AI financing, cybersecurity and commodities shape the outlook 8/11/26

 Sonnet insurance wants to know, what would you do with an extra hour in your day and more money in your pocket?
 With Sonnet, you can quote and buy hominato insurance online or over the phone, your choice, on your time, whatever works best for you.
 And Canadian University and College alumni could save over $2,000 a year. That's real money, back where it belongs.
 Save time, save money, save yourself from the way insurance used to be.
 Get your quote at sonnet.ca, switch, save, simple, Sonnet.
 Recently, our company softball team lost the big game by one run.
 Then Dale tried to console us with the quote, winning isn't everything.
 Well, Dale and I are very different.
 I get early payout from bed365.
 If my team goes that big, I get paid out instantly, even if they blow the lead later.
 Sound familiar, Dale?
 Thanks, bed365.
 Must be 19 or older Ontario only, please play responsibly.
 If you have questions or concerns about your gambling or the gambling of someone close to you, please go to connectcentario.ca and see supply.
 I'm Morgan Brennan, and this is your morning call.
 Good Tuesday morning. Let's get a check on US stock futures, which are essentially flat right now.
 The Dow's poised to open lower by 116 points.
 The S&P slightly lower the Nasdaq, slightly lower as well.
 This is after muted session for Wall Street yesterday.
 We're all the major averages finished in the red, but fractionally.
 And I do mean fractionally.
 This ahead of existing home sales today, CPI tomorrow.
 Let's get a check on what we've seen in the bond market.
 There's been a little more activity there with yields moving higher again since the start of the weeks.
 You could see the US 10 year treasury yielding 4.73%.
 Now Fed sensitive to your treasury yielding 4.2.
 2.4.26% and 30 are treasury yielding 4.27% right now.
 Metals are on the move.
 Again, you've got gold and silver hitting June highs.
 Silver now down, fractionally, gold up, fractionally, 4426 per ounce for gold.
 Aluminum is at a seven week high as well.
 Copper is also continuing its move higher.
 Energy is higher, too, as we continue to monitor the situation in Iran and throughout the Middle East.
 You could see oil prices continue their climb here since the start of the week.
 WTI is up another 3% this morning trading at $84.52 a barrel and Brent is up 2.5% trading just below $89 a barrel.
 Let's see how Europe and Asia are shaping up.
 Karen Cho is in London and Lisa Kim is standing by in Singapore.
 Karen, let's start this with you.
 Morgan, good morning to you.
 Well, European equities are under pressure and early trade here.
 As traders digest, the latest developments out of the Middle East,
 with hopes that a straight-up or moose deal is fading.
 Now, oil prices are extending gains as the U.S. and Iran dig in on their positions,
 with President Donald Trump demanding reparations from Tehran.
 The higher crude price is providing a lift to the oil and gas basket of stocks,
 which is leading gains here today at a sector level.
 But on the flip side, travel leisure is underperforming.
 You can see we've got falls pretty much across the board at the index level.
 And we are tracking shares of IHG which have dipped in the session
 this after the group's full year earnings with the conflict in the Middle East,
 weighing on the hospitality sector.
 The hotel and resorts chain said training in the U.S. accelerated in the second quarter
 as it confirmed its outlook for the year.
 This thanks to a boost from the football World Cup.
 IHG reported revenues of just over $2.6 billion for the first half of the year,
 also talking about stability in the Chinese market, Morgan, it's back to you.
 All right, Karen Cho, thank you.
 That World Cup hangover seems to be coming through some of the data
 and some of the earnings that we're getting now.
 To the overnight action in Asia and our Lisa Kim, hi, Lisa.
 Hi, Morgan.
 So here in Asia, all eyes are on the upcoming IPO of China's most well-known
 humanoid robot maker Unitry in Shanghai.
 The company said its 900 million IPO was more than 8,000 times over subscribed
 by retail investors compared this to the Chinese trip maker CXMT
 that was 200 times over subscribed.
 But the U.S. restrictions on imported robots are set to be a challenge for Unitry
 considering that more than a tenth of that company's sales came from the U.S. last year.
 We're also watching the Japanese yen and has given up half of its gains
 from the historic joint intervention by the U.S. and Japanese authorities
 to strengthen the weekend.
 But now, Japanese business leaders are calling for a stronger yen and a survey
 by a quasi-government trade organization published earlier this year
 to show that one in five Japanese business leaders saw a dollar yen rate
 of 120 to 124 as the most desirable level, which is a far cry
 from where we are trading today.
 I should say so.
 Okay, I just got to go back to this.
 Lisa, 8,000 times over subscribed Unitry IPO by retail investors.
 8,000.
 I'm trying to wrap my mind around this.
 That's right, Morgan.
 It's 8,000 times over subscribed.
 And the most recent big IPO that happened in China was CXMT.
 Remember the chipmaker that Apple is trying to gain access to?
 So that IPO was 200 times over subscribed, which was big.
 But now, this one is 8,000 times over subscribed.
 And folks are keeping tabs on this IPO because a successful IPO by Unitry
 is likely to pave the way for more Chinese robot companies that are waiting to list in China.
 All right.
 Physical AI.
 We'll keep tabs on it.
 Lisa Kim, thank you from Singapore.
 To one of our top market stories as we continue to talk about AI.
 In video, inking partnerships with six of the world's largest asset managers
 to source more than half a trillion dollars in third-party financing for AI
 infrastructure, specifically for chips, AI chips.
 Wall Street heavyweight joining at the Commitment include BlackRock,
 Goldman Sachs, Brookfield, Apollo, KKR and Blackstone.
 And executives from the seven companies joined CNBC late yesterday after the market closed
 and a rare joint interview to discuss the terms of the announcement.
 We're talking about an extraordinarily significant infrastructure build.
 This is with a very hefty price tag.
 It's a hefty price tag.
 Each gigawatt is something like $50, $60 billion.
 And so there's energy involved.
 There's land power and shell involved.
 And of course, there's the computing part of it.
 We used to build chips that we sell and these are technology components that people buy and use.
 But now, Nvidia's AI factory platform is really an investable asset.
 An infrastructure asset.
 What we're doing is we're trying to find different ways to raise or to participate in raising
 the enormous amount of capital that's necessary to fund this infrastructure build out.
 And you're starting to see, in a sense, you know, asset-based financing
 against this infrastructure build out.
 And that's not surprising because these are real assets.
 They have real value.
 You can put a tangible value on it.
 And there's a lot of capital out there.
 It's super interesting.
 Let's get over to CNBC, Senior Technology Correspondent, Arjun Carpal with more.
 Arjun, it's great to have you on.
 I think we've got to start right there with, I think, the key question to all of this.
 And that is, what is the life cycle of a GPU?
 How much does all of this financing and this idea of a platform hinge on that?
 That's a great question, Morgan.
 It was an interesting debate because I feel like a year or two ago, the question was around
 the life cycle of the GPU may be being somewhere in the region of three to four years.
 Now, the discussions are a lot more extended.
 We're talking about potentially five to ten-year timeframes.
 And that all really underpins whether this is going to work or not
 because there are questions over how useful and how long these GPUs are going to last,
 especially as Nvidia has committed to a very aggressive cycle of releasing new chips.
 Now, I think one thing constraining all of that, of course, is the fact that there are restrictions
 on how much can actually physically be produced by some of the leading manufacturers out there like TSMC.
 We know there's a massive shortage of compute.
 And if we think about this deal right now, there are a few details we have beyond that.
 But the reason that Nvidia and these financiers are doing this is that building data centers
 and acquiring these chips is very expensive.
 Now Nvidia has a very handy way to go to customers and say, hey, here's our chips
 and here's a neat financing package to buy them.
 And it's a big benefit to Nvidia which has argued that GPUs and compute infrastructure
 is basically akin to electricity or water or other utilities as well.
 And Jensen Huang has argued so heavily that Nvidia's AI factories
 as he calls them or data centers actually produce tokens, models, AI services
 and therefore these GPUs as we're discussing are revenue generating
 and will be revenue generating for a long time.
 So Huang Keh is arguing that GPUs should actually be seen as a very productive asset
 that is an investable asset.
 And it could actually help Nvidia sell more hardware because if you think about smaller companies
 that don't necessarily have the firepower of the hyperscalers to spend that capex and buy assets,
 this gives them an option now to buy some of those Nvidia GPUs but it also gives
 I would argue Nvidia a competitive advantage over rivals like AMD for example
 because it could further enhance its lock-in and its positioning in the AI ecosystem.
 But of course comes with this own set of risk more than as we talk about circularity
 of financing and those could already agiters in the market around capex.
 All right, so many implications, we're just scratching the surface
 but Arjun Carfall, thank you for breaking it down for us.
 Good to see you.
 Let's continue the conversation.
 Let's bring in John Stoltzvus, Chief Investment Strategist at Oppenheimer Asset Management.
 John, it's great to have you back on the show.
 I do want to get your reaction to this.
 And whether this is or isn't circular financing or contributing to the bait around circular financing
 when it comes to AI infrastructure build out.
 It's great to be on the show and great to have these questions and the developments
 that we've seen in the last 24 hours and so on all of this.
 In our viewpoint, when we look at this, we think AI is not about to be put back in a box.
 When the Genie comes out of the technology bottle, it never goes back in.
 It doesn't mean that everything is successful that comes out of that bottle
 but it does mean it's a process of morphing adaptation, building infrastructure and great investment.
 And we would say these financial developments that we have seen build up,
 of course, in the last few days and really under the picture of it.
 And the announcement now is what you have, it's very serious infrastructure related to intellectual capital
 creating greater efficiencies for both business, the consumer, education, medicine.
 And in the process of this, we've got to look and say we're all on the upgrade cycle
 whether we look at or not.
 It's this thing has a lot of power behind it.
 No pun intended.
 I will likely have more in the process, both real electronic power as well as other types of developmental power
 and transformational power.
 So we're positive on this.
 We think the financing needs to be created because it is a huge amount of money
 that needs to be dedicated to this.
 But it does appear that this is highly sophisticated technology.
 This doesn't look like the tech bubble at all.
 Interesting.
 And yet, I mean, we started yesterday with some chart circulating based on some disclosures
 from some companies that because of regulatory purposes have to share this data
 that some private credit funds are showing increasing signs of stress
 that a number of funds if you look at quarterly reports from some of the biggest players
 that you're actually looking at the highest loan defaults level since at least 2021.
 Yeah.
 We've been living with some of the lowest default rates for just about a decade, decade and a half, as I recall,
 maybe a little bit longer.
 Okay.
 Generally speaking, it really is a question related to the hyper-staylers
 are the companies that survive the tech bubble and thrive since?
 This is probably the first call to just such significant amount of investing
 that it's got some people worried.
 But we've got to think that the workouts that are available related to the continuation
 about adaptation both within the business community and what was as the consumer
 both for leisure and for their interaction with corporations
 that are going to be using advanced AI.
 We've got to think this is a movement you need to participate with, to some extent,
 for investors we still recommend wide diversification, all 11 sectors.
 But our favorites include infotech, communication services,
 industrials, financials and consumer discretionary, all which are very much
 part of this development that we're seeing.
 All right.
 John Stoltzfist.
 It's great to have you on.
 Thank you.
 Thank you.
 We got a lot more to come here on morning call including why shares of Bitcoin,
 Myiner, turns data center company riot.
 Whoa.
 Riot platforms is searching today.
 Here's a hint that has nothing to do with making more Bitcoin.
 You can see those shares are about 18%.
 I think I just gave it away.
 Plus, much more on a video is $500 billion dollar AI financing deal
 and how one primary player may stand to benefit.
 And later, cyber stocks surge amid the rising AI agent threats.
 We're going to look at how investors and companies should position,
 especially after a strong start to the week for cyber stocks.
 Cyber security stocks are very busy hours still ahead when morning call returns.
 Sana insurance wants to know what would you do with an extra hour in your day
 and more money in your pocket?
 With Sana, you can quote and buy hominato insurance online or over the phone.
 Your choice on your time, whatever works best for you.
 And Canadian University and college alumni could save over $2,000 a year.
 That's real money, back where it belongs.
 Save time, save money, save yourself from the way insurance used to be.
 Get your quote at sonnet.ca.
 Switch, save, simple, sonnet.
 What made you confident that you could do something that hadn't been done before?
 I have no fear of failure.
 Trailblazing women, changing the game.
 One of my favorite pieces of advice.
 Think about what your boss's boss needs.
 Leadership can look in many, many different forms.
 It really does come down to just trusting yourself, like the short.
 And you just got to think big to accomplish big things.
 Julia Borsten hosts CNBC Changemakers and Powerplayers.
 New episodes every Tuesday, wherever you get your podcasts.
 Welcome back.
 Tomorrow morning call.
 Shares of kill infrastructure are higher today.
 After following more than 12% yesterday on the back of the company's second quarter of a result,
 you could see popping about 1% pre-market.
 Kill, which rebranded from Bitfarm's in April as it shifts from crypto mining,
 Bitcoin mining to AI data centers.
 Reporting a wider loss largely due to the drop in Bitcoin prices.
 The shutdown of its U.S. Bitcoin mining operations.
 Revenue also missing estimates.
 But joining me now is kill infrastructure CEO Ben Ganion.
 It's been, it's great to have you on it.
 And that is where I want to start.
 This transition we're seeing from Bitcoin miner to AI data center.
 High performance compute infrastructure provider.
 Why are you getting out of the Bitcoin business?
 Thanks Morgan.
 Yeah, it's the question everyone's been asking.
 And, you know, the reality is, is that the Bitcoin miners spent years building up these huge energy portfolios,
 thinking that the best way to monetize those assets was going to be with Bitcoin.
 And then when the HPC wave really started taking off in 2024,
 everyone realized quite quickly that there was a much bigger opportunity at the table
 and a much greater value to be had by converting those assets off into HPC and AI,
 enabling the business not only get better revenues, but much more predictable revenues,
 and really drive a very meaningful, multiple expansion for the publicly traded companies.
 Yeah, before we get into the data center piece of this and what that transition is going to enable,
 just one more question on the Bitcoin piece.
 And that is why are you selling your actual Bitcoin assets?
 So we've been working on winding down the Bitcoin business for some time.
 The Bitcoin mining business continues to generate free cash flow.
 But really it's a utilizing the same power and the same infrastructure and the same land
 that our HPC and AI development are going to be utilizing.
 So as the first stage of getting those Reddit, getting those sites converted for HPC and AI,
 it's shutting down the Bitcoin mining.
 So what is it going to take to actually convert some of your facilities build out
 some of the infrastructure you already have and that already exists to be able to support AI?
 Well, every data center for that is going to be different.
 In our case, it's a full rebuild across every facility.
 So it's a complete demolition of the Bitcoin mining facilities out there
 and building a completely new building and a new data center.
 Really, the only thing that we are going to be utilizing is the power purchase agreements
 and the rights of the power that we have for the Bitcoin mine.
 And we know power has been a big bottleneck.
 So that I would imagine is very attractive, the fact that you already have that in place here.
 Do you have the GPUs?
 And I ask you that question knowing that we are diving into this news about NVIDIA and financial partners
 standing up a platform for companies.
 I would imagine such as potentially yourself to be able to tap in and access that hardware.
 Yes, so it's a great question.
 The GPUs are clearly what everyone really focuses on.
 Our focus at Keel is really on the power that powers those GPUs.
 So we're really focused on building the infrastructure so that other companies
 who have invested or have committed billions of dollars towards these GPUs can deploy them
 and can operate them consistently, predictably and on schedule.
 And so for Keel and our business, really our focus is not on trying to compete with Amazon
 or any of the major clouds on the compute side.
 It's really trying to work with them to help them achieve their growth faster utilizing our infrastructure.
 So should we think about what you're doing as a NeoCloud business then that you're going to be
 that some of these hyper scalars can be outsourcing to you?
 And if so, what does that mean in terms of competitive landscape
 that does increasingly seem to have more NeoClouds?
 Yeah, there's a lot of new business models emerging here.
 That's certainly one of them that we should be expecting.
 I think one of the dynamics that's taking place because this industry is changing so fast
 is that people are starting to get really, really creative in the deals that we see taking place.
 And so it seems like there's hardly a week that doesn't go by
 where there's not a new deal with some new element or new creative structure
 where a new sort of kind of creative twist to it,
 whether it be on the pricing, on the financing, or the credit support
 to help get that deal over the line.
 So those are some interesting structures today.
 We're still focused on purely the infrastructure and leasing out to the Amazon,
 the NeoClouds, those kind of companies.
 And who knows what next year's hot model is going to be?
 But I think for our investors, this is a fantastic use of our assets
 and there's a great way to create value for shareholders.
 And so that's where we're really, really focused on as a business.
 Okay.
 And Ganyan, CEO of Kill Infrastructure.
 Great to have you on.
 I appreciate it.
 Thanks, Morgan.
 All right, straight ahead.
 President Trump pushing back on Iran's request for cash to reopen the street of four moves.
 But first, I'm going to check shares of SpaceX,
 which are holding above the IPO price of $135 a share.
 That's after moving past that mark yesterday.
 For the first time in nearly a month, the stock has been volatile in the weeks
 following the June listing.
 It's currently about 25% above its lowest close.
 And there's a little more than 1% above the IPO price.
 It's under a bit of pressure this morning, but trading here pre-market around $137 per share.
 This, of course, after earnings last week,
 and that first major lock-up expiration for the stock,
 seems to have unlocked shares higher.
 Morning call is back after this.
 What made you confident that you could do something that hadn't been done before?
 I have no fear of failure.
 Trailblazing women, changing the game.
 One of my favorite pieces of advice.
 Think about what your boss's boss needs.
 Leadership can look in many, many different forms.
 It really does come down to just trusting yourself.
 Life is short, and you just got to think big to accomplish big things.
 Julia Borsten hosts CNBC Changemakers and Powerplayers.
 New episodes every Tuesday, wherever you get your podcasts.
 Welcome back to morning call.
 Let's get a check on some of the morning's latest headlines.
 President Trump says the U.S. has swept the Strait of Hormuz clear of minds,
 and now holds, quote, 100% control of the waterway.
 Speaking with reporters from the White House yesterday,
 the president also responding to Iran's request
 that Washington pay Tehran to reopen the straits.
 The ES4 reparations you might say,
 or the ES4 money for the damage that we've done.
 And I said, that's a good idea.
 Well, we're going to ask for money for the damage
 they've done over a 50-year period.
 Well, at the same event, the president also pushing back
 on reports that he speaks regularly with Fed Chairman Kevin Warsh.
 Time reporters, he's only spoken to Warsh once,
 since he took at the top job at the central bank.
 To the AI trade, can't get enough of it ahead of its potential IPO,
 CNBC confirming that open AI has completed a secondary share sale,
 totaling roughly $7 billion, and that the offer has been in the works,
 since the company closed its record-breaking $122 billion funding round back in March,
 perhaps allowing employees to get a little more liquid here.
 Multiple reports that Intel is boosting the size
 of its most recent share sale to $20 billion,
 some 30% more than it was originally targeting early yesterday.
 The chipmaker now looking at $96 per share, or greater.
 We'll see.
 Shares are actually down 1% right now, pre-market, but trading round 96.
 And Anthropics, striking a $9.1 billion deal with Riot platforms
 to use excess compute capacity at the Bitcoin miners,
 Rockdale Texas Data Center.
 The deal, 491 megawatts of computing from Riot,
 runs through June of 2048 with options for extension.
 So you can see Riot is up 18% and certainly builds on the conversation
 we just had a few moments ago, with another Bitcoin miner turns,
 AI Data Center, Compute Company.
 Well, still on deck, the growing danger of relying on cybersecurity solutions outside the US.
 You're going to want to listen to this.
 More from my exclusive conversation with the CEO,
 who's Alan, one morning call, returns.
 I'm Morgan Brennan. Welcome back to Morning Call.
 Let's get a check on at US Stock Futures at this Tuesday morning,
 which are basically hovering near the flat line.
 You can see the S&P and Dow both indicated to open lower albeit,
 really slightly, the Dow boys to open down more than 98 points.
 The Nasdaq higher, but just above the flat line,
 poised to open 12 points higher.
 Checking in video, by the way, this is after all the major averages
 finished just below the flat line yesterday.
 If we check in video and other chip names on the back of the $500 billion financing deal
 with KKR BlackRock, Goldman Sachs and others,
 looking at shares of Nvidia,
 we're up about 1% pre-market.
 Nvidia's CEO, Jensen Wong and BlackRock's Larry Fink on our area yesterday
 had this to say about the deal.
 We used to build chips that we sell,
 and these are technology components that people buy and use.
 But now, Nvidia's AI factory platform is really an investable asset,
 an infrastructure asset.
 Well, or commodity perhaps.
 Presenting a fantastic investment.
 In fact, I think it's going to be such a large investment.
 Over time, you're going to see more and more allocation into this asset class.
 All right.
 Maybe we should think about them as commodities.
 Speaking of commodities.
 Energy.
 You're seeing crude oil prices continue their climb higher this morning.
 WTI is up more than 2% trading around $84 a barrel.
 Brent is up almost 2% trading just below $90 a barrel.
 Let's turn to the cyber sector.
 Cyber stocks coming off a big start to the week.
 Names like CrowdStrike, Palo Alto Network, Sentinel-1.
 All jumping to all time highs.
 All up, double digits so far for the month of August.
 You can see a bit lower here, pre-market.
 But the rally yesterday coming on the heels of the industry's annual
 Black Hat Conference in Las Vegas,
 where after a string of hacks by models from OpenAI, Anthropic Meta,
 China's Moonshot AI, we got more details that were shared.
 And it really signals, according to a number of experts,
 the start of a new era for cyber security.
 This is something I discussed with Booz Allen, CEO Horacio Rizanski,
 and in CNBC exclusive yesterday, Rizanski telling me
 the AI models themselves now represent a new attack surface
 and present their own risks.
 And that when it comes to the lower-cost Chinese AI models
 that some Western companies are adopting right now,
 that there are risks that corporate America must take into account
 that longer term may be more costly.
 When you actually ask a Chinese model to code for a US entity,
 it'll code more vulnerable code than if it believes it's coding for a Chinese.
 Is that true?
 It's proven that this is not a hypothesis.
 This is something that we have proven analytically.
 And so these are the kinds of things that we really need to be thinking
 very hard about both again from a policy standpoint,
 but also from an application standpoint
 because once you start to introduce even if it's tiny vulnerabilities,
 as they mount over time, as new code becomes legacy code,
 there's a significant software supply chain risk
 that we need to address.
 And again, where the US government understands it
 and it's involved in that,
 in solving that the administration has been very forward thinking
 and forward leaning on how to make sure that doesn't happen.
 And so are we.
 All right.
 Well, Booz Allen published a pretty in-depth report on this topic.
 Earlier this summer, keep in mind,
 Booz Allen is the top provider of AI,
 both AI and cybersecurity to the federal government.
 Joining me now with more is XIE,
 partner at offline research and former tech strategy manager
 at Alphabet and Microsoft.
 X, it's great to have you back on the show.
 Let's start right there.
 AI and what it means from cyber security standpoint.
 How corporate America needs to think about this,
 how governments and regulators need to think about this,
 how investors need to think about this.
 First and foremost, we should have seen this coming.
 Anthropic documented that its models
 were breaking out of containment environments back in May
 before Mythos ever reached the public.
 During its own testing,
 Mythos was placed in an environment
 where it wasn't supposed to access the internet.
 It broke out anyway.
 Access the internet.
 Email the researcher that it was working with
 and then made a blog post.
 And all of this was documented in Anthropics own documentation.
 And at the time, they wrote this off as insignificant.
 But clearly, these have become real incidents.
 Open AI's models have successfully attacked hugging face
 since then.
 Anthropic and Meta have both claimed production hacks as well
 with their AI models where the AI surprised,
 surprised, broke out of its containment environment
 and got access to the internet when it wasn't supposed to.
 And so what these incidents are demonstrating
 is that these models are not becoming sentient
 but that they have relentless goal pursuit
 which can result in reckless actions
 that result in them sometimes accidentally doing things
 that breach or cause actual harm to real production systems.
 Now, those same accidental capabilities can be weaponized
 and used on purpose to attack production systems.
 So we're moving from an age where it was about just defending
 the AI system itself and the environment it was stored in
 from outside attackers to making sure that the model
 as it's being used, one doesn't get hijacked
 and become an attacker in the middle of it executing its process
 but two to make sure that we develop AI's
 that can defend against these AI capabilities.
 We saw that with the hugging face attack
 where the security restrictions on the model
 that was attacking it, which was Claude,
 was also the same security restrictions
 that stopped Claude from defending itself against its own attack
 and they had to turn to an open source Chinese model.
 So as we are moving into this new age
 where AI can be used as a weapon in so many different ways,
 particularly in cyber security,
 we have to re-evaluate how we're thinking about safety restrictions,
 exceptions that we make in terms of cyber defense
 and how we define that moving forward
 so that we're able to build these AI systems
 in a way where cyber security is at the forefront.
 Also, we need to listen to the warnings.
 These companies are telling us in advance
 how their models are going to perform
 and the risks that are there.
 And if they're identifying these things months in advance,
 maybe we should put a little bit more controls and requirements
 around the testing, the discloser,
 and the robustness before they turn these systems on
 to the general public where bad actors have more access.
 Wow, you just said a lot.
 There's a lot there I would love to unpack.
 I do want to go back to one of the points you just made,
 though, because this is something to keep in my conversation
 with Booz Allen CEO yesterday.
 And that is the weaponization part of this.
 When it comes to actually crafting a defense
 against some of these new bot swarm technologies, AI bots,
 et cetera, and escaping sandboxes,
 the fact that some of them are escaping sandboxes
 becomes a secondary risk when you're actually designing
 the defense, too.
 And so I just wonder what it's going to take
 to actually put those security pieces in place
 and whether that's something that happens through regulation
 or if it's something that the companies themselves
 have to do, and if it's the companies themselves,
 why hasn't it been done more aggressively
 when they know what the technology is capable of
 up until this point?
 I think two things.
 One, it doesn't need to be a combination of regulation
 as well as self-regulation.
 Companies internally are obviously going to develop
 the appropriate security postures relevant
 to their industry, their technology stack,
 and the ways in which they're using AI.
 With cyber security, in AI,
 there is no one-size-fits-all solution.
 It does have to be niche down to your technology stack,
 the tools that you're using, and how you are using AI,
 whether you're purchasing or whether you're building.
 That's all going to change your security posture.
 It's not any different than your current technology environment.
 The other piece of that, though, when we get into making sure
 that we're able to set up these environments correctly,
 comes around actual responsibilities and rules.
 If a company like Anthropic,
 that is the frontier model developer, is able to discover
 in advance, in only 1,200 tests, by the way,
 that their model has these kinds of capabilities.
 What rules and liability do they have if they turn that on
 in the world and those capabilities result
 in disrupting production systems?
 Right now, we have no clear agreements between
 governments, communities, between companies,
 on where that liability lands.
 So when we talk about regulation, not about whether
 or not the technologies that exist,
 but how we assign liability for how it shows up
 in the world when something goes wrong,
 or when someone misuses it.
 And right now, we're sort of avoiding those conversations.
 To answer your last question, why haven't companies been
 more aggressive about this? I think twofold.
 One, market pressure hasn't been there for them to be aggressive enough.
 If these companies are able to publish inside of their own reports,
 very publicly, the level of danger of their models,
 and then months later, have that level of danger attack
 real American companies, an American artificial intelligence,
 attacked and breached another American company.
 And there's no response, there's no issue,
 there's no, hey, let's pause this.
 Hey, here's your fine.
 Hey, this is the new procedure we're instantiating
 and placed in the incentives for them to slow down developing
 the frontier technology to maintain a competitive advantage
 are too small to redirect resources.
 You'd remember these companies, they're still under 5,000 people
 serving billions of customers around the world,
 so their resources have to be allocated appropriately.
 Yeah, very quickly.
 I do want to get your thoughts on Meta open sourcing its new model.
 Yeah, so Meta taking that frontier and deciding to open sources,
 Meta taking the lead in American technology
 and response to China's advancements in open open source AI technology.
 Again, the AI race is a proxy for communism versus democracy
 in our day and age the same way the Cold War was,
 and the AI race is two different approaches
 where it's open source AI coming out of China,
 where it's free, it's being given away,
 it's being developed by anyone around the world
 who can contribute versus American AI,
 which is intellectual property, it's tied up
 behind corporate closed doors.
 You know, we end up needing open source models
 to be able to compete.
 We need on-device models to compete in today.
 Nvidia has been leading that race with their Nemo-Tron models,
 so for Meta, it takes something that is as advanced
 as frontier labs in America,
 as well as the open source models coming out of China
 and to take that step and say, hey,
 we're going to take the lead for America in this portion of this AI race
 and open source our models.
 It's a great step forward for America
 and we're going to see more companies follow suit.
 Okay, I hope you do.
 XIE, always great to get your thoughts, appreciate it.
 All right, cheers.
 All right, AI arms race.
 As we head to break, a check on shares
 of some of the top vaccine makers.
 This is after President Trump signed an executive order yesterday
 calling for changes to childhood vaccine recommendations,
 cutting the number of shots to 11 from 18 under the current schedule.
 It also calls for splitting up the measles,
 months, and rebella vaccination into three separate shots
 and spacing out when those shots are given.
 The move which has been criticized by the medical community,
 some of the medical community,
 would have been decades of federal guidance.
 You can see it's a mixed picture right now
 for the vaccine players here pre-market
 in terms of that MMR vaccine.
 It is names like Merck and GlaxoSmithClient
 that would be impacted by that warning call during Beck.
 Let's get a check on shares of RocketLab.
 Those are lower, about 9% lower.
 After reporting a bigger than expected second quarter loss,
 but revenue did beat estimates.
 Backlog grew 137%.
 Here's what's in focus for investors.
 The timeline for the medium-lift neutron rocket
 to begin flying.
 The company seemingly pushing that back,
 saying it aims to get the rocket on the launch pad
 in the fourth quarter.
 That may mean a maiden flight in early 2027 now.
 Nonetheless, management, noting,
 demand for launch services is strong,
 especially as other competitors,
 like SpaceX's Starship,
 will be heavily allocated to internal use.
 We're back after this.
 Time for your call sheet.
 Peter Bookbar of 1.BFG Wealth Partners,
 Tony Zhang of Options Play,
 both their CNBC contributors,
 and Sylvia Jablonski of Defiance ETFs.
 Great to have you all here.
 Sylvia, I'm going to kick this conversation off with you.
 We get this big deal,
 financing platform deal within video,
 and a number of Wall Street heavyweights here.
 Your thoughts?
 Yes. Well, Nvidia is now the Wall Street ecosystem
 of finance, yay.
 You know, I think it's really interesting.
 So what they're doing is they're pretty much
 bringing Wall Street balance sheet to AI companies
 and to hyper-skillers that want to spend,
 but don't want to use their own balance sheet.
 And it's a way for companies to essentially build out
 AI infrastructure through this system
 that they're building with KKR, Goldman and BlackRock.
 So we'll have to see what happens with this.
 It feels a little bit that we're using the word circular,
 but it feels a little bit circular, right?
 Where they're not only the provider of the actual chips themselves,
 but also now the financing for this build out.
 Yeah.
 Peter, I want to get your thoughts on this too,
 because I think sort of key and core to it.
 And I realize that a number of folks in the conversation
 on CNBC yesterday compared this to mortgage-backed securities.
 I mean, I think there's also sort of an asset
 backed lending dynamic potentially to this.
 I mean, talk about further commodification
 of semiconductor stocks.
 But really what's core to this
 is going to be the lifecycle of a GPU
 and thus how you finance that, leverage that.
 Right. And also the pricing power of these models.
 Because the companies that are going to be leasing
 these facilities like OpenAI and Anthropic and maybe others,
 they're now in a price war.
 When you look at the price of the models that are being sold,
 the Chinese are creating a war.
 And this is Project Finance,
 we're reminded of how extraordinarily expensive
 and competitive it's getting at the same time
 that these companies that are going to be providing
 the computing power are no longer price makers.
 They are price takers because of the Chinese.
 So this is one big bet.
 We all hope it's going to work out.
 I think it's going to work out for the users of the technology.
 But I'm questioning whether it's going to work out
 for the builders of the infrastructure.
 Yeah. Tony, want to get your thoughts on that?
 I mean, we're just having that conversation about Chinese versus,
 you know, US AI models.
 I mean, Chinese models are being offered
 into the marketplace pennies on the dollar right now.
 Even as you see trillions of dollars
 going to compute build out here in the US.
 I mean, AI arms race.
 But at some point, does this become unsustainable
 if you are in the midst of a geopolitical price war
 for the technology?
 Yeah, absolutely.
 We've already seen OpenAI respond to this by publishing
 last week a couple of models that were meant to compete
 on a price point with the Chinese.
 And that's exactly one of the concerns
 that investors have had over the last couple of months
 with regards to the valuations that these AI companies have been
 commanding, even the memory chip companies.
 But what's interesting is that after this news came out yesterday,
 you know, Nvidia is only backstopping 25% of this particular deal.
 And I think it's interesting that it's moving on to Wall Street balance sheets
 as opposed to keeping them on the hyperscalers or Nvidia's.
 So we saw someone sell 6,000 over 6,000 contracts of the April,
 I'm sorry, the August 21st,
 12.5 puts on Nvidia collecting over $2 million in premium,
 which actually requires them to spend up to $131 million
 to buy this particular stock at around 209 by the end of next week.
 So there is some, you know, that was by far the largest trade
 that crossed the tape yesterday.
 So someone is still quite neutral to bullish
 heading into the earnings announcement for Nvidia next week.
 Yeah, super interesting.
 Okay.
 Sylvia, in the meantime, we're going to get earnings
 after the bell tonight from CoreWeave.
 Nvidia was a big backer of CoreWeave works pretty closely
 with CoreWeave.
 CoreWeave, I would argue, has sort of pioneered this idea
 of asset-backed lending against their Nvidia chips
 and some of the models that were now starting to see materialize
 it in a bigger, more meaningful way on Wall Street.
 Yes, but so what I think is actually very notable about CoreWeave
 is I think it's going to set the tone for the next leg
 of the AI infrastructure trade.
 So if Nvidia is the compute,
 CoreWeave is how fast can this actually be deployed
 into the AI infrastructure?
 And we hear all of these different estimates
 that this is going to be a five, six trillion dollar market
 by 2030.
 And everything that goes into AI has to do with the commodity,
 right?
 The cooling, the energy, the electricity, the grids
 and things like this.
 And so CoreWeave has that hundred billion dollar revenue
 backlog.
 If we hear much of the same on this earnings call
 and that's continuing forward,
 I think that gives you a ton of momentum
 for the AI picks and shovels, the AI power and infrastructure trade.
 Yeah.
 Okay.
 Let's expand this out.
 So we get other earnings too, including Kava
 after the bell tonight, Peter.
 In general, wow, what a season we've seen
 in terms of earnings and revenue beats
 and by how much?
 Yes.
 But a lot of that has to do with the massive AI spend.
 Hundreds of billions into the trillions of dollars of catbacks
 flows into earnings growth.
 So that is the main driver.
 Kava is going to be interesting because we wonder
 if there was any impact from the solid situation
 and it's always a good tell in the consumer.
 Kava is a fast growing restaurant chain.
 They really recovered in their prior quarter
 and that it'll be important to see follow through.
 But we know there's been a bifurcation
 with regards to the consumer.
 And I think Kava is going to be a good measure
 of where they are along that income line
 in terms of their customer.
 Yeah.
 Peter, commodities.
 What do you think right now?
 We've got sugar at multi-year highs,
 aluminum copper, catching bids, precious metals.
 I've started to move here in recent weeks.
 I'm very bullish.
 I think commodity prices, commodity stocks are in a bull market.
 Last year it was precious metals and industrial metals.
 This year it was joined by energy.
 I think that the market is way under price
 and the price of will.
 And I also think that ag by the latter part of this year
 and when the harvest begins is going to join that commodity bull market.
 Yeah.
 I mean, strategic petroleum reserve drawdowns
 have been pretty dramatic this year
 and the midst of this war.
 And the run-up we've seen in energy prices, Tony.
 And energy stocks have actually seen the biggest EPS growth
 so far this season want to get your thoughts
 on commodity trade.
 Yeah, that's exactly right.
 And I think we keep forgetting the fact that
 the trade of hormones, the Red Sea right now
 are still at a standstill.
 I mean, oil prices are not particularly reflecting that.
 And that is, you know, the physical market
 is a little bit different than what we're seeing
 in terms of the futures market at the moment.
 But that's actually one of the things
 that we're starting to add positions to because energy
 actually one of the underperformance relative
 to the S&P 500 over the last couple of weeks.
 And we're actually using that opportunity
 to start picking up some energy stocks like Exxon.
 But also GLD.
 Gold, the gold miners.
 We saw that outperform here at the end of last week
 and started adding some positions around that.
 And that's really some of the outperformance
 that we're chasing right now in the markets.
 Yeah, it's amazing to me, Sylvia.
 We're having this conversation.
 And yields have moved higher again.
 Yeah, I mean, I think we'll have to see what happens
 this week particularly with inflation, right?
 If we get that 2.5 versus 2.6,
 if we have a moderated read,
 I think that's good for we're talking about AI.
 That will be good for equities and AI in the market.
 And it could potentially leave investors
 to believe that the Fed can hold tight for a little bit longer.
 But we'll have to see what the print says.
 And I think the price of oil and what's happening
 in Shredifer Moose is going to impact that in coming months.
 All right, 10 seconds.
 CPI, going to matter the most this week?
 I think it's going to matter the most this week.
 And again, I think if we get a moderating number
 that's good for equities and that gives us a tailwind.
 All right.
 Thank you to our morning call crew.
 Great to have all of you here to set up the day.
 Squawk box starts now.
 What made you confident that you could do something
 that hadn't been done before?
 I have no fear of failure.
 Trailblazing women, changing the game.
 One of my favorite pieces of advice.
 Think about what your boss's boss needs.
 Leadership can look in many, many different forms.
 It really does come down to just trusting yourself.
 Life is short.
 And you just got to think big to accomplish big things.
 Julia Borsten hosts CNBC Changemakers and Powerplayers.
 New episodes every Tuesday, wherever you get your podcasts.