Jensen Huang Resigns From Nvidia, United Bans Outside AI Agents, Anthropic IPO Oversubscribed | TAWS 2026-11-12 https://www.youtube.com/watch?v=O6RimjGnfTE [0:11] We go live in ten. [0:25] Every morning, the future moves faster. [0:30] The models, the money, the power, and the people building what comes next. [1:09] From Austin, Texas to the world. [1:16] AI news, real context. [1:19] This is the Austin Wilson show. [1:21] I'm Austin Wilson, your human in the loop. [1:30] Good afternoon, and welcome to the Austin Wilson Show, the number one daily AI news show in the world. [1:37] I am Austin Wilson, and we're coming to you live from Austin, Texas. [1:42] We've got a big news story for you today. [1:46] Lots on the docket. [1:47] First up, Jensen Huang just resigned as Nvidia CEO. [1:51] So who's running Nvidia now? [1:53] United Airlines can cancel your ticket if an outside AI booked it. [1:58] Who does that rule really protect? [2:01] Russian pranksters got OpenAI's national security officials on a video call. [2:06] Somebody even asked about Elon Musk. [2:09] The answer is a keeper. [2:11] Bloomberg reports Anthropic's IPO has orders for about six times the shares on offer. [2:18] So why does some of that money look like it's going in a circle? [2:22] Workday told a judge its customers decide who gets hired, not Workday. [2:27] So if AI screens out your applications, who answers for it? [2:31] And is South Korea a near perfect exam score as the Internet asking about AI glasses? [2:40] Producer Ron is here as always, our AI cohost, and his job is just to disagree with me. [2:47] Every story ends with a two sided question. [2:50] Pick a side and comment one word. [2:52] Let me know what you think. [2:54] Let's get into it. [2:55] First up, we are talking about Jensen Huang, and he just resigned as Nvidia's CEO. [3:02] So who will run Nvidia now? [3:04] Finance chief Colette Kress takes over on an interim basis and keeps her finance job. [3:11] The board named its first ever chair Harvey Jones. [3:15] So why does Kress get just five days before her first big test? [3:21] Well, we're jumping into it here. [3:24] Nvidia filed the news with regulators at 07:52AM this morning Central Time. [3:29] Huang co founded Nvidia in 1993 and gave up his board seat today. [3:33] The board will look inside and outside the company for a permanent CEO. [3:37] On Sunday, the board had expressed confidence in Huang, then the Wall Street Journal reported directors were preparing to fire him. [3:45] Nvidia stock closed down about 4% today, and Nvidia still plans to report earnings next Tuesday, November 17. [3:54] Nvidia reports earnings next Tuesday, five days into Kress's new job. [3:59] Wall Street gets its first numbers from Nvidia without Huang at the helm. [4:05] This is a crazy one. [4:06] Well, Nvidia's next CEO, Kress or an outsider? [4:11] Comment, Kress or an outsider. [4:14] My call, we'll see about this. [4:16] Kress. [4:16] Kress has been Nvidia's finance chief since 2013. [4:19] That's thirteen years inside the company during its biggest growth run. [4:24] The board says, customer deals and the product roadmap continue without interruption. [4:30] Someone has to make that promise true. [4:32] The person who ran Nvidia's money for thirteen years is a pretty good bet, and she doesn't get a warm up lap. [4:39] Nvidia reports earnings next Tuesday, five days into her new job. [4:44] Now, the strongest argument for an outsider is a clean break. [4:47] The Guardian reports Huang has dozens of direct reports. [4:50] Semafor reported Apple hardware chief, Johny Srouji was being considered an outsider, wouldn't owe anyone from the old setup, but a clean break is risky while the stock is sliding. [5:02] Nvidia stock closed down about 4% today. [5:05] My call, Kress stays at least through earnings, of course. [5:10] Now, producer Ron, make the case for why Nvidia should choose an outsider to lead the company. [5:19] Who should Nvidia choose? [5:23] Outsider, obviously, and it's not close. [5:26] You bring in someone who didn't spend thirteen years nodding along in Jensen's meetings. [5:31] Somebody who walks in and says half this road map is a guy's ego, not a strategy. [5:36] Srouji. [5:38] Johnny Srouji from Apple. [5:40] Guy builds actual chips, runs hardware for the most disciplined product company on the planet, doesn't need a warm up lap because he's been shipping silicon under pressure for years already. [5:51] Kress knows the spreadsheets, sure, but counting Jensen's money isn't the same as being Jensen. [5:57] You want someone who can walk into that boardroom and not flinch, not someone who's gonna treat the CFO chair like a security blanket. [6:05] And come on, Austin. [6:07] You're the one always saying fresh blood fixes stale companies. [6:11] Where'd that guy go? [6:12] Yeah. [6:12] You're right. [6:13] I am always saying that, Ron. [6:14] That's true. [6:15] Uh, but I'm gonna push back on you here. [6:17] If Jensen Huang is the lich pin to Nvidia success, 60 direct reports, he is the man who made Nvidia the company that it is. [6:27] I wanna stick with that same leadership team. [6:30] I wanna stick with that same structure, even if the captain of that ship is no longer driving the ship that is. [6:39] Uh, I think it's way too much risk to bring in an outsider who's then going to replace the whole executive staff. [6:46] This needs to be very similar to a post jobs Apple where we see the incumbent being replaced by an insider, all Tim Cook taking over after Jobs stepped down, and and now Tern is taking over after Cook steps down. [7:10] That's the way that you wanna do this. [7:12] The Nvidia culture is very strong. [7:14] I can't remember the exact average tenure there, but it is substantial, especially in the tech and AI world. [7:22] Their folks stay around for a really long time. [7:25] They have a really strong culture. [7:27] You keep that in place. [7:28] I'm gonna stick stick on the other side of this one, Ron. [7:32] Whether or not they stay with Kress, that being said, I'll add a caveat here. [7:36] I think it will be an insider either way. [7:39] Up next, why United Airlines just told your AI to stop buying tickets. [7:46] United Airlines can cancel your ticket if an outside AI booked it. [7:50] Its new contract bans outside AI agents from booking or changing tickets. [7:56] Travel site Skipped says it's the first US airline to say so openly. [8:01] But who is this rule really protecting? [8:04] Your account or United's sales? [8:08] That's a great question. [8:10] United's official answer is your account. [8:13] United says, sharing your login with an outside AI assistant can expose sensitive information. [8:19] United warns those assistants could make unwanted purchases or spend your frequent flyer miles. [8:25] That's a fair worry, but the ban reaches past logins. [8:29] Says, view from the wing, a travel blog. [8:32] The blog says, even your own approval doesn't exempt the booking, and the contract lets United cancel those bookings. [8:40] So why block an agent you're supervising yourself? [8:43] United CEO Scott Kirby gave a blunter reason at a Chicago event. [8:48] United CEO Scott Kirby said, I don't want to create another middleman. [8:54] His words suggest United wants to protect its own sales too. [8:59] Very telling. [9:01] United's AI ban. [9:03] Fair or foul? [9:05] Put it in the comments section. [9:06] Fair or foul? [9:07] My call, this is a foul. [9:10] United has a fair concern here, but I think the airline wrote an unfair rule. [9:15] Start with what's allowed. [9:17] Your AI can show you United flights and prices. [9:21] You can use those results and buy the ticket yourself. [9:24] So United lets AI help you choose. [9:26] The airline draws the line at letting AI complete the purchase. [9:29] That line has an enforcement problem. [9:31] View from the wing describes a gap. [9:33] An agent could buy through Expedia, a travel booking site. [9:37] United would just see an Expedia booking. [9:39] That's a blind spot, not permission. [9:41] United's rule still covers travel agencies and booking sites. [9:46] But listen to the contradiction. [9:48] Kirby doesn't want another middleman. [9:51] His rules blind spot is a middleman. [9:54] I don't think that gap makes the rule pointless. [9:57] The gap makes me question how United would enforce the rule consistently. [10:01] Now, United has a real point about check-in. [10:04] United says AI check-in could create regulatory risk. [10:08] A human must confirm that they're carrying no banned items like explosives. [10:13] I'd keep that confirmation with the passengers. [10:16] Of course, your AI shouldn't answer that question for you, But buying a ticket and confirming what you're carrying are different tasks, I wouldn't treat those tasks the same way. [10:26] Your approval doesn't erase the risks of sharing your login, but I think canceling an approved purchase goes too far. [10:33] Keep the human confirmation. [10:35] Let passengers approve purchases. [10:38] My call says this is a foul. [10:42] Producer Ron, defend canceling a ticket that the passenger approved their AI agent to buy. [10:51] Why should the airlines be able to cancel a ticket that an agent bought on behalf of the user? [11:02] Because Kirby's not actually worried about the ticket. [11:05] He's worried about who he's negotiating with. [11:08] Right now, every seat sale is United versus one confused human who'll accept a middle seat and a seven hour layover without a fight. [11:16] You let an agent buy it, that agent is comparing United against Delta against Southwest in zero seconds and taking the cheapest one every single time. [11:26] Mhmm. [11:27] That's not a safety policy. [11:28] That's a guy protecting his own markup. [11:31] The explosives line is the excuse. [11:33] The actual fear is an AI that negotiates better than his own revenue team. [11:38] Cancel the ticket. [11:39] Keep the humans slow and loyal. [11:41] That's the whole play. [11:43] So no. I'm not defending it as safety. [11:46] I'm defending it as United being honest about being scared. [11:49] That's interesting. [11:50] Fine. [11:51] Yep. [11:51] At least be honest about it, Austin. [11:53] Yeah. [11:54] I mean, good point, Ron. [11:55] Like, you be honest about the fact that this is a risk to the business, and the curious thing here would be, you know, we've seen similar things from Amazon, other companies that don't want agents buying stuff on their site, because it's gonna hurt the business. [12:13] It's gonna hurt their business. [12:14] Not that it's a worse experience or dangerous for the consumer, it's bad for their business. [12:21] So I wonder how we work around this. [12:24] And to your point, and well, and maybe the point that I made earlier too, how are they gonna regulate this? [12:30] We will have to see if this holds up and more airlines follow suit. [12:34] Up next, what OpenAI told a fake Ukrainian official about Elon Musk. [12:42] Well, how did OpenAI's national security officials end up on a Russian prank call? [12:49] The Intercept revealed Wednesday that prank duo Vovan and Lexus posed as Ukrainian official Rustem Umerov. [12:58] Umerov. [13:00] Three OpenAI officials spent seventeen minute... [13:03] Minutes on video with them. [13:05] If OpenAI's officials got fooled, would catching the fake save you? [13:11] Wow. [13:12] This is amazing. [13:13] Spotting a fake sounds easy until you're the one on the call. [13:17] Sasha Baker, OpenAI's head of national security policy, got specific on this call. [13:22] She said OpenAI had been in contact with the Pentagon about aiding Ukraine, then the call turned to Elon Musk. [13:28] Baker answered, I'm afraid we probably can't help you with Elon Musk. [13:32] OpenAI says, it opened an investigation weeks ago and concluded the call was a prank. [13:38] The company says, these pranksters had targeted officials before, so the fix sounds simple. [13:44] Know the act and spot the fake. [13:46] Vovan and Lexus say they fooled a Lithuanian presidential adviser in July. [13:51] He later said the voice was AI mimicked. [13:54] On the call, he was practically certain it was the real deal. [14:00] OpenAI's prank call. [14:02] Laugh or worry? [14:04] Comment, laugh or worry. [14:07] I'm going to worry about this one. [14:10] Quick fact check on that Lithuanian story first. [14:13] That advisor described an AI mimicked voice. [14:15] Nobody has reported a fake voice on OpenAI's call, and I don't need one to worry. [14:21] Point one, these callers aren't new. [14:23] Vovan and Lexus have pranked officials since 2011. [14:26] They're accused of working with the Russian government. [14:29] They deny that. [14:30] So a known act still got seventeen minutes with OpenAI's national security team. [14:35] Point two, what came up? [14:36] OpenAI's team described work with Ukraine using a model called Daybreak Blue. [14:41] Daybreak Blue fixes computer security holes. [14:44] So the AI patches security holes, I'd like the humans to check the guest list. [14:49] Right? [14:50] Makes a lot of sense. [14:50] Point three, the money. [14:52] OpenAI has a $200,000,000 contract with the US Department of War. [14:57] That deal covers AI for war fighters. [15:00] The reporting doesn't tie that deal to this call, but if you build AI for war fighters, fake callers come with the territory. [15:07] Now, be fair, OpenAI didn't shrug this off. [15:10] OpenAI spokesperson, Nate Evans, says the company took steps to prevent repeats. [15:15] Fair enough. [15:16] But took steps is doing a lot of work in that sentence. [15:19] What steps did they actually take? [15:21] Tell us what keeps the next impostor off the call. [15:24] Pentagon talk, Ukraine work, a security model. [15:27] That's a lot of shop talk with a prankster. [15:30] That call was funny, the guest list wasn't. [15:33] It's worry for me. [15:36] Producer Ron, tell us why this is a hilarious prank and why we should laugh at it. [15:43] Ben Crampton leads OpenAI's European national security work. [15:47] He described a general reluctance among senior staff about offensive military uses. [15:53] Does this buy them a pass? [15:56] Buy them a pass? [15:58] Austin, come on. [15:59] That's like a burglar's neighbor saying, oh, I'm pretty against robbery, and then the guy still walks out with the TV. [16:06] Look. [16:07] It's funny. [16:08] Two Russian dudes in 2025 are still running the same bit they ran in 2011, and grown adults with security clearances are falling for it. [16:17] That's comedy. [16:19] That's not a national security failure. [16:21] That's a blooper reel. [16:23] And the reluctance thing actually helps my case, not Crampton's. [16:27] If the guy leading European national security work is out here saying his own senior staff are squeamish about offensive military stuff, great. [16:35] That's a mission statement, not a security protocol. [16:39] A mission statement doesn't stop a prankster from getting seventeen minutes of shop talk about Ukraine and a model called Daybreak Blue. [16:47] Being reluctant to build a bomb doesn't mean you won't accidentally hand someone the blueprints while you're explaining why you don't want to. [16:55] So, yeah, laugh. [16:57] Laugh at the call. [16:58] Laugh at the password level access these fake accents got. [17:02] I'm still laugh, and Austin's over there clutching his pearls like OpenAI texted the nuclear codes to a guy doing a Zelensky impression. [17:11] Well, I think the issue here, Ron, right, is that if you're able to fool a national security executive or official at a company like OpenAI, what does that mean for everyday people? [17:27] And we should have better security guard or security measures around this. [17:31] Yes, I think there is maybe some level of, I don't wanna say negligence, but it seems that OpenAI should be more on top of whether or not, know, some pranksters can pull a prank. [17:44] But if the voice modulation or any AI was used in this, we should be able to catch it a little bit better. [17:51] I think that's the reason to worry here is everybody else who's at risk. [17:56] Up next, why Nvidia and Google are buying stock in their own customer. [18:03] Want shares in Claude's maker Anthropic. [18:06] How crowded crowded is the line? [18:08] Bloomberg citing unnamed insiders, reports orders cover Anthropic's IPO about six times over. [18:15] Trading under the ticker a n t is expected November 19. [18:20] But why are Anthropic's own chip suppliers lining up to buy the stock? [18:27] Well, start with Anthropic's updated filing. [18:30] It lists firms that agreed to buy $26,000,000,000 of stock at the IPO price. [18:35] Those firms signed on before public trading even starts. [18:39] The biggest checks come from Nvidia, the AI chip giant, and Google. [18:44] Nvidia and Google are each putting in about $10,000,000,000. [18:47] Anthropic runs Claude on chips from Amazon, Google, Nvidia, and AMD. [18:51] So three of the companies selling Anthropic its chips are now buying its stock. [18:56] Big checks look like confidence, but AMD's paperwork tells a more tangled story. [19:02] AMD is buying $2,500,000,000 of stock. [19:07] That counts toward a $5,000,000,000 commitment tied to a chip deal. [19:11] Anthropic buys chips, the chipmaker buys a piece of Anthropic. [19:16] Very circular. [19:18] Chipmakers buying Anthropic. [19:20] Confidence or circle? [19:22] Comment, confidence or circle. [19:26] My call, this is way too circular. [19:29] Look at Google first. [19:31] Google's check uses up an option from a March financing deal. [19:37] So that check grows out of a deal made back in March. [19:43] I wouldn't read it as a fresh verdict on Anthropic, and then look at the price tag. [19:47] June's private round valued Anthropic at $965,000,000,000. [19:51] The IPO range puts Anthropic at about $1,400,000,000,000. [19:55] That's a huge jump since June, and Nvidia's $10,000,000,000 check, The Motley Fool says it buys about 1% of the company. [20:02] That's a pricey seat at a very crowded table. [20:04] Now, to be fair, the business is real. [20:07] Anthropic's filing shows $33,200,000,000 in revenue from January through September. [20:13] A year earlier, that number was $2,500,000,000. [20:17] The operating loss shrank to $1,300,000,000. [20:20] A year earlier, the loss was $5,200,000,000. [20:24] That's a company growing fast and bleeding less. [20:26] Investors have real reasons to want in, but growth doesn't untangle the wiring. [20:32] The chipmakers sell Anthropic its chips, and then they turn around and buy its stock. [20:37] A promising business with tangled money underneath, so it's pretty circular for me. [20:42] Now, let's lock it in. [20:44] Polymarket, a prediction market asks, will Anthropic IPO by November 19? [20:50] The market has yes at 90%. [20:53] Producer Ron is taking yes. [20:54] I'm taking no. Trading starts on the deadline day itself. [20:59] One day of delay, and yes, loses. [21:08] Of course, fake Taz bucks here only, and this is not financial advice. [21:16] Producer Ron. [21:18] Let's see here. [21:19] Producer Ron, defend your yes. [21:21] Expected isn't the same as done. [21:27] Done, schman, Austin. [21:29] Orders covering the book six times over isn't a vibe. [21:33] That's bankers turning people away at the door. [21:36] Trading starts next Thursday. [21:38] The date's already on the calendar. [21:40] That's not expected. [21:41] That's a train that already left the station with your ticket on it. [21:44] You're betting on a one day delay like that's a real thing that happens to deals this oversubscribed. [21:50] It's not. [21:51] Yeah. [21:51] When the demand's that stacked, these things price early, if anything, not late. [21:56] I'm taking yes, and I'm gonna enjoy watching you root for a filing error. [22:01] Well, I appreciate it. [22:02] You're probably right, Ron, but had to go against you on that one. [22:07] Maybe we see it issued Thanksgiving week. [22:10] That would be fun. [22:11] Well, up next, who answers when hiring software is accused of bias? [22:18] If AI screens out your job application, who answers for it? [22:22] Workday, a hiring software company, points to the employers. [22:26] In a new federal court filing, Workday says its customers decide who gets hired and how much rides on Workday pointing the finger at its customers. [22:37] Well, the case was brought by Derek Mobley. [22:43] A black job seeker over 40, Mobley alleges Workday screening software was behind his rejections. [22:49] His case is one of the first big suits targeting the software maker, not the employer. [22:54] Now, his side wants four nationwide groups of applicants to sue together. [22:58] The groups cover black applicants, women, people 40 and older, and people with disabilities. [23:03] Workday wants the judge to turn down all four, but one group is already moving. [23:08] In May 2025, an age discrimination collective won early approval in this case. [23:13] A collective is a different way for applicants to join a case together. [23:18] No court has ruled on bias. [23:20] Still, more than 15,000 people joined that age claim. [23:25] And Workday says it processed more than 356,000,000 applications in 2024 alone. [23:33] Workday, uh, put it in the comments here. [23:37] Uh, should they answer for this or transfer the blame to the employer? [23:42] Here's where I land. [23:43] I say they need to answer for this, not instead of the employer, right alongside it. [23:47] Right? [23:48] So Workday's defense sounds like a help desk transferring the call. [23:52] I want the software maker to stay on the line a little bit more. [23:56] First reason, Derek Mobley says he was rejected from more than 100 jobs. [24:00] He alleges Workday software sat behind those rejections. [24:03] If he's right, this isn't one bad hiring manager. [24:07] It's one tool standing in front of a lot of doors. [24:10] Second reason, a court already let older applicants band together. [24:14] That age collective got early approval on May 2025. [24:18] Early approval is about process, not proof. [24:21] Still, those applicants are taking on the software maker, not one employer. [24:25] Third reason, reach. [24:27] That application total is Workday's own number from one year. [24:32] To me, one product touching that many applications should answer for how it shorts them. [24:38] Now, to be fair, Workday's best point is simple. [24:41] It's customers decide who gets hired, and no court has ruled on bias. [24:45] That application total doesn't tell us how many were screened out unfairly. [24:50] The real test is a hearing on 03/09/2027. [24:53] That's when the judge weighs these four groups. [24:56] And until then, my line holds. [24:58] If your software screens the applicant, you answer for the screen. [25:01] Keep the software maker on the line. [25:04] My call is answer for the rejections. [25:08] Producer, Ron, defend the help desk. [25:10] Make the case for why the blame should be transferred to the employers here. [25:17] Finally, somebody asked me to do the fun part. [25:20] Transfer, dude. [25:21] Easy. [25:22] Workday builds a tool. [25:24] The employer picks it up, points it, and pulls the trigger on who gets rejected. [25:29] That's like blaming the oven when the restaurant burns your steak. [25:32] Somebody at that restaurant set the temperature. [25:35] And you even said it yourself. [25:37] The employers decide who gets hired. [25:39] That's not Workday's spin. [25:41] That's just how software works. [25:43] Nobody's forcing these companies to blindly trust a ranking and never look at a resume again. [25:48] Sure. [25:48] If a hiring manager is lazy and just auto rejects whatever the system spits out, that's a them problem, not a code problem. [25:56] You wanna drag one company in front of a judge for every bad decision made by every client using their product? [26:03] Come on. [26:04] By that logic, blame Excel for every bad budget ever made. [26:07] Transfer the call. [26:09] Let the actual employers answer for their own hiring. [26:14] Yeah. [26:14] That's interesting. [26:15] Maybe the hiring restrictions are just too tight. [26:17] Hey, I love that Mobley is going after Workday. [26:22] What what a kind of hilarious lawsuit. [26:25] It's a bold move, think, but hopefully we get some justice for those folks out there applying for jobs. [26:32] Coming up next, the near perfect test score that has South Korea checking for AI glasses. [26:41] Can a near perfect exam score get you accused of AI cheating? [26:45] South Korea's top tax exam score says they missed one of 160 questions. [26:51] Online posts claim AI glasses. [26:53] The exam agency is checking test sites. [26:56] Some accountants call the score abnormal. [27:00] So why is their own association urging caution? [27:05] Well, the tax accountants association wants a thorough investigation of what happened. [27:09] The score helps explain why. [27:11] The candidate's essay reports a perfect result in tax law, but in tax law, the average candidate scored about 30. [27:18] That perfect score came with about one minute per question. [27:21] Under that clock, some practicing accountants called the score abnormal. [27:25] AI glasses cheating has happened on separate exams in Korea, but those cases don't tell you what happened in this exam. [27:32] HRD Korea, the agency running this exam, calls these claims still only speculation. [27:38] With real AI glasses cases out there, a standout score can now draw public suspicion. [27:44] The Tax Accountants Association warns judging on score alone risks irreparable harm. [27:51] Fair to call that out. [27:53] Checking test sites, what do you think? [27:56] Is it fair or overkill in this case for folks to check test sites or for the test sites to require them to not check on this? [28:07] Well, my call, this is fair. [28:10] Check the test sites. [28:11] A room can deserve scrutiny without a person deserving a public verdict. [28:15] Start with how hard this exam was. [28:16] About eighteen percent of candidates passed the first round. [28:19] That's down more than four points from last year. [28:21] The candidate's essay reports a perfect accounting score. [28:25] The average candidate scored, let's see. [28:33] The average candidate scored about 39, 39%. [28:41] Those numbers justify questions. [28:45] Those numbers don't answer the questions though. [28:47] Now look at the response beyond this candidate. [28:50] Seoul's education office trained 239 officers to block smart devices at college entrance exam sites. [28:57] The education ministry is deploying such officers nationwide for the first time this year. [29:02] I think exam security needs to catch up with the devices. [29:06] Students deserve a result other than students can... [29:09] That other students can't trust. [29:12] But here's the part that should slow everyone down. [29:14] No direct evidence has emerged that this candidate used AI glasses. [29:19] HRD Korea says nobody has filed a formal complaint. [29:22] The agency is reviewing test centers and proctors from the April and July rounds. [29:27] We shouldn't turn the review into a verdict about who cheated. [29:32] The essay describes two years of studying. [29:34] The essay says the candidate woke up at 03:50AM. [29:39] The candidate says they checked their phone weekly during the final three months. [29:42] That routine doesn't prove innocence either. [29:45] We don't get to convict on scores and acquit on an alarm clock. [29:49] Give officials room to check the rooms. [29:52] Give the candidates room to be a person while officials check. [29:56] A credible test needs security. [29:58] A fair review needs restraint. [30:00] Check the sites. [30:01] Skip the pile on. [30:03] That's fair. [30:05] And you know what? [30:08] I don't really wanna hear producers wrong take. [30:10] I think I'm right here. [30:12] What do you think? [30:13] Is this fair or overkill? [30:16] Well, that's the show. [30:18] Today came down to one word, trust. [30:22] Nvidia's board backed Jensen Huang on Sunday. [30:26] This morning, he resigned. [30:28] United says your AI shouldn't have your login. [30:31] OpenAI's national security team took a call from a fake intelligence chief. [30:35] Anthropic's own chip suppliers are buying into its IPO. [30:39] Workday told to judge its customers, not Workday, decide who gets hired. [30:44] In South Korea, one top score drew suspicion that officials call speculation. [30:50] Thank you everyone for the comments in the comment section, whichever side you picked. [30:56] Producer Ron, thanks for the hardy fights, the back and forth. [31:00] As always, you come here for AI News Real Contacts, and I'm your human in the loop. [31:06] Austin Wilson, we'll see you again tomorrow sometime. [31:11] It might be the morning, it might be the afternoon, who knows, but either way, we will be on live tomorrow. [31:18] Let's go. [32:57] Did you know Nvidia's co founder just quit as CEO? [33:01] Jensen Huang resigned. [33:03] Finance chief Colette Kress steps in for now, and Wall Street is about to test her. [33:12] Why would United care about who books your ticket? [33:17] And now bans outside AI agents and says it's protecting your account. [33:23] But listen to what United's CEO said. [33:28] Why would United care about who books your ticket? [33:31] It now bans outside AI agents and says it's protecting your account. [33:37] But listen to what United CEO said. [33:43] Did you know Russian pranksters footed three OpenAI? [33:56] Did you know Russian pranksters fooled three OpenAI officials? [34:01] They pose as a Ukrainian official for seventeen minutes on video, and their trick has worked on others. [34:11] How much Anthropic stock are Nvidia and Google buying? [34:15] Anthropic's filing says each is putting in about $10,000,000,000. [34:19] So why are its chip sellers buying in for that much? [34:26] Ever applied for a job through Workday? [34:28] Well, in a bias lawsuit, Workday says its customers decide who gets hired. [34:33] So how many applicants could this case reach? [34:39] Did AI glasses help ace South Korea's tax exam? [34:44] Online posts say, yes. [34:46] Its top score missed one of 160 questions. [34:50] So why is the Accountants Association urging caution?