WEBVTT 00:00:00.000 --> 00:00:04.560 Wow, tech is still making the news, is it? Yeah, what this show is. 00:00:04.560 --> 00:00:09.200 I mean, I would have thought they'd have gone through it all by now. So, this must be the last of it here. 00:00:09.200 --> 00:00:16.720 What are you doing? A gaming handheld with a folding screen may be shown off by Lenovo at MWC Barcelona 00:00:16.720 --> 00:00:22.640 next week, according to Windows Latest, who posted a bunch of apparently leaked images. 00:00:22.640 --> 00:00:29.920 The Legion Go Fold yourself has a 7.7-inch display already bigger than the Steam Deck OLED, 00:00:30.000 --> 00:00:33.440 that can fold out to 11.6 inches. 00:00:36.000 --> 00:00:39.760 It also has detachable controllers that work in either mode 00:00:39.760 --> 00:00:43.360 and can attach to a wireless keyboard and touchpad polio. 00:00:43.360 --> 00:00:47.120 What can't it do? I mean, game comfortably? 00:00:47.120 --> 00:00:54.080 The right controller doubles as a vertical mouse, and appears to include a tiny round display showing a watch face 00:00:54.080 --> 00:00:57.520 in the same place where the Legion Go 1 and 2 have a trackpad. 00:00:57.520 --> 00:01:04.320 Even more strange are its specs. It's reportedly running an Intel Core Ultra 7 258V from 2024, 00:01:04.320 --> 00:01:09.360 rather than a newer Panther Lake chip, along with a 48Wh battery and 32GB of RAM. 00:01:09.360 --> 00:01:15.040 Meaning that by the time this thing hits store shelves, it's going to be slow, expensive, and have kind of crappy battery life. 00:01:15.040 --> 00:01:19.600 You could buy it as an investment. And that's if it ever hits store shelves. 00:01:19.600 --> 00:01:24.160 Windows Latest says that the Legion Go Fold will be introduced as a concept of a device. 00:01:24.160 --> 00:01:29.440 But Lenovo has actually launched a few of their wackier concepts as real products in the past. 00:01:29.440 --> 00:01:34.080 In fact, I showed a couple of them to Jimmy Fallon. I appreciate Lenovo letting this one leak ahead of time. 00:01:34.080 --> 00:01:39.680 So before I show it to Jimmy, I can get used to it and make sure it will do what it's supposed to do. 00:01:39.680 --> 00:01:46.240 Anthropic has officially rejected an ultimatum issued by the U.S. Department of War, 00:01:46.240 --> 00:01:52.240 demanding that Anthropic allow the military to use their AI model cloud without guardrails 00:01:52.240 --> 00:01:55.600 or face being deemed a supply chain risk. 00:01:55.600 --> 00:02:01.520 CEO Dario Amade wrote in a blog post that he believes deeply in the importance of using AI 00:02:01.520 --> 00:02:06.240 to defend the United States of America, but that in a narrow set of cases, 00:02:06.240 --> 00:02:10.080 AI can undermine, rather than defend, democratic values. 00:02:10.080 --> 00:02:15.120 Amade published this post on Thursday ahead of the deadline the Pentagon had set for Anthropic 00:02:15.120 --> 00:02:18.720 to agree to their terms. And despite Under Secretary of War, 00:02:18.720 --> 00:02:24.800 Emile Michael coming in hot with a very public crash out on Twitter, calling Amade a liar with 00:02:24.800 --> 00:02:30.560 a God complex who wants nothing more than to personally control the U.S. military, good lord. 00:02:30.560 --> 00:02:36.240 Amade stuck to his guns and late on Friday, Trump ordered all U.S. agencies to stop using 00:02:36.240 --> 00:02:40.880 Anthropic's tech. Ultimately though, Anthropic fighting to keep guardrails on cloud is good, 00:02:40.880 --> 00:02:44.400 because in the wrong hands, the model can be dangerous. 00:02:44.400 --> 00:02:49.040 As proven on Wednesday, when it was discovered that a hacker used it to breach multiple Mexican 00:02:49.040 --> 00:02:54.240 government agencies, stealing 195 million taxpayer records along with voter data and 00:02:54.240 --> 00:02:57.760 employee credentials. Oh, I see how disentropic. 00:02:57.760 --> 00:03:01.360 You don't want to give unrestricted cloud to your nation's protectors, 00:03:01.360 --> 00:03:03.360 but you will give it to Mexican hackers? 00:03:04.880 --> 00:03:10.640 Meanwhile, Meta's AI alignment director, Summer Yu, had nearly her entire email inbox 00:03:10.640 --> 00:03:14.240 deleted by the popular AI agent Open Claw this week. 00:03:14.240 --> 00:03:20.400 If you haven't seen the hype around it, Open Claw, a.k.a. Moltbot, a.k.a. Clawadbot, 00:03:20.400 --> 00:03:25.360 is an AI personal assistant that can work independently and will interface with users 00:03:25.360 --> 00:03:29.440 through common messaging apps like WhatsApp, performing tasks on their behalf. 00:03:29.440 --> 00:03:35.120 Yu told it to go through her inbox and suggest what to archive or delete, but not act without 00:03:35.120 --> 00:03:39.760 permission. The bot agreed and then immediately forgot Yu's instructions and went rogue, 00:03:39.760 --> 00:03:45.920 deeming all emails older than February 15th to be useless and therefore garbage, bulk 00:03:45.920 --> 00:03:53.360 deleting hundreds of emails in seconds. When she told it to stop, it ignored her instructions and she had to rush back to her 00:03:53.360 --> 00:03:57.360 computer and kill the process the bot was running on in order to get it to stop. 00:03:57.360 --> 00:04:02.320 Some commenter speculated that, given Yu is one of Meta's top safety researchers, 00:04:02.320 --> 00:04:06.160 the whole incident must have been a test that she was running to check the effectiveness 00:04:06.160 --> 00:04:12.720 of AI guardrails. But Yu admitted that no, it was in fact just a rookie mistake, saying that 00:04:12.720 --> 00:04:17.280 alignment researchers aren't immune to misalignment, even if that immunity can be 00:04:17.280 --> 00:04:22.400 fairly easily achieved by just not giving an infamously unreliable and insecure AI agent 00:04:22.400 --> 00:04:27.440 total control of your email. So AI safety might be in trouble over at Meta. 00:04:27.440 --> 00:04:30.960 But hey, at least on Instagram, teen safety might be improving. 00:04:30.960 --> 00:04:35.760 Meta announced parents using supervision mode for their kids' accounts will be alerted 00:04:35.760 --> 00:04:39.200 if their teen repeatedly runs searches involving terms related to self-harm. 00:04:39.200 --> 00:04:42.480 This is just rolling out though, so we can't say how effective it will be, 00:04:42.480 --> 00:04:46.320 but we can say that you should check out our sponsor. Okay, that's gotta be it. 00:04:46.320 --> 00:04:53.280 Surely there's no more tech news out there. Wait, there's still some small pieces that we can do quickly? 00:04:53.280 --> 00:04:59.040 You've done this before, I don't- This is crazy. NVIDIA is rolling out a new update for the SHIELD TV, 00:04:59.040 --> 00:05:04.320 surprisingly proving their commitment to continued support for the now 10-year-old hardware. 00:05:04.320 --> 00:05:09.600 The new update, version 9.2.4, focuses on stability and security patches, 00:05:09.600 --> 00:05:13.280 ensuring that the last great set-top box remains reliable. 00:05:13.280 --> 00:05:17.520 Unlike newer devices, it thankfully still runs the classic Android TV interface, 00:05:17.520 --> 00:05:22.160 rather than the cluttered Google TV shenanigans that are seen on a lot of smart TVs these days. 00:05:22.160 --> 00:05:25.200 What a breath of fresh air. Hey, our inbox is open. 00:05:25.200 --> 00:05:29.920 If any other companies want to email us and explain why all of their devices get ens**tified, 00:05:29.920 --> 00:05:35.120 when even one of the biggest contributors to our current AI dystopia can manage to do this. 00:05:35.120 --> 00:05:43.440 But not all is green flags for Team Green. NVIDIA has officially polled their latest gaming GPU driver, 595.59, 00:05:43.440 --> 00:05:49.280 after reports that it was causing critical fan issues across RTX 3040 and 50 series cards. 00:05:49.280 --> 00:05:52.720 So if your GPU has been acting like a single fan fossil, 00:05:52.720 --> 00:05:57.280 you might want to roll back to the previous stable driver 591.86. 00:05:57.280 --> 00:06:03.760 We've got to make sure we protect the cards we have, especially since NVIDIA CFO Collette Cress warned during the company's recent earnings call 00:06:03.760 --> 00:06:08.720 that RTX 50 supply will be very tight for at least another six months. 00:06:08.720 --> 00:06:12.240 So that means it'll be over in six months, right? 00:06:14.640 --> 00:06:19.680 Right? Block, the company that owns Square and Cash App, 00:06:19.680 --> 00:06:24.560 announced in a shareholder letter on Thursday that they're laying off nearly half their staff 00:06:24.560 --> 00:06:30.560 in what CEO and guy playing a bongo too loud in the park, Jack Dorsey called 00:06:30.560 --> 00:06:33.600 a deliberate and bold embrace of AI. 00:06:33.600 --> 00:06:38.960 You know, it's really nice for CEOs these days that they can use AI to pretend that 00:06:38.960 --> 00:06:46.720 mass layoffs is just them being ready for the future, rather than them trying to boost the perceived value of their company by lowering their expenses. 00:06:47.680 --> 00:06:53.040 But hey, I guess it worked. Block shares soared 24% after the announcement. 00:06:53.040 --> 00:06:56.480 Hooray, think of all the bongos you can buy now, Jack. 00:06:59.440 --> 00:07:04.640 A company called Techno is bringing modularity back at MWC this year with 00:07:04.640 --> 00:07:07.920 the world's thinnest modular smartphone ecosystem concept. 00:07:07.920 --> 00:07:11.520 The Moda edition sports a Greek inspired aesthetic, 00:07:11.520 --> 00:07:16.320 while the Atom offers a white and red look akin to Kratos from God of War. 00:07:16.320 --> 00:07:20.400 These devices use magnetic pins to swap cameras, batteries, and game pads. 00:07:20.400 --> 00:07:28.080 It's kind of like MagSafe, but on steroids. Finally giving us a way to lose our phones, camera, and batteries individually, instead of all at once. 00:07:30.240 --> 00:07:34.400 And Burger King is deploying Paddy. Oh, Lordy, I saw this. 00:07:34.400 --> 00:07:39.840 An AI system designed to monitor employees and ensure that they are remembering their 00:07:39.840 --> 00:07:47.520 P's and Q's when they interact with customers. Beyond policing manners, though, this digital supervisor apparently has sensors of some kind 00:07:47.520 --> 00:07:55.040 to detect when restrooms need a scrub. It's an ambitious and somewhat dystopian move towards taking over all the jobs with AI. 00:07:55.040 --> 00:08:00.480 But I mean, hey, maybe this was inevitable. This concept was shown to us years ago in SpongeBob, 00:08:00.480 --> 00:08:04.640 with Plankton's computer wife, Karen, just this time she's gone corporate. 00:08:05.280 --> 00:08:12.160 Straight to the top, baby! And I expect you to come straight back here on Monday to get more tech news. 00:08:12.160 --> 00:08:15.440 More tech news? Are you kidding me? There can't possibly be more tech news that- 00:08:16.000 --> 00:08:17.120 How are we ever gonna finish it?