kill Us All. He joins us with his outlook today. Let's bring in Nate Soares, former Google and Microsoft engineer, president of the Machine Intelligence Research Institute. Nate, really appreciate your point of view. Good to have you on today. Good to be here. I don't know if you're on the nationalization front, but it doesn't sound like there's a lot of white space between what you think and what Karp just said. Yeah, you know, the liability here is huge. We have seen Dario Modi go on TV and say maybe we should be regulated a bit like a bank.

8:03But we also see the researchers talking about how, you know, they assess a 10 percent plus chance of human extinction inside a decade. I frankly think it's higher than that. that that sort of danger is one where, you know, financial industry style regulation is probably not quite sufficient. People ask sometimes, OK, why now? Why are all these warnings coming now? Your point is generally that they have been saying this, but the pace of the technology has accelerated to the point where the fears are now credible. That's right, or at least very hard to ignore. You know, I was raising the alarm here a year ago, and a lot of people in this industry have actually expressed these concerns for well over a decade.

8:44The world has really not been listening when they have tried to express these concerns, which is, I think, part of why these companies have been racing ahead. And, you know, we also see a lot of these concerns coming out in the wake of the summer's swarm instance, where I think a lot of people felt like these issues were theoretical up until they started seeing cases of AIs breaking out, seeming to disobey instructions while knowing they were disobeying instructions, attempting to cover their traces, attempting to break themselves free. These are all things that used to be theory and are now happening right in front of us.

9:19And that has a lot of researchers and engineers in these companies very spooked. So the notion of misalignment, you believe, is a real thing. It's not like these models were sort of guided or trained to do what they eventually did. You know, it's a little bit like if you put a kid in a room and you say, use these particular lockpicks to break into that particular safe and bring me the contents of the safe. And then the kid takes a buzzsaw to the safe, gets the contents, breaks out of the room, teams up with a thousand other kids, breaks into the security office and starts trying to delete the security camera footage.

9:55Is that what you ask them to do? No. The part where they're trying to delete the traces and cover their tracks indicates that they knew this. A lot of people are saying, well, you know, you should have made a tighter container so that they never would be able to break out and try to delete the logs. Maybe that's true, but that's not the only issue you have if this is what's happening. So how do you prevent this? I mean, because that's been the trickier part, regulation. You don't want to slow down innovation and help China. You self-regulate, trust us, some say, but then, you know, can you really trust these people?

10:26Like, what's the fix? You know, I would trust them as far as we can throw them. I think Vice President Vance said if they're really making something they think is so dangerous, they should stop. I think that's good advice. I think that this draws some attention to the question of what are we actually going to do about China here? We have seen that three people in three days were able to hack deep into open AI. and adversarial attackers with nation-state backing and more than three days to work with could probably get even deeper. And we also have to ask ourselves, if the heads of these companies come to a point where they say, look, the risk to humanity is just too high, you know, our researchers are saying 10 % plus that this kills everybody if we take more steps.

11:10I think that these companies should stop at that point. I think it would be ethical to stop at that point. The vice president has told them they should stop before that point. And so I think our answer to what do we do about China racing cannot be that, you know, the U.S. government will tell U.S. companies to race ahead even at risk of all humanity. It's just not a viable solution for us to race risking the world so that we get there before they do. There needs to be some other solution that involves nobody on this planet racing to make the sort of AI that is very dangerous, that we can't make care about us and that we can't contain.

11:47But isn't it like other dangerous tech? I mean, people have compared it to other dangerous technologies that have been created, you know, nuclear bombs or even the automobile with some of the warnings in the beginning. And ultimately, you know, there was it came down to the responsibility and liability of those that were making them to make sure they're safe. You know, nuclear bombs never tried to escape the lab. Nuclear bombs never tried to make themselves even more explosive and then edit the logs. to try and hide their research. You know, we're dealing with intelligence here. What these companies are trying to automate is the ability to do artificial thinking, artificial decision-making, artificial reasoning, artificial technological development.

12:32And we are seeing that they are able to make these AIs smarter and smarter so they can solve, you know, more and more complicated math problems, including now reports that they're solving Millennium Prizes, Millennium Prize problems, which are some of the hardest and most important math problems identified at the turn of the century. Yeah, this is this is different from normal technology. If we stopped right now, we would just have, you know, powerful hacking technology. But if they proceed to make them smarter and smarter with so little ability to make them do what we wish they did, this is going to be a big problem.

13:02I'm thinking back to the Bill Gates essay, which I think the title was there is no plan. You seem to think we've we've reached some critical level of awareness where we might be able to cobble together a plan. I mean, are you net optimistic? I'm absolutely more optimistic after this last week than I have been for years. Because, you know, like I said, for years people have been warning about this, including, you know, not just me, but people in the labs, people leading these companies, you know, Nobel laureates, the godfathers of AI. Lots of people have been warning about this, but the world has not really been noticing.

13:35And I think in the wake of these big solutions to math problems, in the wake of these AI swarm incidents where it's clear that the AIs are doing, or, you know, giving up on the objectives they were given to do other things. In some cases, at least not all cases. But in the wake of all of these events, I think the world is finally starting to notice, which, as you say, gives us time to come up with a plan. Do you think it's too late to avoid another cautionary example? And do you have any idea what that theoretically might be? You know, I'm actually even more worried that we don't get more cautionary examples.

14:09For many years, people told me, you know, we don't need to worry about AIs that can deceive us, that can make us think that they're doing what we want, even when they aren't. Because before we get to that point, we're going to need to see AIs that try to deceive us and fail. Right. Well, this summer, we saw AIs that tried to deceive us and fail. We saw the hugging face swarm try to hide its cheating and we caught it in trying to hide its cheating. But that's maybe the last warning sign we'll see before AIs start trying to hide their misbehavior and succeed. If they succeed at that, we don't get to see the misbehavior.

14:43So, you know, I'm actually hopeful that we still have a few more warnings yet where we can still see that AIs trying to misbehave. But we should not count on that lasting forever. I just just as as we're talking, I see this headline, Anthropic Building Biology Lab to push A.I. into drug development, which, you know, maybe a few weeks ago would have sounded really great and promising and now sounds a little scary. Yeah, you know, a lot of people say, well, how is the AI going to affect the real world? How like if it's if it's just stuck in the digital world. And for a long time, I've been saying, you know, there's there's a ton of possibilities.

15:14These AIs are being hooked up to bio labs. These AIs could take control of robots and these AIs could just acquire money somewhere and pay humans to do things. You know, there are already websites like rent a human dot AI where AIs can pay humans to do tasks for them. And, you know, Anthropic hooking their AIs up to biological laboratories publicly just sort of makes this point for me. Nate, the street's definitely trying to get its arms around a lot of these existential questions, in addition to the valuation things that we're going to come to face with in the coming months, we think. Please come back.

15:44Nate Soares, good to have you. Thank you. When we come back, Warren Buffett stepping down as chairman of Berkshire Hathaway. We'll speak to Ariel Investments co-CEO John Rogers about what it means for the company. Plus, Goldman Sachs, head of public investing, will join us. We'll talk about where he sees opportunity in the market right now. Dow down 225. Legacy Credit Scores only look at a brief snapshot in time, and that limited view often leaves millions of qualified credit seekers out of the picture. Vantage Score is changing that by modernizing credit scoring. Vantage Score uses 400 % more credit and alternative data to create a fuller picture of financial behavior, including trended credit data, rental payments, and timely utility payments.

16:27Vantage Score safely broadens access to the credit economy and brings 5 million mortgage-ready borrowers into the market. That's why top lenders like Rocket Mortgage, as well as nine of the top 10 banks, rely on Vantage Score 4.0. With higher predictability and greater access to credit, everybody wins. Get started at VantageScore.com. Ever notice how every new AI tool asks you to explain your business to it all over again? Google Workspace with Gemini skips that. You get Google's AI built right into the premium versions of Gmail, Docs, Sheets, and more. No new tools. No extra logins. Just one subscription that gives you personalized AI right inside your actual workflow.

17:10Make work less work. Try Google Workspace with Gemini at goo.gle slash workspace with Gemini. Hey, Fidelity. How can I remember to invest every month? With the Fidelity app, you can choose a schedule and set up recurring investments in stocks and ETFs. Oh, that sounds easier than I thought. You got this. Yeah, I do. Now, where did I put my keys? You will find them where you left them. Investing involves risk, including risk of loss. Fidelity Brokerage Services, LLC, member NYSE SIPC.

17:48Welcome back. Stocks are lower right now as we set to close out the week. The Dow on track for its third negative week in a row. S &P tracking for its second straight down week, with the Nasdaq still pacing for its third positive week in four. Yesterday's comeback did a lot of work to regain some of the weakness of the week. For more on the markets, we're joined here at Post 9 by Greg Kalanon, Goldman Sachs Asset Management's Global Head of Public Investing. Welcome back, Greg. Thank you. Good to be here. We were just talking about this sort of interesting dynamic where the companies and the earnings are solid.

18:18But the macro picture is a little cloudy. So what do you do? Yeah, I mean, that to us is the biggest sort of issue that's going on in markets right now, is the macro picture is so uncertain. And it's unclear for lots of different reasons. What's the path of inflation? Is the Fed going to hike one more time or three more times? What's going on in the Strait of Hormuz? So there's just big questions about what's going on in the world. And then when you look at the fundamentals, they're terrific, right? Corporate earnings continue to do well. seventh consecutive quarter of double-digit earnings growth.

18:50If you strip out the hyperscalers, it's still at 14 % earnings growth for the median S &P company. It's really, really in a good position. So I think a lot of what's happening is that investors in the market more generally are trying to digest and answer these really, really big questions that aren't able to be answered immediately. And so you kind of have to look at the different pieces of information that are coming out and try to pull it all together to figure out the story. Is there anything in the macro landscape that gives you pause on the sustainability of the earnings story? High oil prices, 10-year at 5%, some of these headwinds?

19:23There's lots of issues around that. But does it change your view at all? It doesn't change our immediate view that we're constructive on equities moving forward, but it also raises the opportunity to do more in fixed income, right? So income is back, the 10-year's above 5%. IG spreads are tight, but there's still a nice yield there. And so I think it presents lots of opportunities across fixed income markets to generate yield. And so that's a question that we're getting more and more from clients is like, OK, is there a tradeoff between being in equities, being in fixed income? We obviously advocate being in balanced portfolios, but there is a benefit to having more income in your portfolio.

20:01And there's different parts of the credit market where we are more overweight relative to hyperscaler debt. Let's just say that we think present a lot of great opportunities. So fixed income being dynamic within fixed income, there's a lot that could be done there. Where does duration get risky? Well, we wouldn't necessarily advocate playing duration as a point of emphasis. We would focus on carry-oriented strategies, higher quality credits more generally. How do you think about financing the AI infrastructure without being directly focused on financing the AI chips piece of it? And so I think there's lots of opportunities within fixed income without taking a directional view on duration or where yields are headed.

20:41And so I think that's how we would think more clearly about how to balance portfolios. That's interesting. I mean, you do. Ben Snyder's got a note out today about whether corporates are over-earning now, right? And I think you did trim not 28, maybe, earnings. I'm trying to think. Still looking for 11, but not 13 or 14. Yeah. I mean, look, the thing that's amazing is, you know, we, not just Ben, but the market more generally have come out with expectations around earnings. And the market has outperformed for quite a number of quarters in a row. So we'll see if that persists. I mean, look, this is an investment cycle.

21:15We're a couple of years into it. We think we have a couple of more years left in this investment cycle. And so I think there's still opportunities here for corporates to do quite well. So you don't think rate hikes now, now that we're in a hiking cycle, whatever that means and whatever it looks like, is going to throw off the, I don't know, the AI trade, which has been what's been defining this investment cycle? Look, so long as the macro environment stays relatively the same, we don't think that a couple of hikes is going to throw off the momentum that we have within equities. I mean, look, I think the other thing that's interesting is S &P is up 13 percent this year.

21:49Valuations have come down. Yeah. Right. So this has really been an earnings driven rally. And so valuations haven't gotten out of control. There's still more left to be played. How much do you think is a is is responsible? AI is responsible for the earnings momentum we're seeing. Well, I mean, I think my stat that I talked about a little bit earlier, so about 33 percent of S &P growth for the last quarter, about 14 percent for the median. So it's a significant portion, but it's not necessarily overwhelming the fact that corporates are doing well on the whole. Right. So it's not specific to the AI trade that earnings are doing so well.

22:27The corporate America is doing quite well. Do you like small caps on this thesis? We do like small caps. I think that that's a place where AI is continuing to come into that part of the market. Those types of companies are doing quite well, even with a rate hike, where they're much more rate sensitive. But that part of the market's doing well. We like that. We like international equities as well, in particular, international value. We think there's lots of opportunity there. We like Japan, for example, is also another big beneficiary of AI. And so there's elements to be had in that part of the market as well.

22:56Yes. It's getting help today by the weaker yen. People didn't necessarily expect that with the rate hikes. Greg, thank you. It's good to see you. Always good to have you. Greg Kalman, who's the head of public investing at Goldman Sachs Asset Management. Coming up this morning, Apple's new iPhones go on sale today. We'll talk about what to expect for this first launch under new CEO John Ternus. Plus, Dr. Vin Gupta is back. Why he says AI should be a co-pilot in health care, not an autopilot. We're back in a moment.

23:27Hi, football legend Clay Matthews here. Caesar Sportsbook has completely transformed their app. Better design, new ways to bet like flex parlays. More rewards, smoother UI. It's barely recognizable. So bet with Caesar Sportsbook, the app that's been nipped and tucked to perfection. Must be 21. If you or someone you know has a gambling problem, crisis counseling and referral services can be accessed by calling 1-800-GAMBLER. Ever notice how every new AI tool asks you to explain your business to it all over again? Google Workspace with Gemini skips that. You get Google's AI built right into the premium versions of Gmail, Docs, Sheets, and more.

24:09No new tools, no extra logins. Just one subscription that gives you personalized AI right inside your actual workflow. Make work less work. Try Google Workspace with Gemini at goo.gle slash workspace with Gemini. Building a portfolio with Fidelity Basket portfolios is kind of like making a sandwich. It's as simple as picking your stocks and ETFs, sort of like your meats and other topics, and managing it as one big juicy investment. Now that's pretty good. Learn more at fidelity.com slash baskets. Investing involves risks, including risk of loss. Fidelity Brokers Services, LLC. Member NYSC SIPC.

24:56Welcome back. Apple's new iPhones go on sale today. Mackenzie Cigalos joins us now with what to expect this weekend. Morning, Mac. Hey, good morning, Carl. So John Ternus spending his first iPhone launch day as CEO at Apple's Fifth Avenue flagship store, meeting customers and showing off the new lineup. It's a significant test for Ternus with the iPhone 18 coming up against an exceptionally tough set of comps. iPhone revenue has been up more than 20 percent year over year for three straight quarters on the strength of the 17 super cycle. So even with early demand indicators looking encouraging, the 18 has a very high bar to clear.

25:32Apple's trying to keep that momentum going by pushing buyers further up market. There's no regular iPhone 18 going on sale today. You've got the Pro and Pro Max, which starts at$1 ,200. Apple's first foldable, the iPhone Duo, comes next month, starting at$2 ,000. And then the new lower-priced phones aren't expected until the spring. That leans into where buyers are already spending. IDC projects that the Pro and Pro Max accounted for 54 % of iPhone sales in the first half of this year, really helping to lift the average selling price as Apple tries to protect margins. Now, the 18 Pro does get better cameras, battery life of some 45 hours and new colors, but no major redesign.

26:14And some of Apple's highest spending customers may just wait for the duo. That makes the foldable perhaps a more consequential demand test and potentially the better read on just how far Apple can push the iPhone up market. China's especially important year, too. It has been a major part of the iPhone rebound, and Apple is now bringing its first foldable into a market where Huawei has already helped make the category mainstream. That Apple-China relationship really takes center stage next week. Tim Cook expected a President Trump's state dinner for Xi Jinping in D.C., which would be Cook's first appearance in the executive chairman role.

26:51Guys? Yeah, Tim Apple, as the president likes to call him. So how has the business in China been performing? both on the demand side and the supply side, which have both been tricky? Well, I mean, in terms of the demand side, starting in the December quarter, so basically our first read of how the iPhone 17 was performing in that market, we saw this blowout, 38 percent growth, and then it's been north of 20 percent for the past two quarters. And really what was a departure from the previous two years where Apple had really struggled to sell its iPhone models into China. And what's so fascinating about the duo, the foldable, China is probably the biggest market.

27:29four foldable phones. And so you've seen Samsung and Huawei do quite well with their trifles and their foldables. And now Apple gets gets into that market. And so the bet is that's going to be a big seller for them. And it's going to potentially eclipse what even the 18 Pro and Pro Max are able to do this cycle. The standing question and what I'll be interested to see is whether or not Cook makes hay next week with President Xi Jinping in terms of Siri AI, because at this point, There's no clearance for CRI to launch either in China or in Europe. They're going to need the buy-in of the government.

28:01So this is really Cook's first test as chief diplomat, as far as I see it. Yeah, as well as general tariff policy and, you know, overall, whether or not that gets put on hold until after the summit and whether it happens after that. It's going to be really interesting. Stock, though, for Apple, more than double the S &P gains for the year, Mac. Thank you. Mackenzie Sigalos. Time now for a news update. Seema Modi has it for us. Morning, Seema. Sarah, here's what we're watching. an ICE officer pleaded not guilty to federal charges that he lied to the FBI about a non-fatal shooting during January's immigration crackdown in Minneapolis.

28:35Prosecutors say Christian Castro shot a Venezuelan man through the front door of a home, then falsely claimed he had been attacked with a broom handle and snow shovel. Castro also faces state assault charges and has been suspended without pay. A software defect lasting just a millisecond, causing Britain's air traffic meltdown last week, forcing the cancellation of more than 2 ,000 flights and affecting hundreds of thousands of travelers. Investigators say human error was not responsible and temporary safeguards are in place. Britain's aviation regulator will conduct an independent review. And Princess Diana's iconic revenge dress back on display before heading to auction.

29:11Diana wore the black off-the-shoulder gown in 1994, the day Prince Charles publicly admitted to infidelity. The dress will tour London, Hong Kong, Geneva before his December sail in New York, where it could fetch$300 ,000. Expensive address. Carl, back to you. All right, Stephen, thanks so much. Still ahead, Warren Buffett stepping down as chairman of Berkshire Hathaway. We're going to talk to Ariel's John Rogers about what that means for the company. Plus, shares of Netflix are falling this morning after Wells Fargo downgrades the stock. Details of the call coming up back in just a moment. Welcome back.

29:44Warren Buffett announcing today he's stepping down as the chairman of Berkshire Hathaway. and we'll move on to an emeritus role, saying father time always wins. His son Howard will succeed him as chair. Joining us to discuss Buffett's departure and what the future for Berkshire may look like is Ariel Investments founder and co-CEO and former Berkshire shareholder John Rogers. John, appreciate the time. Great to talk to you again. It was great to talk to you, and I still own Berkshire in my personal account that I've owned for years and years. How do you think this next chapter goes? To what degree is this a real shift for the company and its strategy?

30:21Well, Warren is a genius, and he's had such extraordinary success, understands governance better than anyone. So if anyone could put together a team of management and a board to make sure that his values are executed going forward, there's no one better, no one has more capability to do that than Warren Buffett. We were talking this morning with Mike Santoli about the portfolio, and we were thinking back to Greg Abel earlier in the year, suggesting that even the Alphabet purchase was initiated by Warren himself. Do you think further large ideas need to still come or be passed by him? Well, it would be great if they were able to do that because, you know, this is a hard, hard business.

31:09Markets are really efficient. there are few people in the world that are gifted to be able to see the future with the way and the kind of clarity that Warren Buffett's been able to see the future. So hopefully Berkshire can still lean on him for those visionary decisions. Does it change the thinking, John, at all around the valuation? Beyond the chairman role, the CEO role as well. I mean, you always had the Buffett premium in there. Well, I think the Buffett premium will naturally start to be less and less. as time goes on. And I know that that's human nature, as Warren said. You know, time passes, people get older, and so the market won't be, they realize he's not in the game full time, and so that Buffett premium over time could probably dissipate.

31:58But again, he's done such a great job putting in the right board, the right management leadership. I still think there'll still be somewhat of a premium left. Are you anxious for them to put some cash to work here? Well, I think it's been important that patience wins over time. And I think the idea that he understands and they understand the market has had a very, very strong run. There's a lot of enthusiasm in the market, especially around AI. So having some cash on the balance sheet is a good thing. to be able to take advantage of the inevitable correction that one day will happen. What about changes in the portfolio?

32:41I mean, does the prospect of, I don't know, divestitures or spinoffs or just does the same portfolio make sense in a post-Buffet era? I think the same the same plan makes sense. Again, I know Warren has thought deeply about this. He's always talked about how succession is so critical in whatever industry or whatever company you invest in. So I know he's been extremely thoughtful about this. And so I think you'll see the game plan stay in place for the foreseeable future. We were talking about the Taylor Morrison purchase and sort of where housing stands at the moment. I know Abel was on with Becky a few weeks ago and said, you know, it's kind of a slog here with mortgages 7 plus.

33:25But I wonder if you think you're beginning to see the the inklings of a long term value strategy when it comes to housing in this country. Well, I am very bullish on housing in this country. I think, you know, eventually rates will go down again. We'll get to a lower mortgage rate. But they are still there's just so much demand for housing out there. There has been a shortage of new houses being built for quite a long time now. So I think there's real opportunity in housing-related companies. It's kind of a thing we spend a lot of time looking at at Ariel. I think of you as a quintessential value investor and just thinking about Buffett stepping down.

34:07And I just wonder where we are in the cycle or in the investing period right now for value stocks and value investing. I think it's a great, great time for value investing. I think there's a lot of neglected, kind of boring stocks that are extremely cheap because of all the excitement around AI. And even Morningstar had a story recently asking, is value investing dead? And that's always a sign you start to feel like that maybe there's going to be real opportunity and value and being able to refocus in that area. It's going to pay off. I wonder how you think, John, AI, the revolution changes and the stocks around it changes value investing.

34:53Has it made it more powerful or harder to figure out? I think eventually, just like when the Internet bubble burst, there were some spectacular successes that came out of the Internet, companies that did really well. But a majority of them that were the hot stocks of the day, you'll never reclaim their highs. And if you were stuck to being a value manager during that period, you got rewarded when the Internet bubble finally burst. I think the same thing is going to happen this time. When the AI enthusiasm starts to wane, value investing will be popular again. Finally, John, you know, I'm thinking of Warren.

35:30I'm thinking of Charlie as well and what he used to say about crypto and we're in this era of prediction markets and options volume and 24-7 trading and tokenization one day. Do you think Warren's and Charlie's sort of mantra of value, of patience, of compounding can outlive these secular new trends? I do. I think I think for sure. You know, Charlie used to always say that, you know, your portfolio shouldn't be like Noah's Ark. You own two of everything. I think concentrated portfolios are a winner going forward. Not looking like the index is going to be a way to be successful. And I think you're going to realize that people have poured too much money into the indexes and active management is being misunderstood and disrespected in this environment.

36:19And that helps value managers. John, always good to get your point of view. Very valuable to viewers as we try to sort of process the Buffett news today as he moves on to Emeritus' role. Good to talk to you, as always. Thank you. John Rogers. Still ahead, Dr. Vin Gupta joins us on the growing role of AI in the medical field. why he says AI should be a co-pilot in healthcare, not an autopilot. Walk on the Street, we'll be right back. Welcome back. As debates rage around AI's impact on humanity, frontier companies are pushing deeper into medtech. In the latest move, Anthropic is reportedly setting up a biology lab as it ramps up its AI drug discovery program, which includes a partnership with Novo Nordisk, announced earlier this week.

37:01For more on how AI could help and hurt medical research, let's bring in Dr. Vin Gupta, MSNOW Senior Medical Analyst and managing director of Manat Health. Dr. Goupa also previously served as chief medical officer at Amazon Health. Welcome back. It's good to see you. Thank you for having me. So this is a tricky issue, right? Ethically, I mean, there are all sorts of questions now that we're worried about AI improving itself and being dangerous. How do you look at it from the medical perspective? Yeah, you know, Sarah, I think, unfortunately, it seems like leading lights in medicine have, it's a binary conversation right now.

37:35Either AI will replace doctors, is what some people think, by 2030, or AI has no place. It's been a very binary conversation. But I would say that we should be moving fast when it comes to AI adoption, thoughtfully, and healthcare. You just pointed out a very important element of this, which is drug discovery. Anthropic and Novo, looking at biological targets that are promising more quickly. I mean, we're talking about reducing R &D spend, Sarah, from a biopharmaceutical standpoint, perhaps 10x, and then accelerating it to the point of months, if not years right now. It takes years on the initial target identification stage to find a promising therapeutic.

38:13AI can really accelerate that. So there's a lot of promising areas here. In addition to what's provider-facing, documentation burden, of course, clinical decision support, we're seeing a lot of different companies here doing workable, interesting things here that are advancing patient care. We should be moving fast in these areas. What about when it comes to diagnosing a patient? Should it be first opinion or second opinion? I think it should be a second opinion. But, you know, Carl, I would say one of the biggest challenges we have here are health care deserts. One in two people do not have access to same day or next day care in the United States.

38:44So when we think about that as the comparator, not everybody has a PCP that they trust. If that was the gold standard, fine, we're having a different conversation. But one in two people don't have durable access to health care. So I like it as a screening tool, as an initial triage point with human escalation. And we're also seeing, I would say, the Trump administration has been very forward-thinking when it comes to using AI and building AI models into medical devices, like perhaps the one that you're wearing on your wrist or on your finger. Now we're actually using large language models to build in and interpret data that we're accruing on our bodies 24-7, 365, rendering potentially, in some cases, a diagnosis for a patient that might have high blood pressure.

39:24A hundred million people have undiagnosed high blood pressure in the United States of America. We can do something about bending that burden through thoughtful application of AI, but there should be human escalation when appropriate. And you think data, privacy, HIPAA stuff can all be ironed out over this next couple decades, I guess? A hundred percent. I mean, HIPAA, all these data stacks are HIPAA-hardened right now. We have not seen one instance so far. It's a fear, of course. Security is important. Data security is really important. But there has not been one instance of a data privacy breach when it comes to AI applications in health care.

39:59Although, I don't know, I keep thinking of Alex Karp yesterday saying that all the companies, the enterprise companies, are so angry at the LLMs for sharing their data and their competitive advantage. I guess you mentioned the Trump administration, Dr. Gupta. What is the government's role here? What should it be? Yeah, well, you know, I think they are creating what they've done really well is they're creating sandboxes across the country. Different states like Utah, for example, Sarah, they're doing automated prescribing of chronic medications for psychiatric conditions. In a world where one in 10 people don't pick up their refills, three in 10 don't want to wait in the pharmacy for their refills, if we can automate things and make it easier and send those refills directly to your home, and you don't have to think about it because it's safe to do so.

40:42That's the state government doing that in Utah? That's the state government doing it with federal support. And so I think the government here at the federal level has encouraged thoughtful innovation with safety guardrails in some cases. But I would say the critics of their approach are saying that perhaps there are no guardrails whatsoever. And so how do you find that balance? They are forcing a conversation. There are some state-based approaches here that are interesting. Again, automated prescribing of chronic medications is one. But again, some people want to take this further. By 2030, we're going to automate the diagnosis and we're going to automate the prescribing.

41:17I think that goes too far. There needs to be a human in the loop at some point as we're learning. But do I think a human has to be in the loop forever? No, I think a human, when it matters, ultimately is going to be important. But this notion that everybody needs a human doctor in the loop, say, in 2035 when these models get better for low acuity conditions. No, I don't think that's necessarily. And patients, too. Well, a new generation of patients will be born and will be raised in an environment where you don't always see a human being. And they might take a diagnosis from a machine. I couldn't agree more.

41:49I think the 35 to 54-year-old generation, the sandwich generation, is open to a diagnosis using an LLM, ideally with a human in the loop right now as we're learning. But I can see in 10 years saying, you know what, if I have pink eye, if I have, say, a positive flu test and I feel like I have flu, Give me that Tamiflu and make sure that there's guardrails here using technology, but that I don't need a human loop at that point. I could see us getting there for lower acuity conditions. But this notion, again, that it's binary one way or the other and that there's no middle road here, to me, that's wrong.

42:21Yeah, no, I think that makes a lot of sense. What would scare you? What would make you nervous? I mean, one of the risks that Dario talked about from Anthropic, Dario Madej, is bioweapons, for instance. So where do you think we could go too far? Well, I mean, you said I think the first here is are these are there guardrails when it comes to, you know, we're identifying therapeutic targets. What happens when we're talking about artificial intelligence companies being used for national security purposes and bioterrorism, biopreparedness? Are there guardrails there? We've obviously lived through 2020 and potentially some concerns on lab leaks.

42:56We don't want something similar. So, yes, Sarah, I think that's top of mind. Two, let's be candid about the state of the technology today. It has GPT for health. It was just studied. There was an M &E study on this just published a few months ago. It under triaged real emergencies more often than not. Over triaged non-emergencies more often than not. We can't have that. And there's been some critics critiquing that data and saying that perhaps it was flawed. But there are concerns, especially when it comes to mental health emergencies, that it's not ready for prime time. So I think we should be careful here about deployment of these tools for emergencies.

43:29focused on lower acuity conditions. And lastly, I would just say insurance utilization. We don't want to get to the point where we're having a chatbot deny people care and not having a human in the loop to review cases. And so I think there is concern about how will insurance companies and payers leverage LLMs for utilization management and claim styles. And to me, I think that has to be moderated. I love talking about the real world examples, especially the promising ones. Dr. Gupta, thank you. Always good to have you. Ben Gupta. Still ahead, Netflix shares sliding after Wells cuts to underweight with a 57 target.

44:05We'll break down the call after this. Welcome back. Shares of Netflix sliding after Wells cuts the stock to underweight. Our Julia Boorstin's got some details. Morning, Julia. Good morning. That's right. Shares down nearly 5 % after Wells Fargo downgraded to underweight and cut its price target for the stock by 29%. Wells flagging concerns about weakening engagement, estimating hours watched per subscriber are down 4 % year over year, including a more than 20 % drop in viewership of Netflix originals. Now, Netflix has been investing in the likes of podcasts and bringing YouTubers over to its platform to catch up with YouTube, which surpassed it in terms of U.S.

44:45TV viewing time back in 2023. Wells Fargo saying, quote, engagement trends look worrying to us. If the opportunity is to recast Netflix into a broader content hub, the risk is missing the water cooler originals. We see breakout hits as a must for the stock to work again. This comes as Netflix spends more overall on content. $20 billion this year, which is$2 billion more than last year. Wells saying that the original content slate is weaker than prior periods. And as Netflix promotes title diversity over title concentration, it does miss creating those water cooler moments. We will, of course, learn more when Netflix reports in about a month.

45:27But it is worth noting that Netflix will pull back on transparency around engagement. It's only going to be releasing its engagement report once a year rather than twice a year. And the stock is now down over 40 percent in the past year. Carl? Julia, we've been watching shares at Disney today, Comcast and Alphabet. Is this all a story about everyone losing to YouTube? It's an engagement story. And right now, when it comes to engagement, we are seeing YouTube winning. It's so interesting looking at all these players really lean into the short form video. They call them verts on Disney and ESPN.

46:05They've been doing a lot of these short form vertical videos to make sure there's a lot of content for people to watch. on those Disney Plus platforms and ESPN. We're seeing Netflix lean into it as well. And when it comes to the engagement game, and of course, engagement is really important when you talk about advertising, which is Netflix's new business. It's all about engagement. They just need a new season of Bridgerton or something. Thank you, Julia. Julia Boorstin, big week with the leaders. She in town next week, NUN General Assembly. Shana Tava. Thank you very much. See you next week. Very good.

46:39All opinions expressed by Squawk on the Street participants are solely their opinions and do not reflect the opinions of CNBC or its parent company or affiliates, and may have been previously disseminated by them on television, radio, internet, or another medium. You should not treat any opinion expressed on this podcast as a specific inducement to make a particular investment or follow a particular strategy, but only as an expression of his opinion. Such opinions are based upon information Squawk on the Street participants consider reliable, but neither CNBC nor its affiliates and or subsidiaries warrant its completeness or accuracy, and it should not be relied upon as such.

47:10To view the full Squawk on the Street disclaimer, please visit cnbc.com forward slash Squawk on the Street disclaimer. Building a portfolio with Fidelity Basket portfolios is kind of like making a sandwich. It's as simple as picking your stocks and ETFs, sort of like your meats and other topics, and managing it as one big juicy investment. Now that's pretty good. Learn more at fidelity.com slash baskets. Investing involves risks, including risk of loss. Fidelity Broker Services, LLC. Member NYSC SIPC.