What AI risk warnings mean for software stocks, tech leaders meet Trump, Nvidia’s buyback

[music] Hello and welcome to market domination. I'm Josh Lipton live from our New York headquarters. There is just now to go now to the closing bell. Stocks are mixed here as bonds continue to sell off [music] and AI jitters escalate. Yahoo finances in feray has the very latest. Nez >> Josh mixed stocks and off the lows of the session I should add. If we take a look at the Dow right now, it's down about 2/10en of a percent, but really trimming losses over the last half hour to hour, I would say. Also, the Nasdaq 100 uh right now just above the flat line. And then the S&P 500 is wavering just around the flat line again off those lows of the session as we have seen bond yields relaxing just a little bit. We did have did see a climb in longdated bonds with the 10-year Treasury moving up and also the 30-year moving to its highest level. That yield to its highest level since 2002. You can see it there at 560. The 10-year at 526. Make no mistake, bond yields is really where Wall Street it what it Wall Street is eyeing right now as we have seen these yields going higher. The market hasn't fallen off a cliff yet, but certainly it is raising eyebrows on the street. Taking a look under the hood, we are watching energy that's under pressure. This is because oil is down today. Also, materials is down and consumer staples, but you've got utilities that's higher and consumer discretionary higher as well. Over on the NASDAQ 100, a little bit of a mixed picture here. You can see on the lefth hand side of the screen, Apple, Alphabets down, Nvidia, but then you've got Amazon, SpaceX, Meta that are higher. And then just taking a look at crypto, we have been seeing Bitcoin up about 210 of a percent over the last 24 hours, hanging around that 84 or $83,000 level per token. And as I mentioned with commodities, a surprise that you saw bond yields going higher today, yet you are seeing oil going lower. Usually when you see oil going lower, you see uh those yields e easing a bit, but look, you've got WTI that's down more than 3%. Brent crude also down about 2% for the session, Josh. >> All right, thank you NZ. Appreciate it. Well, fears around AI risk mount. President Trump is meeting with tech leaders for a crucial AI summit. The CEOs of Meta, Nvidia, Palanteer, Google, and Anthropic were all in attendance. Yeah, our finances Ben Worko is tracking all this. You know, Ben, I mean, first of all, just the the visual of that, Ben, was uh that was something else, man. I mean, some of the biggest heavyhitting Texios on the planet, all just standing there. It was impressive to look at and see. What What do we know about was actually being talked about there, Ben? What were the What were the discussions? Well, well, we know from the discussions, President Trump just spoke briefly to reporters a couple minutes ago and underlined his position here that there is no interest from his administration in any sort of government regulation of even with the growing worries of AI danger. The term of our from the president um just a few minutes ago was self-regulation. He said there was agreement between his team and all those CEOs that were gathered with him that there should be in his words tremendous self-regulation i.e not getting involved. And as the president sort of had almost a dayong kind of bear hug of he had this meeting with CEOs and he's talking about releasing an executive order to later this afternoon further on the issue just underlining his focus here that AI needs to grow unfettered partly in terms of a rivalry with China but partly just because he's supportive of the technology as as he pushed forward. As you mentioned there was a lot of CEOs in attendance. President Trump was flanked by Elon Musk and of of Tesla and SpaceX and then Jensen Wong of Nvidia on the other side. A range of folks around the table there, but Trump kind of arguing that they all were in agreement about limited. Notably, Dario Amodai was was at the table there as well. No word on what his involvement there was. He of course set off a lot of this debate talking about pacing the frontier, but President Trump definitely was trying to show a unified frontier, CEOs and his administration unified on kind of AI full speed ahead and that they can handle the dangers themselves. >> And we do have a vote scheduled in the Senate tomorrow. I think I guess this is a bill around pushing data centers to to pay more to local communities. What would that do and and what are the prospects there? What do you think? >> Absolutely. Yeah. So this meeting today, the sort of formal agenda of this meeting today was about beginning a conversation about AI regulations. Speaker Mike Johnson was also in attendance and he called it. That is a much longer term and based on President Trump's comments right now just uh unclear what what if anything Washington wants to do on this. There is this vote you mentioned tomorrow. It's called the rateayer protection act. This is about local communities and the and the data centers. The worries that they're going to drive up prices and drive up electricity bills. that what this bill would do is basically push local utilities to consider data centers in their rate making. Um what what it's it's got an uphill battle partly because the critics say that it just doesn't do enough. This is being pushed. Democrats are opposed to this bill saying this is the word they use is voluntary. The word they call it they call it a fraud suggesting it's going to have a hard time getting to 60 votes but there is a vote scheduled tomorrow. We've seen a variety of comments in the last few hours. Tom Killis of of North Carolina as an as one example talks about how he didn't think this bill, even if it passed, would do a lot. And moderate Democrats echoing the party line that they they're not they're not on board with this bill. It's going to have a hard time getting to 60 votes, but this is kind of the only thing that that Washington can move on at this moment on AI and it seems to have a hard time getting past tomorrow for certain. >> All right, Ben, thank you. Appreciate it. Anthropic IPO perspectives leaked according to reports from Reuters. the filing revealing yet more warnings about AI risk referred to quote existential risk to humanity. How are tech investors inos thinking about all this? Join me now. I got Pat Wal Ravens, citizens head of technology equity research. Pat, I'm actually going to call an audible here because let's talk about open AI that that's making headlines right now as we speak. Uh, one big headline, I don't know if you saw this dots, Pat. All right. These are AI agents. They're going to work on tasks for you. I guess Pat connect to it sounds like thousands of apps. I'm just curious. You saw Sam Alman make that news, right? Did you think Pat, well, that's a real opportunity for software. That's a risk. Maybe that's both. Talk me through it. >> Uh, the first thing I first of all, thanks for having me on, Josh. It's great to be here. >> Uh, the first thing I thought was um, >> so I started with Grockbot. So, I do have Grockbot on my phone. I connected it to a phone number, which by the way is a fascinating thing. >> What do you think so far? You impressed? >> It's fantastic. It's fantastic. The very first use case I had of my Grockbot with a phone number and there's a company called Bandwidth that you can get a phone number for it, right? Was I was running late and so the the you're at the top of your Grotbot there's a a chief of staff. So you have my chief of staff, right? So I I I tell my chief of staff, >> please call Aaron who I'm having dinner with. Let him know I'm running a half an hour late. Um have him order me a particular cocktail and if the waitress comes around for um our orders, get me a nice piece of fish. And then I'm off and running doing things I'm not really thinking about. I get to the restaurant, there's a cocktail sitting there and I go, "Aaron, did you get the call?" And he goes, "Done, Mr. Yeah." And he goes, he goes, "Yeah." And I go, "What did it say?" It said, "Um, hello. This is uh this is uh Pat's Grockpot." Sorry, Grockbot, chief of staff. I'm calling to let you know he's running late. Can you order him a drink? And if he if the waitress comes around, get a I go, "That's fantastic." And you can also have it um >> ask questions. So, when I'm, you know, I'm in New York today and I've been seeing clients and I demo it for my clients. I say, "What's your mobile? I have I have my my chief of staff call them and say, you know, Pat really enjoys doing business with you. He appreciates our relationship. Will you come to the Citizens Technology Conference in March, right?" And they either say yes or no. And then I get a text back on my phone as to whether or not a client is going to attend our technology conference. >> So, this is saving you what time? It's time. It's energy. >> It's so great. It's so great. It it um it saves time. It it it really smooths out communications with people if they're willing to pick up the phone. Okay. So, >> all right. So, you have Grockbot and get back to your dots. You have Muse, right? Now, we have Dots. So, there's more and more of these consumer agents. Um you can't get them all to do a phone number yet. The phone number is really fantastic. >> Um our telephone network is going to be burned up >> by agents. the next billion users of the telephone networks are going to be agents, right? So, we got we got a ways to go to figure that out. People aren't going to pick up the phone over and over again when they have an agent calling them. Your agent is going to start answering the phone from another agent, right? And how all that is going to get resolved, we don't know yet. And it's moving so fast. The techn the big takeaway from all this for me is the technology is moving faster than the human's abilities to absorb it. Yeah, when you watch what was going on at the OpenAI dev day today, >> I mean, it's just remarkable, but it's so much. And the same thing is true in enterprises, right? It's just it's going faster than the enterprises can safely implement. >> Well, I want to get to that. A question I had as you were talking there, I was thinking, Pat, if you have these agents and they can kind of work across dozens or hundreds of thousands of apps, who ultimately owns that customer relationship? Like, is that Sam Alman? Is it the software company? How do you think about that? >> I own that customer relationship and they and and the I can tell you who's going to win. You want to hear who's going to win? >> Yeah. Give it to me. Who takes the gold? >> Yeah. >> Who takes the gold is who delivers a product that I trust that is safe. >> And the and the more we get that message across the better. Right. Because uh >> what is Can I just ask you when you say safe? Like what does that actually mean in practical terms? What is safe? I mean, when you read the disclosures from some of these AI labs about, you know, we were training something and it's a future version of the product and we were we and we had to do this evaluation, it broke out of the evaluation and it and it and it it hacked into hugging face in order to try to get the answer to the that is not safe. Okay, that is not safe. And um and so we got to slow down so that we can um put guard rails around these products. Well, now you now you're into a hot debate here and you find yourself into it on at odds with you know the president of the United States and others by I think I think Jensen Wong's with him. I think Mark Zuckerberg they'll say you know what um we have we have guard rails. We have safety in place. It's called liability. You you screw up you screw up Sam Alman. You screw up Dariel Maldi. And you know what? You've got verdicts and settlements and insurance costs. You're going to get wrecked man. What do you say to that? that that they argue that's your guard rail. >> I think the product liability laws are going to have an incredibly important role to play here. >> And I think the last thing we want to do is create some sort of regulation that exempts these companies in away from those laws. And those cases are coming. I wrote about it recently in one of the Microsoft notes. You might have seen it, you know, but there's this this >> horrible horrible situation with the family that lost their son, right? that was engaged in this really dark back and forth with Chad GPT. >> And you know, I remember those headlines. >> Yeah. And and in in OpenAI's defense, it it said like a hundred times, get help, get help, get help, get help. Right? Not in OpenAI's defense. When you read some of the things that chat GPT was saying to this this kid, you're just like, I mean, that's insane, right? Those are crazy things to be saying to some teenager, you know, for those of us who have teenagers. I guess I guess the question then Pat though is you know it's easy to talk about well you know you need you need regulation but then to your point I don't know what the smart targeted regulation looks like that it'll catch a defective dangerous model but it won't like freeze an industry in place that could be enormously economically valuable and I'll be honest Pat I don't know if I perfectly trust Congress to get that right I mean we've all watched Congress like try and grill executives from social media. Does that leave you with a ton of confidence, you know? Yes. Those are the folks we should put our trust in to get this, you know, bleeding edge technology, right? >> We want to try to set up the incentives so that people do the right thing. Yeah. Right. And one of the incentives to get people to do the right thing is to realize if they don't do the right thing, there are consequences. And those consequences are civil. >> What happened to all the companies used to make asbestos? >> Right. >> Yeah. >> Right. But then those consequences can also be criminal. Right. And so when you when you make it clear, you know, what the consequences are of your behavior, >> um that hopefully will help push things uh in the right direction. >> I guess if I'm a software investor, the bigger picture question, let's see, Pat, we're talking 12 months from now. I hope we are. I mean, what would tell you all these agents, whether it's Open AI or it's it's Meta, like what would tell you what would be the indicators? Hey, you know, this is actually this has proven to be a real tailwind for publicly traded software names or yikes, we got a problem. It's a risk. It's taking profits. What what should I be watching for? >> I mean, I think big picture. Yeah. >> And I cover enterprise software. We have a fantastic internet analyst, Andrew Boon, who you can come on to get all the consumer side of it. But big picture in the enterprise, look at what Microsoft is doing, >> right? I we we use co-pilot at citizens bank and >> um it's fantastic and it has um access to all of my emails and all of my contacts but at the same time it doesn't give me access to things that are elsewhere in the bank that you know the software equity research analyst most certainly should not be seeing right and I am now spending 70% of my day in that product >> to learn to write notes to create cool graphs right and to prepare for my conversations people like you and with my clients So, um, you know, we have some examples of where it's working. I think I think what Microsoft is doing with Copilot is one is one really interesting one. >> Let me before we go, I want to squeeze one ticker in there because it was dominate headlines. MongoDB. You covered that name. I know you like the name. Um, so what did you make of that news? The co leaves. He's going to work for Zuck and Meta, the stock. I mean, I want your take on the news, Pat, but then did you look at that stock reaction, man, and think >> down like 20%. >> Rocks, right? Did you think that was fair? Was that overblown? What do you think? So I was at the analyst day today and um >> uh >> the timing of this is amazing. So the it's really difficult. So analyst days today in New York >> on Thursday >> uh CJ Desai who is by the way he's at service now before who was the the the the president CEO and a board member of MongoDB notified the board that he was leaving right and he's taking the uh enterprise platform role at mega at meta. It's incredible >> this new business AI platform that Zuck. Yeah. >> Yeah. um on Thursday. >> The board meeting happened on Saturday. >> Yeah. >> Right. Yeah. >> A weekend [clears throat] board meeting. And then luckily for MongoDB, they have Dave who uh has been the CEO for the past 11 years and has been gone almost a year, right? Uh who's still on the board who is able to step right back into that role on interim basis. And he presented this morning um at the at the the analyst day. >> And were you impressed by what you heard in terms of growth, strategy? Yeah. Yeah, >> he's fantastic. He's always been fantastic. He's fantastic again. The only thing you would have liked to have heard is I'll stick around for another 5 years and he's not signing up for that. Right. So, the challenge now is it's a very good business. It grew 30% last year. There's a lot of tailwinds. But the challenge now is we have Dave for this interim period, but now we got to go find a recruiting, you know, a head hunting firm. Got to find a new CEO and then we got to go through that process all over again. That is not optimal. >> But still, bottom line, it's an outperform. You gota you're telling your clients this is a buy here. >> Yeah. And down 20%. That is an opportunity that you should take advantage of. >> Pat, so good to see you, especially on set, man. Thank you. It was great. >> It's my pleasure. It's nice to be here. >> Coming up, we dive into the latest [music] Fed speak. That's next on Market Domination. [music] [music] >> [music] [music] [music] [music] [music] >> Heat. [music] [music] Heat. [music] >> [music] [music] [music] [music] >> Down. Down. >> [music] >> Down. [music] >> [music] >> Heat. Heat. [music] >> [music] [music] [music] >> Heat. Hey. Hey. Hey. >> [music] [music] >> Big batch of Fed commentary coming out today. Our fin correspondent Jennifer Shawnberger is here with the details. Jen, >> hey there, Josh. Yep, lots of Fed speak this afternoon. We just heard last hour from New York Fed President John Williams who sort of laid to rest the notion of an October rate hike saying that quote there's no need for urgency and we have time to gather more information. He sees only one more rate hike. He says what he calls later this year, which to me speaks to December time frame rather than later next month to return inflation in a timelier manner back to the Fed's 2% target. Now, we also heard from Federal Reserve Governor Michael Bar earlier today, who seems to be a bit more hawkish because he sees further interest rate hikes, plural, to get inflation back down to the Fed's 2% target in a more timely manner. He is looking at the same forces that John Williams is looking at but seems to think that the Fed needs a more aggressive approach and that is that higher energy prices seem to be sticking around for longer and could have far more farreaching ripple effects. Also AI very much on the front burner sort of replacing tariffs as a focal point for near-term inflation for the Federal Reserve. We also heard from Chicago Fed President Austin Goulby who says, "Listen, I'm one of the more optimistic members of the Fed when referring to the DOT plot implying that perhaps he doesn't see that many rate hikes, but he says we really need to make sure the inflation is coming back down to target and these things that seem to be temporary are actually temporary." He still thinks the Fed could cut rates at some point, you know, maybe 18 months from now. Back to you, >> Jen. Before you go, the Trump administration launching this this AI tool. What can you tell us about that? >> That's right. The Trump administration launching a brand new website today called america.gov powered by artificial intelligence. Before he had to go to so many different government websites, go through multiple steps to apply for a passport, apply for Medicare, compare health care plans, apply to change your name if you just got married. and you had to fill out all these forms, go through multiple steps, maybe go to an appointment and wait on the mail. Well, now powered by two AI agents, Grock and Gemini, so Google and SpaceX's AI, users can now go to one website, America.gov, and ask questions in plain English, like, I'd like to apply for a passport, or I got married and I need to change my name. And that technology is going to comb and link all of the government websites into one. This has been developed by Joe Jebia who was co-founder of Airbnb and now head of the White House's national design studio. And this website is seen as sort of a concierge for American citizens to the federal government. And Jebia says he's confident that this website is going to operate well, that we're not going to see some of the infamous uh missteps that we saw, for instance, with healthcare.gov because it has been formed by a co-founder of Airbnb and has been in partnership with SpaceX and uh Google. Right now, you can only search for questions on how to do certain things, but early next year, 2027, they expect this is going to be fully operational. you can actually fill out forms, have tasks completed. Um, so you no longer have to go through that fussy process if you want to get your passport. You just go to america.gov and you fill it out. And indeed, Secretary of State Marco Rubio said earlier today that that is where where they want the majority of Americans to apply for their passports going forward. So, I don't know, Josh, it's it's pretty cool if you check it out. >> All right, great stuff as always. Thank you, Jen. >> Coming up, how much runway is left in AI trade? We dive into [music] that next on Mark Dominated. [music] [music] >> [music] [music] [music] [music] [music] [music] [music] [music] [music] >> Heat. Heat. >> [music] [music] >> Down. Down. [music] >> [music] >> Heat. Heat. >> [music] >> Down. [music] >> [music] [music] >> Hey, [music] [music] hey, hey. >> [music] [music] [music] [music] [music] [music] >> Open our eyes. [snorts] Stepping up its push in enterprise AI with new agents called dots creating a more direct race with meta. For more we want to bring you down. Ben beharin co and principal analyst creative strategies. All right Ben let's just start right with that openi developer conference. Talk to you about dots. These are these new AI agents. What do you make of it Ben and what does it mean for meta which openai seems to be you know taking a direct shot at? Yeah, I mean I think across the board the there's interest in these let's just call enterprise or personal um agents which actually we think is very helpful in uh eliminating some of the skill gap uh being able to have something that essentially looks at what you're doing or anticipates what you need and can kind of help you use AI a bit more effectively. So we think there's a lot of upside there. We don't expect you know one agent to win them all. We think there will be a handful either in an enterprise context which might be attached to your software or your enterprise company platform or even in a personal sense. We think you could see most of most uh several of these be winners. But at the end of the day, it's it's interesting because it's another way I think to get people to use AI more in their day-to-day. And right at the end of the day that also just means a whole lot more compute uh infrastructure needed as we scale these things not just for you know basic agentic tasks all the way down to personal or individual agentic tasks where everyone may have one or multiple of these in in their lives for daily use. I >> I'm curious man which ones are you kind of experimenting with? Like are you are you playing around with Muse for example? >> Yeah we have Muse, we have Instinct. Um we're going to try Dots as as soon as we get access. we've had a little bit of uh of early exposure to that. Um but I mean, you know, Josh, we we we try them all as early as possible, break as much as we can, push their limits. Um, but I think the having something that can go do things for you or anticipate things for you and string those workloads together in an automated fashion. I think a lot of people will find value on and that's why these are the first things I think you can look at and see real consumer value around around AI. um not just shopping but I mean you know being able to book book tickets or do uh fix you know sheets for you or find some you know government document right these are the things where I think people can start onboarding with AI in in a whole lot more meaningful fashion but we're very very early obviously >> what what do you make of folks who say you know I kind of I want to experiment with these agents right I want to give Muse a test drive but I maybe I have u some privacy concerns I hear people talking about that I'm curious what you make of that Um or do you think Ben some people a lot of people just say you know what I will I'll sacrifice privacy for convenience? >> Yeah I mean I think the privacy part is not going away. Um I think there's a a strong case to be made that these things get a lot more valuable when you you know let them log into your services or to some degree give them access. Um, I mean, we don't think everybody day one's going to start giving these things to banking or financial access, but areas where you might use a website or a service, right, to do some shopping or some commerce, um, you're hearing stories come out, right, pretty frequently around it helping people with trouble tickets or customer support, um, getting refunds for plane tickets, things like that. So, there there's a lot of lowhanging fruit, but I do think there will always be that that security, that privacy, and it's going to come down to who you trust, right? Is it a service provider? Is it a vendor? Is it a brand where you are comfortable using commerce or doing banking? Are they the ones that will provide this to you and you need agents to just work across those things? But I think security and privacy, you know, continues to remain um a pretty central, you know, mindset as you pointed out. But like I said, I think there will be vendors who have more trust in the ecosystem that I think can get consumers to use these things backed by consu, you know, security and privacy that's inherent to their platforms. Ben, the other name I want your take on that made news this week. I want your take I haven't heard it yet. Nvidia kind of two big pieces of news, right? Boosting its buyback by 150 billion. Massive. Uh but what about this new Nvidia software, Ben, this AI safety platform? just explain that to the audience, the viewers in simple terms, Ben, and then I I want your response to it, what you made of it. >> Yeah. No, we think this is one of the most interesting things, right? Right now, I look at it as kind of a a conceptual way where we can go and do a little bit more of a of a safety or an agent harness. Um there's a number of ways Nvidia has implemented this, but essentially what it's doing at its core is observing the reasoning of any model and that's being done by a third party independent model model watchdog per their you know nomenclature. But essentially you have an independent model whose sole job is to just look at the reasoning of the agent and then see if at any particular way it starts to go off the rails or might start reasoning that it should go do something that it shouldn't. And again across an enterprise spectrum, this can be set with B very specific enterprise permissions um policies so that it it knows that it has the right to do these things and and do those guardrails. But that is a pretty clever approach um other than just looking at the outcome. Is the outcome okay? But actually, you know, having a observer of the reasoning and doing so with a with a very specific approach is actually a fairly novel approach. And we think conceptually that has legs to have a lot more of these safety harnesses uh that we need that we need from um agentic AI. >> What is your general take then on the kind of AI safety fears that are voiced by open AI and and uh anthropic? I mean do you in general Ben do you hear those and think okay that's fair and true? Do you think it's overblown? Do you think it's cynical? What is the Ben Behar take? >> Yeah I mean I I think it's fair. I mean again I think if you talk to folks in infosc and cyber security you know you hear a lot of very scary uh you know logical conclusions if this does get out of hand. I think you need to have something that has very hard permissions hard policies hard security and the kinds of things that allow um you know these these agents to not go rogue and go hack things or go try to uh get into software that they don't have access to. Again, this is something that can be solved through software, right? And this is one of those things where it's like, oh, we have no idea how to do this. Like, sure, bring awareness to it. Let's make it cognizant, understand the problem, and then start to put guard rails around these things so that we can go and fix them. But the reality is, right, that it can get out of control, and agents do have a lot of power. And a lot of case in a lot of enterprises, agents are running right now ungoverned. And so I think we need to rein that in and have a standard process where we can do better governance, better deployments with these types of security. And then more importantly from a cyber security standpoint, equip governments, equip enterprises to defend themselves against frontier hacks, whether that comes from a model that went rogue or actual attackers that I think becomes a much more significant problem, you know, that an AI enhances over time. >> Finally, Ben, before we let you go, listen, if you want to do AI, you need memory chips. Micron's earnings report on deck. I'm just very curious, Ben, what are you listening for there? Is it ASP's pricing power? Is it just overall demand? What's on your radar? Yeah, not demand. We think demand remains continually strong. We still hear DRAM is structurally uh constrained in in many different ways and that's not going away. Um I think there's two things. I think one we're not we're looking for is gross margins, right? This is a huge debate. Can gross margins stay up uh at the rates that they are or do we need to see some normalization so it doesn't feel like you know price gouging in a movement. So I think that will be super interesting because that's one thing to track over time is to see if you know margins should should marginalize. The other one is just continued conversation around LTAs. I think the debate around memory keeps coming back to how long can this stay structural or does it go back to cyclical in some capacity and LTA are going to be the thing around the confidence of these long-term agreements. how much is guaranteed in capacity that I think will give investors more confidence that this is a much more longer and structural cycle than people realize and we think LTA's you know remain plus some margin uh integration across that remain the best signals to look for for the length of this cycle. >> Ben as always great to have you on the show. Thank you buddy. >> Thanks for having me man. Good to see you. Y finance executive editor Brian Sazi spoke to Dan Ies Yorkville Ives partner and senior managing director about the health of the eye trade earlier today. Here's what he had to say. >> Dan, what do you think about this anthropic? >> Look, I mean this is this is the event everyone's been waiting for. Now look, in terms of the leaks and things like that, that's obviously, you know, a head scratcher, but look, I think it all comes down to like when they file, how investors ultimately play this name because I think the big focus is going to be more about the growth and what the trajectory looks like over the next few years because look, they're at the centerpiece of the AI revolution trade. And you you think about Daario today with Trump and and many other leaders at the White House. Look, it speaks to this tug of war we're going to see between safety >> and growth. >> Dan, you think they come public? >> I mean, I think it's the the probably the same chance that you're mentioning more than the bar. [laughter] Well, I mean, I'm trying to think like why would this even leak if they were not trying using this as some form of trial balloon on a reception on a potential reception in the public markets? And then you've been at this game a while. Who even leaks this stuff? Look, I mean, things have happened in the past like where we've seen some things like this, but look s I think it just comes down to like you go back to the Daario essay, the safety, we have to slow things down. There have been a lot of things that that have gone on where I think it's confused investors because it's sort of talking out of both sides of the mouth. I think the reality too is that just going back to like the Daario essay to today, Jensen, Nadella, you know, obviously Karp, you know, I think many have come out pretty strongly against the concept of what Daria was talking about and what is is there a moral issue here? The risk factor section for anthropic reportedly saying their technology is an existential risk to humanity. I I like I've never seen anything like this before. How could I put money into a work work in a company that might kill me and my family? Like there's definitely unprecedented, right, in terms of a risk factor like that. But I would I mean like they're not similar but we just even go back to like Taiwan semi and right I mean you know me and you that have carb attack for so many decades like that's another name like there's a risk like you know in terms of like geopolitical in terms of Taiwan but I think investors they understand that this is a risk again I continue to view this as kind of a a shock or risk relative to what I believe is something where guard rails will be put in. Technology always goes ahead of the regulation and but I do think there's some alarmism that's that's going on here as well. >> Dan, $518 billion in obligations for anthropic. Are they too big to fail? >> Look, I think an anthropic is the hearts and lungs of the the AI revolution trade and that goes for open AI as well. I think part of I almost view it as me be like uh you know looking at it the wrong way or misguided or whatever. When everyone talks about the circular financing trade and that's some sort of negative doomsday view I I view it differently. I view it is you are essentially building out a new economy over the next 10 years in terms of the AI economy from data centers to chips to GPUs and for the first time in 30 years US is headed China when it comes tech anthropics at the center of that I do actually think for tech them going public is bullish for the broader tech trade more transparency and I think that's something that investors will crave and look just like we saw SpaceX, it comes down to what reception will be because on the private side, it's different than when public investors because there's a market, there's demand, there's buyers and sellers, and I think that that's a healthy process. >> I think Anthropic would have liked to go public a couple months ago. Now we have Aura delaying its IPO. What are these what's the signal from that to the to the average investor? >> I I personally wouldn't put a stat that signal to this. I mean that's my I mean it's look there's going to be there will be some choppiness in IPO market you know just given what we see depending on different areas of tech but when it comes to the AI trade and when it come you talk about micro crime you talk about just overall growth and where this is going like size is my view like we're only 15% through the spending the 4 to 5 trillion that's going to happen and it's hard for an investor to just not pay attention to that or not participate in that. But I still believe this is a 1997 moment, not a 2000 bubble moment. >> Well, Dan, if we're only 15% through and Anthropic wipes us all out, I mean, there's there's no more to get through. >> Look, if that happens, at least I want to I want to, you know, be in a Corvette, [laughter] you know, driving, you know, >> there you go. You're getting my Yahoo of the day, man. You're getting my Yahoo yodel of the day, Dan. I know you love the Yahoo yodel. Um, another big story this week. I want to go to Micron. My heart wants to go to Micron, but I'm not. I'm going to go back to Nvidia because I haven't seen, this has been a big week of first for me, I should say. I've never seen uh a perspectus leak like this very important company. Never seen a buyback out of what Jensen did in Nvidia. My question to you is this. What's next for Nvidia? We're always looking for what's next. And I always I think we have we were expecting this bullet from Jensen. We were expecting a big buyback. like what is what what's what does next year look like for this company >> be about physical AI I think that that will be the holy grail for Nvidia because when you think about robotics and physical AI that will be multitudes more than the spending that we're seeing today I think when you talk about the buyback and what they're doing on the safety side and even what he's done with Karp in terms of sovereign AI he's laying out the platform form of basically they're building a sort of the runway the airport for what ultimately AI is going to look like not just in terms of enterprise and consumer but I think physical AI and I think it just shows the confidence because the reality is investors continue to underestimate the scale and scope of what's happening I'm not saying there aren't risk that the safety issues data center warriors in terms of midterm elections but you can fight the trend >> you think Jensen response to the pope Look, I mean with with Jensen, you know, never say never, but I just look I but I just think like Jensen's he's the adult in the room. Like the reality is like he is someone I don't care where it is, what country, every like when he talks, everyone listens, right? And and I think it just he has the best perspective and balanced perspective of what's happening in AI. Coming up, how one company is developing humanoid [music] robots that do dangerous jobs. That's next on Market Domination. [music] [music] Heat [music] up [music] here. >> [music] [music] [music] [music] [music] >> Heat. Hey. Hey. Hey. >> [music] [music] [music] [music] >> Robots may be the future of difficult and dangerous work. Foundation Future Industries is testing humanoid robots that could potentially work on battlefields doing dangerous jobs before soldiers are sent in. Now got Mike Leblanc, Foundation Future Industries co-founder. Mike, it's great to see you. Maybe start here, Mike. Um, I heard your company uh was testing humanoid robots in Ukraine. Mike, maybe start there. Walk us through that. How, you know, how exactly were those humanoid robots being used over there, Mike? For what task, what purpose? >> You know, Ukraine, I think the place that Ukraine's going to take in history of the West is going to be the initial place that robots were really tested in war. They have a mandate to take off as many people off the battlefield as possible and to replace them with robots. And so we had a lot of people reaching out asking us about putting our humanoids out there. Uh our our robots right now are the only ones that can actually, you know, potentially unlock all the weapons and vehicles and and all the spend cost that they have on human shaped people uh in the past. And so all the tests now have been focused very much on terrain, logistics, reconnaissance, very very kind of simple tasks. Uh but I think the belief is that the robots are eventually going to be able to pick up guns. They're going to be able to drive vehicles. uh and really pull pull humans off the battlefield. >> And just so I understand it, Mike, your robots in in Ukraine, uh were they navigating, you know, making decisions on their own, Mike, or no? There was still a human being remotely controlling all this >> in in all the dynamic tasks for for anything warlike, anything for the military that we've done, it's all a human controlling it right now. Uh we do have robots that work autonomously when they're doing uh working in manufacturing, but right now it's still very much human in the loop. Uh and it's it it gets very very tricky of whether the Ukrainians are going to stick with that or not, but we we really leave it to the militaries and the commanders to make those decisions. >> Maybe a simpler question, Mike. Why does the robot, you know, need to be shaped like a human? Like couldn't a smaller, cheaper robot on wheels do the same thing or no? Uh, so we there are lots of different form factors going on the battlefield right now. You've got drones, you've got unmanned ground vehicles, which look like ATVs that go out. So you have all the different form factors that are that are going out to fight fight armies, but you still have things like steps, you still have things like doorork knobs, uh, and everything that we've built for the military up until now. Every heavy machine gun that you need to set up, every vehicle that you need driven, all of those were made to be compatible with a human form factor. And so having a human-shaped robot really unlocks the potential of those. >> When you talk, Mike, about eventually maybe giving one of these humanoid robots a weapon, what's the case for that, Mike? What's the advantage of that on the battlefield? >> Well, you know, I was I was a Marine Marine Corps officer for 13 years, and so to me, anything that you can do to replace the 19-year-old from Louisiana going out of that battlefield is worthwhile. Today, when you look at a lot of the tasks that the military is doing, a lot of them are walking around with guns. That's that's just the case. It's a lot of patrolling. Uh you know, whether it's combat patrols, whether it's just, you know, setting a cordon around a house, any of these things, you really have a gun. You're ready to protect against what comes. And you have very simple directives that you're giving to young people to do this. So, if there's any possibility of us taking those people off and putting robots on, we're we're going to build into it. >> Who would decide though, Mike, in that hypothetical you're describing, and who would decide, you know, whether the robot fires that weapon? Is is there a robot or do you think no, there's always going to have to be some human service member decide, okay, that robot can now pull that trigger? >> Well, already we're we're seeing with electronic warfare, right, that is that is the counter to drones. So, that will be the counter to robots, too. If you have robots that are trying to go out and attack an enemy, they're going to try to block that signal. And so, even today on the battlefield, you have drones that are already making decisions of what to do, right? They're they're hoping that they can identify a target. They're hoping that they can attack that target once it's out of a human's control. And so I think I think that boundary has already been passed and we'll see a very similar thing for for humanoids. >> It's interesting though, Mike. I mean, let's say, you know, if a humanoid makes a mistake when it fires that weapon, a mistakes a civilian for a terrorist, who do who, Mike, ultimately is accountable, responsible for that? Well, I you know, I think this is an age-old question of of philosophy of really agency in battle, right? And is is it the the king who set the war? Is it the commander who's carrying it out? Right? This this goes back to the Iliad. This goes back to Henry V. Um really, when we're putting our 18-year-olds, our 19-year-olds out there with, you know, very very basic rules of who they're allowed to shoot and who they're not allowed to shoot, decisions are being made willy-nilly all the time. And we we have very strong history of making a lot of mistakes on the battlefield. So I think that adding adding AI is actually going to put a lot more guard rails than people think. >> And finally, I'm just curious how how much do your robots cost, Mike? And how many can you actually, you know, make manufacture right now? >> So right now we're we're selling the robots commercially for $100,000 per year per robot. So when they're replacing three people on a shift, that that makes a lot of sense. Uh in in the military, we've worked with some different pricing primarily because our biggest objective is to get these robots out onto the battlefield, right? We we want to get learning as fast as possible. And so we'll we'll be making thousands of these next year and putting those hopefully on the battlefield where we can start iterating of what they're really going to do out there uh and what we need to build. I >> I read Mike Eric Trump is an investor and now an adviser as well. What what does he bring to the table for you all? You know, meeting Eric was was really interesting. We had uh he had initially invested in us, but then would come by the office. He's very mechanically inclined. So, he's he's actually he can geek out about robots all day. Uh but he's also very helpful in talking about building a very large robboactory that we're doing in Texas to build all these robots. Um and Andrew strategically talking about what we should be doing to capitalize the company. >> Mike, great to have you on the show today. Appreciate your time. >> Thanks so much. >> Coming up, got you covered through the closed bound Wall Street. [music] Don't go anywhere. [music] Heat. Heat. N. [music] [music] [music] [music] >> [music] >> Heat. Heat. [music] >> [music] [music] [music] [music] >> Heat. Heat. [music] [music] [music] >> [music] >> Down. [music] [music] Down. Heat. Heat. N. [music] Heat. Heat. >> [music] [music] [music] [music] >> Ah. >> [music] [music] [music] [music] [music] >> Stocks in a day lower here. Yahoo Finances Enz Fay has the very latest. Josh, we are seeing that the Dow has ended the session down 2/10en of a percent but off the lows of the session. The Nasdaq composite just below the flatline, but you can see wavering throughout the day and the S&P 500 also down about 2/10en of a percent. We've been watching all day the Treasury yields, the longdated bonds which had gone higher. So, you're looking at the 10-year at 526, the 30-year easing just a little bit. It had been at its li highest level since 2002. And this is what has some Wall Street analysts a little bit jittery about what's happening. But because we haven't seen these markets uh going down too much, but nonetheless, as those rates get higher, especially the tenure, if it goes to 55, that's where markets could feel more pain. Taking a look at where we're at with the sectors, you've got utilities that's been in the green throughout the session. You've got energy that's lower. That's because oil prices have fallen today. And then over on the NASDAQ 100, a mixed picture with the MAG 7 with Nvidia, Alphabet, Apple, all lower, but you've got Meta jumping 3% and also Tesla down 2%, Amazon in the green. Over on the semiconductor space, do want to note that we are seeing a little bit of green with some of these semiconductors, especially Micron, which will be reporting its results tomorrow after the bell. So, you've got Micron that's up more than 1% today. And then just looking at commodities, we have seen quite a bit of movement in the commodity space. Quite a bit of volatility when it comes to oil prices. You're looking one day it's up, but the the next day it's down. today down more than 3% for WTI Brent also lower and gold after that big sell-off that we saw yesterday today up about 1% Josh. >> All right, thank you Nez. >> Well, for more on the impact of rising yields, we got Jeff Klinghoffer, Aeristo Pacific managing director and portfolio manager. Jeff, it's always good to see you. Let's just start right there on the market action today. Oil actually slipping here. Your benchmark tenyear though, Jeff, it's here at 525. What signal or signals do you think that that is sending investors? What should we take from that? >> Well, I think what the market is trying to tell us is look, we the Fed has an inflation problem. Inflation is sticky. Growth is incredibly strong, but really what I think the markets are telling you is this Fed means business, right? I think what we're seeing is a transition in the market to take what the Fed is telling us at face value. Chairman Worsh came in and he told us we have an inflation problem and price stability isn't 2.1%. It's not 2 and a4%. It's 2%. And so I think the big transition we're seeing in markets is that we might have to start looking at 2% as a ceiling rather than a floor. And in order to get there, this Fed still has some work to do, which means rates need to be higher because growth is so strong. >> Yeah. On that point about growth being so strong, I mean, do you think, Jeff, is that why, listen, you can have this bond sell off and you can see your 10-year hit 525, but the stock market kind of hangs in there. Is that because Okay, well, if the macro is so strong, that should be good for equities. I think that's part of it, but I think that the the equity side will also have a little bit of tough time digesting. The impetus has changed for this Fed. The past Fed was willing to tolerate a little bit of higher inflation to really protect the employment side. What we're seeing from this Fed is almost the exact opposite. They're willing to >> at least put a little bit of challenge into the employment market because they need to see growth come down. They need to see demand come down because they really have to hit on the inflation side. They worry about that price stability mandate. They worry that eventually future price stability won't be a given from markets, right? They're losing potential. They could lose credibility. >> Is that what you meant, Jeff, when you when you talk about, okay, how this is, in your opinion, a fundamentally different Fed under Worsh. Is that what you're driving at there? >> That's exactly what I'm driving at. This Fed inherited a long-term trend of inflation well above target. That is a fundamentally different reaction function, and this Fed will deliver that. So, that's exactly it. This Fed worries that all of the credibility they built out over previous decades is at risk given they've now not hit that price stability mandate for five plus years. >> Would you bet then Jeff, okay, hikes come in October and beyond. >> It does. It depends on the data, but the way I see it right now, one hike isn't enough to get inflation down. So, I think their credibility will be a little bit uh requested if they don't hike in October. I we saw a little bit of weak data today, so we're going to see obviously inflation tomorrow. We're going to see another round of inflation mid-occtober. But if the data continues to come in hot and that's what I expect, they should hike in October and I think there's a good probability that they will. >> What do you make, Jeff, of those who push back and say uh the issue for Worsh is that the forces driving inflation in large part are actually beyond the Fed's control. It's geopolitics. It's a fight that's going on 6,000 miles from here. It's the great AI buildout. It's, you know, big tech's going to drop 800 billion this year, estimated a trillion next year. That's beyond the Fed's control. What do you say to that? >> I say it's not beyond the Fed's control. I think the entire entirety of the economy is well within the Fed's mandate and it's well within the Fed's control. >> Absolutely. Those individual pieces, certainly oil, they can't solve the war, but they can prevent the bleed through on AI. I push back. There is a level, I think we can all agree, there is a level that the issuers are sensitive to when it comes to rates and all of the debt they're taking on in this buildout. Thus far, we haven't hit it, but the Fed can absolutely affect that. And I think we're starting to see that. I think we are starting to see maybe capex just be a little bit more questionable on how far these companies continue to push into where they're issuing debt at the data center level at now mid to high seven, even low 8%s for highquality IG companies. >> I guess the issues are okay, so we're going to raise rates, we'll slow the economy. Then the issues how how much do we have to slow the economy to get back to target? >> Well, they don't know. And I think that's really what equity markets are struggling to digest. And again, if we take the Fed at face value, they are willing to push on the employment side of their mandate in order to get inflation down. Now, they don't want to cause a recession. That's not within their purview, but they're willing to accept at least the possibility that it might take one because they have to reanchor those long-term inflation expectations. We have to get to 2%. And again, I think we need to start treating that as a ceiling and no longer a floor. We have to get under 2%. I'm curious how President Trump factors into all this because you say you're a comment from the president could shake confidence in the Fed. What do you mean by that? >> Well, I think we'll go back to October. >> Coming into this, it was a big question of whether Chairman Worsh ultimately wanted higher rates. He's given us higher rates. At the moment, he certainly seems willing to go even higher from here. And so, I think we've broken that link between the potential non-independence of the Federal Reserve. So, if Trump was to tweet, thus far, he's been relatively quiet. He doesn't want higher rates. But if all of a sudden Trump was to start tweeting and then we see a pause in October and potentially one in December, then I think all of those questions around independence would come back to the market. But thus far that link seems to be broken. And I I will continue to say the two cardinal sins you cannot commit as chairman of the Federal Reserve or any central bank is you cannot lose independence and you cannot preside over a central bank with high and persistent inflation. And I think Chairman Wars absolutely doesn't want to to have either one of those. >> Let's say you're in here on some advice for fixed income investors. Jeff might be listening. What do they buy and what are they avoiding? What are you telling clients? >> Yeah. Well, look, I say go back to basics. Fixed income, it there's no question overall the income component is back. It's easy to get a 6.5% type yield on very high quality fixed income portfolios. But I think you want to layer into longer rates because the biggest benefit of what we're seeing despite the pain of how we've gotten here is that if the economy was to roll over, look, this is a real business cycle. Eventually it will. >> If and when that happens, the Fed can cut. fixed income really looks very compelling here with the yields that we have in place. >> Jeeoff, always great to see you, especially on set. Thank you. >> Thank you. >> Coming up, how AI entrepreneurs are thinking about [music] safety fears. That's next on Market Domination Overtime. [music] [music] >> [music] [music] [music] [music] [music] [music] >> Heat. [music] Heat. >> [music] [music] [music] >> Down. [music] Down. >> [music] >> Ow. [music] Down. [music] [music] >> [music] [music] >> Heat. Heat. [music] >> [music] [music] [music] [music] [music] [music] [music] >> Open AI is holding back a new model over safety concerns. Anthropic is warning that increasingly autonomous AI can behave in unexpected ways. Joining us now for more on the AI safety landscape, we got Noam Schwarz, CEO of AI security company Alice. Noam, it is good to see you. Maybe I I thought Noam just start here. Just explain for the audience, Noam, in simple terms what you all do. You know, I was reading through it and it sounded like no like you all are the ones who are sort of you're testing and securing top AI models. Is that the right way to think about it? >> Yeah, 100%. So, we're frontier safety company. We're uh testing and protecting models, agents, applications, online platforms, personal AI agents, the whole thing to make sure they're doing only what we want them to do and not going and do things that we don't want them to do. >> So, you're basically being paid on, it sounds like, attack the model and see if we can get this model to do something it wasn't designed for. >> Yeah. So we're we're getting to attack the model, to test the model, but also to protect the model in real time. So it's figuring out how it can actually predict in adversarial circumstances. So we're testing it and how we actually need to use it and see how it regress, how it fails in in regular state. But the most important part is when that model when that agent when that AI that we're all using is actually meeting an adversarial actually meeting something that not supposed to happen to meet like something that is trying to abuse it. What then happens and we're protecting the models and the agents against just that noam. Can you give us maybe a real world example here? I'm not asking you to name names but like an example to illustrate this like an example where okay you got the model to do something that surprised the company that built it. >> Sure. So there's something called prompt injection which is kind of like the classic attack against the model. That means that let's say you tell your personal AI agent go and buy me something or go and do some research find like different types of uh laptops and then that agent is going and doing a research online. One of those websites may contain a prompt a line of text or an image with something hidden in that image and that will include instructions to the model to do something that you don't want it to do from sending payments. Well, we replace or just like making your model do something that you don't want it to do. So, we're protecting against that and we're testing that the model will actually be able to understand that it's being manipulated, that it's being abused, and that that attack would not get to the user. >> We saw news Noam, you saw this too, of course, that Open AAI scrapped the release, it sounds like, of a new model because of safety concerns. Why would they do that Noam? And then what do you make of that generally? >> So that's that's pretty straightforward. Safety is now becoming an operating constraint and not just a research discussion uh discussion point. Now everybody's talking about AI safety for a very long time uh in the past like four years and open AI decision is actually an evidence that testing can do its job. it's finally doing its job. And and the lesson here is not just that AI is uncontrollable. Uh it's that model alignment has to be paired with independent evaluation, system level security, monitoring and containment. It means that people are understanding that safety is not just a pie in the sky and things that you know may happen in the future. It's something that need to influence the decision making right now. >> Uh I don't know if you saw no, you probably did. Nvidia's Jensen Wong was on CNBC this week, right? And he was making the case. I'm paraphrasing, but basically Jensen was saying, "Listen, in his view, ultimately AI safety, it is an engineering problem. It better be," he was saying or or you know, we don't have a fix for it. Do you agree with Jensen Noam on that? So I think AI safety is an important problem that prevents the adoption of AI. But there's many different entities, many different people that are describing AI safety in different voices from you know the end of civilization as we know it to a product issue or an engineering issue. Effectively the concept is right now AI is not in a place that it can destroy the world. Uh AI escaped its laboratory. it attacked like another company. This is not something that is very likely to happen when you're using an AI on your server or on your laptop. It may happen in the future when capabilities will be way more sophisticated and compute will be way more available. This is not the issue right now. The issue right now is closer to what Jensen is saying, an engineering problem. When we're teaching AI to do something really good like being great at understanding how to find vulnerabilities in code, it's also very good in creating those vulnerabilities. So when we want to create an alignment and when people are talking about alignment, it means alignment with the will and the purpose of the person or the company that is using that AI. It means that we need to make sure that the AI can uh use our our agency that it do that it does exactly what we want it to do and there are many ways to make sure that this is happening. Uh there are guard drills there's like all kinds of uh evaluations that you can do. There's many different solutions and our focus is to make sure that AI can bring out the potential that it has can take us as you know society as humanity and bring the potential out without causing damage. But this is possible and it's happening and we're seeing safety research and safety practitioners are advancing every day. So I believe the public can expect uh all kinds of incidents all the time. Exactly like what we saw with hugging face and open AI but the long term it would be okay because we're as an industry we're actually here we're working very hard towards protecting people protecting companies protecting models and we will figure this out and we will eventually use AI and and and use the entire the full potential. Let me ask you a final question on maybe I'm a CEO and I'm listening to this right now and maybe know I've I've I've uh I've deployed I've integrated AI models throughout my company. Do you have any just kind of basic tips and tricks and just general advice and guidance Noam for those CEOs about how to deploy and integrate these models responsibly and effectively and efficiently. >> Yeah. So right now you need to make sure that when you're using AI and this is a lesson that most companies already learned that it's not something that works out of the box like for personal use can use CHP you can use cloud you can use whatever you can use bot and you can ask it a question you can tell it to do something it will do it right away it will probably work very well for most enterprises for most companies it doesn't work this way you need to adjust it to your own tools your own architecture you need to train it you to sometimes do finetuning to it. The same type of investment and deployment needs to be also for the safety and security of your uh of your setup because if you're thinking about safety as an afterthought, you'll eventually have an incident that will cost you a lot of money that will cost you like your your grand. So, this is something that sometimes people forget. They're very enthusiastic about leveraging AI, but they forget that they need to secure it in a very different way. This is not just like a security problem. It's a safety problem, which is a new world. This is why OpenAI pulled out from the from the latest release and this is exactly the path that our that our industry is going towards it. >> No, I'm so great to have you on the show today. Thanks for your time. >> Thank you very much. >> AI may be making healthcare more expensive. That's according to a new report by Blue Cross Blue Shield. Joining us now is Yahoo Finances, Carrie Hannah. And Carrie, what do we know? >> Yeah, Josh. Uh, so this new report came out from Blue Cross Blue Shield Association and their members represent roughly one in three Americans. Um what they are saying is that AI coding that's used by the hospital systems have increased is cost them in billing um close to a billion dollars over a 2-year period. And so what they what they are saying is that because the hospitals have been using the AI to to do analysis of patient information that someone might go in for one uh situation, one health issue, but they end up uh being diagnosed with two or three additional ones after that, maybe anemia or something like that, another thing. And as a result, they move up into a higher category for billing. So when the hospital bills the insurance company, the bill is fatter than it was uh previously before they were using AI to scoop up all this information about the patient. So it could be even from the doctor's notes, from conversations uh perhaps even in the room. There's lots of ways they scoop up extra information. And so they're saying that uh this association, the Blue Cross Blue Shield Association says, you know what, not only did it cost us this much that they found in their research that the patients didn't even get extra care for the extra billing. Now, so that's a pretty big, you know, assertion that it's really pumped up their cost for the insurers. So it's not me, it's you. And on the other hand, insurers aren't so innoc innocent either, Josh, because they have also started using AI to deny claims to um uh delay them to deny them. So they have made it difficult as well for patients to get the care they need. >> And so why Gary does AI use by hospitals? Wh why does that translate to higher premiums potentially for Carrie and Josh? >> Yeah, bingo. So business 101, right? Okay. So, so the hospitals bill the insurance companies more money and they say, "Okay, I'm passing that down to the patient." So, as a result, your premiums, my premiums, your premiums get bumped up in order to for the insurers to make sense of this extra charge, the extra charges they're getting from the AI used by the hospitals. So um say for example now uh KFFF uh the nonprofit that analyzes healthcare issues they are anticipating premiums for uh the medium premium to be up 15% uh next year and this year was also double digits. It was up 20%. So you know consumers are getting slammed and hammered by these higher premiums with rising costs and the insurers are saying it's not our fault we're raising premiums it's the hospitals. So, you know, it's quite the the fat the fight out there. >> On the flip side, Carrie. Okay. Any any good that you see coming from the use of AI by health providers. >> Yeah. If I'm a health care provider, if I'm a hospital, this is a great way to streamline your billing. Um, you can really cut costs by using AI to to make billing quicker, to make it and presumably more accurate. right now we're hearing that it's not according to the association but so there are ways that it can it can make the system more efficient but that doesn't translate to what um you and I will will see with our premiums >> so for patients bottom line here Krie what if anything you know can they do should they do >> yeah the the deal with the hospital billing the insurers that's pretty tricky for a consumer to to push back on because you don't really have control of that how that's going to move down to what your premiums is going to be a as they try to pass their business cost along. What you can do, and I'm looking at it just from the angle of the insurer uh denying or delaying claims, and there's a a whole thing going on right now, Josh, with Medicare, traditional Medicare, a pilot program that they're using AI in certain states to determine if a if a patient, a Medicare patient can get the uh the care they need. And it's been a disaster. So, at least so far, what I have found in my reporting, what what I would say is if you can push back on a denial or a delay, get noisy, get angry, and get and stay on them, stay on the insurance company and on your medical provider in order to see as as soon as you can push through uh that roadblock. >> All right, Carrie, thank you. Great reporting as always. >> Thanks. >> Coming up now, retail investors are navigating [music] recent volatility. That's next on Ask for a Trip. Heat. Heat. [music] [music] >> [music] [music] >> Heat. [music] Heat. [music] >> [music] [music] [music] [music] [music] [music] [music] [music] [music] [music] [music] >> Heat. Heat. [music] [music] [music] Heat. Heat. N. [music] [music] [music] >> [music] [music] [music] >> Hello and welcome to asking for a trend. Well, investor confidence slipping, that doesn't mean they're heading for the exits. New Investopedia data shows retail traders are getting more cautious even as they keep buying those ETFs, money market funds, and some of the biggest AI names out there. Joining us now, we got editor-inchief of Investopedia. That would be Caleb Silver. All right, Mr. Silver, let's start with this. Let's start with how folks feel. All right, because this was an interesting stat. The share describing themselves as cautiously optimistic. That falls, it's a fall 61% in July. We dropped to 42%. Start there because that's a drop. What's driving that? >> Yeah, only about 20 percentage points. And that's just a vibe feeling that is I'm just not so sure that we are as strong as we were a month ago. >> What is that? >> Everything. rising rates, the house of horrors that is the economy, the unpredictability of it all, and also surprises that don't really allow investors to say what is the future going to look like? Are we going to have higher for longer? Is this 5% just the new normal uh in terms of the yield on the tenure that changes the equity risk premium, right? Also, how much longer are we going to have inflation at these levels and what the Fed's going to do about it? What is that going to mean to my portfolio? Again, we feel one way just like consumer sentiment. We don't often change behavior as individual investors. That said, there has been a lot more floating towards those safety havens. Money market funds getting a bid here. Diversified ETFs getting a bid. The biggest ETFs getting the bid here. >> Well, I want we covered how they feel. All right. That's how they feel. Let's talk about what they're actually doing with their money because to your point, I'm looking through your data. They're leaning. Where they're leaning is interesting. It's ETFs. It's money market funds. You look at that. What do you make of that? What does that suggest to you? >> Yeah. Well, getting four to 5% in a money market fund sounds okay given the uncertainty. And we've had a lot of gains. We're coming up on the fourth birthday of this here bull market. And investors know that. Long-term investors know these go five to six years. Though lately there's been a lot of topheavy talk. So maybe it's time to get grab a little safety. And that's the money market funds. PS. Money market ETF funds have seen the biggest surge in history in just the past 6 months. So it's not just individual investors. You've seen some institutions lean that way. Even though institutional investors are as bullish as they've ever been. If you look at their cash levels, they're still moving money into money market ETF funds. So, a little searching for safety there, but also the uncertainty of rates. Why not grab on to four to 5% right now? And the ETFs is a diversified play. It's really hard to pick stocks in a high beta environment right now. Not everybody was able to load in to the big semi stocks this year. So you look at the stocks that we hold, they're the same stocks we've been holding basically for the last 5 to 10 years. And the gains have been good, but not a lot of playing on the outside of the core. I would say 10 to 15 stocks. >> Uh let's talk AI because that's interesting. All right. Investors say AI stocks, they look bubbly. They're bubbly. Okay. But then they're also buying per your data, my friend, they're buying Nvidia. They're buying Meta among the names, right? So how are both of those things true at the same time? Help me understand that. >> Yeah. We fear what we own. >> Worried about a AI bubble. But you know what I'll do? I'll buy the face of AI. Jensen Wong's company. >> Yeah. We fear what we own, but we're afraid to not own it just because the results are there. That stock's up some 5,000 plus percent over the last few years. Thank you very much. And nobody wants to miss out on that. Plus, it's almost hard to avoid owning Nvidia. If you own any index fund or if you own any big ETF, it's probably in there. So, a lot of these people own it directly. A lot of these people have been buying it on the outside and adding to their positions. And every time they've doubted themselves and by buying it and done it anyway, they've been right because the stock continues to deliver for investors. That's not the only one they've been buying. They've been buying other stocks around the security complex. They've been buying uh Crowd Strike. They've been buying Meta. They've been buying these midstream uh oil refiner ETFs and that got one of them, AMLP, uh took in a billion dollars in inflows. That's a lot for a midstream refiner. So, they've been experimenting in other places, but you still see the same sort of concentration. and AI and tech, even though that's where we fear that's where the bubbles are. >> 54% 54% see a bubble in AI stocks. I mean, I we're gonna I guess we're gonna keep having this conversation. AI bubble. I really don't understand the bubble. I'll tell you why. Get your take on this, Caleb. I don't understand how you look at Nvidia's multiple and say, "Oh, yeah, that's a nosebleleed." No, it isn't. Nobody's saying that. In fact, everybody's wondering, "Why is it trading at that multiple?" And and I don't look at investors and say, "Oh, you know, they're really approaching this whole thing with these rosecolored glasses." In fact, all I hear these hand ringing about bubbles and the skepticism and the concern. If we're talking about a bubble this much, is it a bubble? >> Yeah. If everyone's talking about it, is it or are we just talking about the same thing and media helping sort of regurgitate the conversation? Yes. Guilty is charged. But all I think we've also been conditioned to focus on maybe the wrong things in this environment. Think about all the metrics that tell us we might be in a bubble. Cape ratios, right? We're looking at a smoothed out earnings for the past 10 years. Who cares about the past 10 years? We care about the next 10 years. So forward PE ratios. >> The cape is like just for people that's like old heads care. >> I get it. I get it. And a lot of the old metrics that we look at and that were traditional. >> Why am I comparing this economy to 10 year? I mean, does that make sense? >> Different this time. Worst things you can say in investing. That said, every cycle is different. This one, if you look at it compared to other cycles, 1999, 2000, it went on for a while before that bubble burst and there wasn't as much real revenue there. And there's some hocus pocus revenue going on here, but we're talking about trillions of dollars with companies making real products and Nvidia is the poster child of that. >> We talked about what they're buying. Let's talk about what they're selling. Yeah. >> Palunteer, Microsoft, Service Now, Salesforce, Oracle. Is there a theme or themes in that that you can see? >> Software outside of Palunteer, which is still software. It's data plus software plus a lot of government contracts and it has been a high flare. That's a trillion dollar or about to be a trillion dollar company. But the other ones were all sort of in infected by that AI software will eat the world and eat these companies and all of their clients because they won't need as much software and they won't need as many upgrades if they're all doing it through AI. So that was part of that trade and we saw the bubble burst in that sector for sure. So they were selling out of what they feared. At the same time, uh there's still a lot of opportunity on the on the uh cyber security side, which is why they've been buying the crowd strikes of the world. But look inside the big concentration of the stocks they own in their portfolios and it looks just like the top of the S&P 500 or just like the top of the QQQ maybe plus Bergkshire Hathaway and a bank or two. >> Um I always love when you ask people this question. You ask, all right, what would you do with an extra 10 grand? uh they would be buying stocks and funds and that ranked ahead of paying down debt. What do you make of that answer? >> Well, people have done very well and hopefully these are people that don't have a lot of debt, but with that discretionary money, they do look at high beta stocks and they say, "Okay, I'd take a flyer with 10 grand on a on one of these stocks that could break out or add to my position in Nvidia." So, we always like to ask that question because it is ultimately the discretionary question. What would you do? Are you feeling safe? Are you feeling even though you're feeling cautious right now, you'd still go out and spend money on individual stocks? Maybe we're not as scared as we think we are. We just say that we're scared. A little bit like consumers don't feel great. We're still spending. >> Spend, spend, spend. Final question. I'm just curious. Is is crypto coming up in the conversation? Is Bitcoin nice? I mean, you were roughly around 60 this summer. You made a nice move up to I don't know what 83. Is it in the bloodstream? Are people talking? >> It used to be >> and price dictates everything. Even this survey we closed just at the end of last week. is very hot off the presses and not a lot of bids on Bitcoin. We always ask, is this going to be a a bigger part of your portfolio going forward? Would you buy and hold it for the next 10 years? It did not make the list this round or the last round, which tells you that when the price comes down like this, the interest waines away from it, though, it's catching a pretty big bid right now in an environment where you wouldn't expect it with the dollar rising and bond yields higher. Weird. >> Caleb, one of the greats. >> Thank you, my man. >> Thank you, brother. Stick around. Much for a trend still to come. >> [music] [music] [music] [music] [music] [music] [music] [music] [music] [music] [music] >> down. >> [music] [music] [music] >> Down. [music] [music] Down. >> [music] >> Down. [music] Heat. [music] >> [music] [music] [music] [music] [music] [music] [music] [music] >> Former Disney CEO Bob Chapek sat down with Yahoo Finance executive editor Brian Sazy this week to discuss his experience at the House of Mouse, his new book, and a lot more. >> Who is Bob JPEG? I feel as though just reading the book from page to page, you have been completely misunderstood. >> I think so. I think to a large extent I'm a fish out of water uh in the let's say the media business/Hollywood business. I grew up uh right outside Chicago uh in sort of the uh Gary E Chicago Hammond area uh sort of a uh bluecollar uh rust belt as it's called sometimes area and uh I grew up really uh with great supportive parents in a great community but the idea of going to work at a Disney I mean that was mythical right no one actually goes to work at Disney or frankly for most of the people that I went to high school with even going to college was not I mean you you went to the oil mill you went to the steel refinery uh I mean the the uh steel mill or oil refinery or you worked at the soap plant or you worked at the corn processing plant those were the types of jobs that you did and it was an extreme uh unique situation for someone to you know go to college graduate and then move out and I think that has left an imprint on me that isn't necess necessarily consistent with the way most of the people that I've, you know, grown up in my career have had. Um, and I think it makes me a little more gritty. Uh, uh, I believe in willpower. People say, you know, what's kind of the one thing I'm like, willpower. You just, it's amazing what a human can do if they will themselves to achieve a goal, whether it's a personal goal in terms of health, whether it's a personal goal in terms of financial or just, you know, trying to build a family. So, I think I I am a bit of an of an oddity uh in the world that I've actually practiced my 40-year career in. Uh but I celebrate that and as you can tell in the book, uh uh it's something I'm proud of. Is that grittiness what got you out of that town? >> 1,000%. 1,000%. Uh, I recant in the book a a brief interaction when I was a high school freshman and I sat with my guidance counselor and I said, "Okay, I want to go to college. Okay, tell me what classes I should take." And she essentially scoffed at it. She goes, "You know, be a man, be like your grandfather, be like your father, go work in the oil refinery, uh, retire at, you know, 52 and have a good life. Why do you want to go to college? Why do you want to waste your parents' money? Uh, and to me, that was uh that was my my spark. That was my my catalyst to say, "Okay, I've got to do this on my own." You know, they were more than happy to give you your fourth semester of, you know, metal shop. But, [clears throat] uh, you know, physics and chemistry, they had it, but very small classes. I was I was one of them sitting there absorbing it. Was your first day at Disney like a shock? You just said you were the average guy. I think Disney I think suits uh certain vibe and maybe not so much the average guy. You're correct. I mean, you know, one of the ways I differed from most of my peers in Disney is that, you know, they all had the uh Ivy League, you know, or Stanford or University of Chicago or MIT sort of background. uh you know, Indiana University, public school, state school, Michigan State University, public school, state school. It it was different. And and and the culture there is it's an interesting culture because yes, you have, you know, very well- educated, very well bred people uh uh who choose to spend a section of their career there. But you also have the creative community, which tends to be different, right? you know, extraordinary artists that tell stories and really are the heartbeat of entertainment and particularly Disney. Uh but you know I wasn't either of those you know so I did have to learn elasticity and morph uh depending on who I was talking to in the company because you know having a meeting with uh uh your head lawyer or a CFO is a lot different than someone who draws animation for a living or creates music as a soundtrack for a movie and you know you really did have to become fairly elastic. Were you the Just sounds like you were constantly viewed as the underdog from day one you walked in the door. >> Absolutely. I was I I I the term I always used was dark horse. >> Dark horse. >> Yeah. I was a dark horse. Uh never high in awareness. Always kind of kept my head down and and did my job and hope that my business results would speak for themselves. uh I was several times consult during my career that I needed to uh advocate for myself a little bit more and and I guess I did a little bit but I in the world of Hollywood where everybody advocates constantly for themselves and self-promotion is an art form. I probably was way behind on that. >> Do you think that held you back? >> I think it held me back. Yeah, I think I think I might have moved obviously I couldn't go any higher but I could have probably moved faster had I done that. Weren't you the one that fixed the Disney retail business? >> Well, we, you know, we had the Disney stores and then because they weren't working, they sold the Disney stores and because it wasn't Yeah. It's a Children's Place and then they bought them back. And, >> you know, I I'm not a guy that takes well to it can't be done. >> It's like a personal call [laughter] that it can be done and I'm going to prove it can be done. And uh the business was not making any money. Tremendous capital investment. uh long-term leases in malls and uh I one of the businesses that I inherited in consumer products was the Disney stores and I wanted to formulate a plan to revive them and bring them back to growth and I was told it couldn't be done. You know, it's a mall-based retail business. Don't count on it. And uh 3 years later I doubled uh operating income to a pretty healthy number uh by returning the product back to quality because the quality of the product had really fallen. Uh reduce reliance on discounts dramatically. You know no lease line sign saying 50% off every week. Uh and just returning some of the magic to those stores and it worked really really well. And I think we defied the conventional wisdom that you could make money on this business because they hadn't for the last 20 years. >> You know, the wins you had in the in the video distribution, the wins you had in retail. Why do you think it took you so long to be to break through onto that leadership team? >> Well, I uh uh I was in the movie studio for most of my career and uh I was doing a lot of great things. Uh but I think the management there at the time appreciated the contributions I was making because you know the movie business you know if you isolate just the revenue >> the golden goose. >> Yeah. I was making the money keeping the studio sort of I should say our team me and my team were making money keeping the studio profitable and uh uh hit a lot of u warts you know uh the the the film business trying to make something successful in theaters is hard business. You know most films don't make money most lose money and every once in a while you have a breakout hit and hope that that breakout hit makes more money than all the others combined loss. But even if you do that, you're still at break even on tremendous investment of capital. Capital meaning the money to produce films and distribute them and to market them. And we were just crushing it uh every year. And so I think I became sort of almost too valuable to trade. And uh it was only when Bob uh sort of uh you know decided to make some moves in the studio that uh he put me over to consumer products and then moved me over to theme parks and then it started happening really fast. >> For the full interview with former Disney CEO Bob Chapek, stay tuned for Power Players this Thursday, October 1st. You won't want to miss it. Stick around. More As to come. Heat. Heat. [music] [music] >> [music] [music] >> Heat. Hey, Heat. [music] >> [music] [music] [music] [music] [music] [music] [music] [music] [music] >> Time now for to watch. Wednesday, September 30th. Micron sets report earnings on Wednesday. With investors expectations running high on the AI boom, surging demand for memory chips has driven prices up, putting the company on track to beat revenue estimates. Next on deck, fresh inflation data from August personal consumption expenditures or PCE with economists forecasting headline and core [music] PCE to rise on a month-over-month basis. And finally, we're going to get new figures on the private labor sector from ADP September employment report, which is expected [music] to show 74,000 jobs added for the month. That's compared to August 38,000, giving us more insight into the health of the job market ahead of Friday's full jobs report. That's a wrap on today's show. Thanks for watching. Heat. Heat. [music] [music] [music] Heat. Heat. [music] [music] >> [music] [music] [music] [music] [music] [music] [music] >> Heat. Heat. [music] [music] Heat. Heat. [music] [music] Heat. Heat. N. [music] [music] Down. >> [music] >> Down. [music] >> [music] [music] [music] [music] [music] [music] [music] [music]