Cook Retiring, Ternus Will Helm Apple; Anthropic Gets Another $5 Billion from Amazon; CA Accuses Amazon of Price Fixing; OpenAI Releases Codex for Macs
Posted: April 21, 2026 Filed under: Uncategorized | Tags: AI, Amazon, anthropic, Apple, Artificial Intelligence, chatgpt, llm, technology Leave a commentBig Apple news dropped yesterday afternoon…Tim Cook is retiring as CEO, and Senior VP of hardware engineering John Ternus will take over that big chair. Engadget.com reports that this will happen September 1st. Cook will segue into an Executive Chairman role, so won’t completely disappear from the Apple universe. This is notable in one way as Ternus is 50…the same age as Cook when he took the reins in Cupertino, AND…Ternus worked for Steve Jobs at Apple. He will now be the last CEO of Apple who has had a direct connection to Steve Jobs. Ternus has kept a low profile, although he was featured in the rollout of the new MacBook Neo. He is credited with turning around the Macs and also had a big hand in Apple’s AirPods and Watch.
Anthropic has announced that Amazon will pump another $5 billion into the company, bringing their total investment to $13 billion. There could be another $20 billion later, depending on benchmarks. According to tehcrunch.com, Anthropic has agreed to spend over $100 billion on Amazon Web Services over then next 10 years, getting them a new 5 gigawatts of computing capacity to train and run Claude. The deal specifically covers Amazon Trainium2 through Trainium4 chips, too…even though the 4 chip isn’t yet available. With ChatGPT now worth some $730 billion, venture capitalists have been offering Anthropic additional capital that would bring Anthropic’s valuation to $800 billion or more!
In other Amazon news, California is accusing Amazon of price fixing. Gizmodo.com says California Attorney General Rob Bonta is accusing the online giant of pressuring brands to increase prices for their products on other retailers’ websites so that Amazon would have a more competitive price. The allegations, which were made in a filing that is part of California’s ongoing antitrust lawsuit against Amazon and was unsealed on Monday, lay out a scheme in which Amazon used the leverage of its massive e-commerce platform to pressure companies into raising prices with other retailers or face punishment for failing to do so. According to the attorney general, Amazon demands a vendor “fix,” “correct,” “increase,” “raise,” or “look into” the prices of products on other retailers’ websites. The expectation is that the vendor will ultimately raise its prices everywhere but Amazon. To get that outcome, Amazon would allegedly threaten to punish the brand by restricting their advertising, demanding they pay compensation, or removing their products from Amazon altogether.
OpenAI has released Codex Chronicle for Macs. 9to5mac.com notes that this is something of a ‘super app.’ Right now, it is especially made for agentic coding. ChatGPT remains the more general AI chatbot app. The idea of Chronicle is to make Codex more aware of context without repeating details or being super specific with each prompt. It builds on memory, and that lets Codex learn from conversation history for context. It also can learn from recent screen context. In terms of privacy, Chronicle can be paused or disabled at any time from Codex’s menu bar app. However, OpenAI warns that Chronicle consumes rate limits quickly based on its current design.
I’m Clark Reid and you’re ‘Technified’ for now.
Anthropic Briefed Administration on Mythos Model; FCC Exempts Netgear from Foreign Router Ban; Google Launches Native Gemini App for Mac; Deepfake Nude Crisis in Schools-Worse Than Many Realize
Posted: April 15, 2026 Filed under: Uncategorized | Tags: AI, Artificial Intelligence, chatgpt, deepfakes, FCC, gemini, llm, mythos, Netgear, technology Leave a commentThe co-founder of Anthropic has confirmed that the firm briefed the Trump administration on their Mythos AI model. Techcrunch.com notes that the new model will not be released to the public, due to it being ‘too dangerous’ and having powerful cybersecurity capabilities. Jack Clark, the co-founder of Anthropic said in an interview “Our position is the government has to know about this stuff, and we have to find new ways for the government to partner with a private sector that is making things that are truly revolutionizing the economy, but are going to have aspects to them which hit National Security, equities, and other ones…so absolutely, we talked to them about Mythos, and we’ll talk to them about the next models as well.” This comes after reports last week that Trump officials were encouraging banks to test Mythos, including JPMorgan Chase, Goldman Sachs, Citigroup, Bank of America, and Morgan Stanley. Let’s hope this claimed super Mythos model helps with cybersecurity and doesn’t make it easier for bad buys to crack security!
The Federal Communications Commission has just exempted Netgear from its ban on foreign routers…but strangely, gives no reason for the move. According to arstechnica.com, when the FCC banned foreign-made routers some 3 weeks ago, they said there could be exemptions in cases where the Department of Defense or Department of Homeland Security determines that the router does not pose national security risks. Router makers seeking conditional approvals must submit a justification for the use of foreign manufacturing and a “detailed, time-bound plan to establish or expand manufacturing in the United States.” The FCC and Netgear didn’t say what kind of justification or plan was submitted by Netgear.
Google has released a native Mac app for its Gemini AI…which up to now had just been on Android and iOS. 9to5google.com says This “native desktop experience” is launched via an Option + Space keyboard shortcut. There’s also the Gemini spark in the Menu Bar. You get a pill-shaped “Ask Gemini” bar with Liquid Glass. The “Add files and tools” plus at the left lets you upload Files, Drive, Photos, and NotebookLM, as well as Share window. Below, you will find Tools like Create image, Create video, Create music, Canvas, Deep research, Guided learning, and Personal intellegence. The window is identical to the one at gemini.google.com. You can share anything on your screen with Gemini, including local files. You can download it from theApp store.
The use of deepfakes in schools is apparently worse than most know, despite laws that have been passed banning or restricting the tech. Apparently, teen boys are working industriously…downloading pics of girls they know from school from Instagram and Snapchat, then running them through ‘nudify’ apps that created fake nude photos or videos of them. Wired.com notes that the problem really started showing up in schools a couple years ago, but now have hit at least 90 schools worldwide and some 600 students…that’s according to a review of public incidents by Indicator. Generative AI has really opened the door to more of this. There are 30 reports from North America since 2023…and you can bet that incidents are drastically underreported. In a number of cases, things have been handled by schools, and haven’t hit the press. Deepfakes have been around since 2017, but have exponentially increased.
I’m Clark Reid and you’re ‘Technified’ for now.
Amazon Merges with Sat Provider Globalstar; Google Chrome AI ‘Skills’; OpenAI Buys Hiro-AI Personal Finance Startup; Lucid Gets New CEO & Cash Infusion
Posted: April 14, 2026 Filed under: Uncategorized | Tags: AI, Amazon, Apple, Artificial Intelligence, chatgpt, gemini, Google, llm, technology Leave a commentAmazon is merging with satellite internet provider Globalstar. Engadget.com reports that this move should at some muscle to Amazon’s rival to Starlink, Leo. Globalstar is the satellite company behind Apple’s emergency SOS feature on iPhones and Apple Watches. Interestingly, Apple already owns 20% of Globalstar. Amazon and Apple have agreed Leo will “power satellite services for supported iPhone and Apple Watch models.” And that this support will continue as Leo’s network evolves, as well as collaborating “with Apple on future satellite services using Amazon Leo’s expanded satellite network.” Leo’s own direct to device service won’t start until 2028…the deal actually closes next year.
Google Chrome has a new generative AI feature, and this is pretty cool. It’s called Skills. I know, really original name. At any rate, according to wired.com, Skills are repeatable AI prompts you can run in Chrome with a keyboard shortcut. That sort of feature ought to attract a lot more people to AI who don’t want to learn or mess with trying to write and refine prompts. You can set up your own Skill using Gemini, Google’s chatbot, through the Chrome browser, or you can choose from the premade Skills Google released alongside this feature. The more than 50 presets in the Skills library cover a range of prompts that instruct Gemini to summarize YouTube videos, maximize your protein intake via recipe substitutions, or evaluate job listings. If you want to try out Skills, open up the Gemini in Chrome sidebar by clicking on the “Ask Gemini” sparkle icon in the upper-right corner of the screen. Then, type a forward slash in the prompt box to pick which Skill you would like to run. Have fun!
OpenAI has bought an AI personal finance startup called Hiro. Thenextweb.com notes that all 10 of the startup’s staff will join OpenAI. Hiro has built an app that offers AI-powered financial planning for consumers: users entered information about their salary, debts, and monthly costs, and the platform modeled different what-if scenarios to support financial decision-making.
Silvio Napoli is the new CEO of Lucid Motors, the EV maker, after a year long search. TechCrunch.com reports that Napoli has mainly managed at Schindler Group, which makes elevators and escalators. Hey…they’re electric! Napoli will join the board. In another related story, Lucid has gotten another $200 million cash infusion from Uber, which will buy up to 25,000 of Lucid’s upcoming mid-sized vehicles to use as robotaxis. The majority owner of Lucid, Saudi Arabian Public Investment Fund, has also kicked in by buying another $550 million of the company’s shares. Lucid’s upcoming mid sized vehicle will sell in the $50,000 range, a much larger pool of customers than the $100,000 or really $150,000 plus subset of buyers.
I’m Clark Reid and you’re ‘Technified’ for now!
Apple is 50; SpaceX Files for IPO; 100 Baidu Robotaxis Froze in Traffic; Claude Code Leak Was an Accident, Not a Hack
Posted: April 1, 2026 Filed under: Uncategorized | Tags: AI, anthropic, Apple, Artificial Intelligence, chatgpt, claude, llm, SpaceX, technology Leave a commentIt’s not an April Fool’s Day joke. Apple Computer got its start 50 years ago. They have been able to come up with a couple of truly world-changing devices.,.the iPod and the iPhone. Will they eventually come out with a third device that can have that kind of impact like some smart glasses? Time will tell. Meanwhile, mashable.com notes that there is a really cool graphic sequence on Apple’s home page celebrating the company’s devices using colorful brush strokes. It’s a bit Google-like, frankly. Imitation as the fabled sincerest form of flattery.
SpaceX has filed for an IPO…initial public offering of its stock. Engadget.com reports that while this was expected, most saw it happening in July. The Musk-owned company is looking for an IPO valuation of $1.75 trillion, which would make it the biggest IPO in history. SpaceX is the parent of X (formerly Twitter) and Grok, as well as xAI. The company is wanting to get its Starship rocket program on track, and has aspirations to build a base on the moon…and of course, Mars one day. They also plan for data centers for AI in space, orbiting the planet, as is in the works at several other tech companies.
We have had a few instances of some robotaxis stalling out and jamming up traffic…notably in San Francisco. Now, according to thenextweb.com, the US robotaxi makers like Google’s Waymo have been one-upped big time…and not in a good way. Over 100 Baidu Apollo Go robotaxis froze mid-traffic in Wuhan…blocking many hundreds of commuters. There were some crashes, although police say there were no injuries. This is really scary when you know that Wuhan has over 1,000 driverless vehicles rolling around the city. It’s an embarrassment for Baidu, which has vehicles in 26 cities globally, and claims to have orders for some 20 million vehicles. Welcome to the future…where you can have 100 instantaneous, random traffic jams to ruin your commute.
Claude’s source code got into the wild, and it turns out that it wasn’t due to hackers. 9to5google.com says the code was mistakenly published by Anthropic in the middle of the night. That’s a hell of a mistake! Ahthropic has been aggressively promoting Claude as superior to ChatGPT, and touting tools to migrate your ChatGPT work over to Claude. So how did this massive screw up happen? Well, apparently at about 4 am Tuesday morning, Anthropic pushed out what was supposed to be a routine update to Claude. Apparently, included in that update was a source map file that led right to Claude’s source code. The debugging file contained 512,000 lines of proprietary TypeScript code, which was initially spotted and posted by someone on Twitter/X. It wasn’t long before that entire code package was downloaded and circulated to thousands, though this leak doesn’t seem to include Claude’s model data. Still, this interface code is a costly loss for the company. In other words, a pretty gigantic ‘Oops.’
I’m Clark Reid and you’re ‘Technified’ for now.
Apple Maps Get Ads; FCC Bans Non-US Routers; Google Android Auto Moving into Car Control; Claude Code Can Now Take Over Computer for Tasks
Posted: March 24, 2026 Filed under: Uncategorized | Tags: AI, Artificial Intelligence, chatgpt, llm, technology Leave a commentIn a move that will evoke no joy in any quarter…except Apple’s sales folks, bean counters, and advertisers…Apple Maps is getting ads. Techcrunch.com reports that it is starting to allow advertisers to target customers on Apple Maps, starting in the U.S. and Canada later this summer. The ads will be available to any size business that has a physical location and has already created a business listing on Apple Maps. Users, meanwhile, will see the ads appear next to relevant search results. Apple is folding this ad offering under Apple Business, which includes integrated email, calendar, and directory service. This is irritating, but really just Apple joining the party…Google has had ads for years. Theoretically, this could tack on billions to Apple’s bottom line.
The FCC is banning all new routers not made in the US…unless they have already received FCC authorization. According to gizmodo.com, it “does not prohibit the import, sale, or use of any existing device models the FCC previously authorized.” The rub of this is…right now it doesn’t appear that there are ANY routers made in the US! Why did the Federal Communications Commission do this anyway? It’s all due to a series of attacks and hacks. There is a loophole that will continue to allow foreign routers, though…there is always a loophole. If the product is on the so-called ‘Covered List,’ of products and services that “pose an unacceptable risk to the national security of the United States or the security and safety of United States persons.” Many foreign routers already have conditional approval. They are primarily made in Taiwan, Thailand, and Vietnam.
Google is rolling out a new version of Android Auto. Android Auto has been a good open-source infotainment system, but the new software will be more of a car control system. Theverge.com notes that as cars become more and more a rolling computer, a unified software system will be a plus for a lot of can makers…or so Google thinks. Google is promising faster over-the-air software updates, better voice assistants, and more proactive vehicle maintenance alerts. Non-driving functions like climate control, lighting, and seating adjustment would fall under Android’s control. And the system would move beyond basic infotainment to create a unified ecosystem for features like remote cabin conditioning, digital key management, and personalized driver profiles. For the car makers, the unified Android Auto means spending less on developing their own expensive software systems.
This sounds pretty Orwellian right out of the gate, but Anthropic is now getting into the game with AI agents that can take direct control of your computer desktop. Arstechnica.com reports that this is for both Claude Code and the more casual user-oriented Claude Cowork. The Agents can now “point, click, and navigate what’s on your screen” to “open files, use the browser, and run dev tools automatically” when necessary to complete tasks. The new feature is now available to Claude Pro and Max subscribers using MacOS in what Anthropic calls a “research preview.” That means the system “won’t always work perfectly” and will sometimes require a “second try” for complex tasks, Anthropic warns. Completing tasks via “computer use” also “takes much longer and is more error-prone” than performing the same task via Connectors, the company writes.
I’m Clark Reid and you’re ‘Technified’ for now.
Google Gemini Will Power Apple’s Siri; Amazon-97% of Its Devices Run Alexa+; Anthropic Bows Version of Coding AI for Regular People; Meta will Cut 10% of Reality Labs
Posted: January 13, 2026 Filed under: Uncategorized | Tags: AI, Artificial Intelligence, chatgpt, llm, technology Leave a commentWe have reported on this for months…Apple has been shopping around, but mainly talking to Google…as they plan to drop ChatGPT which hasn’t worked well to beef up the ailing Siri. Now, it’s official…Macrumors.com reports that Cupertino has signed a multi-year deal with Google which will have Gemini help power a more personalized version of Siri, and also support a range of Apple Intelligence features that are in the pipeline. Note that everything will still run on Apple’s Private Cloud servers or your local device, so you won’t have more of your info basically uploaded to the world. The upgraded Siri should be out in March or April…about a year later than Apple originally promised. I don’t have high expectations about a more personalized Siri…I’d just like it to actually work decently…with any luck, adding some Google AI brainpower will make Siri at least somewhat competent.
Backwards compatibility is a beautiful thing…it’s a big pain …to say nothing of expense…when tech companies put out something and it won’t run on any of your gear. According to techcrunch.com, Amazon says its Alexa+ AI platform will run on 97% of the devices they have ever shipped. That is a lot of devices…about 600 million! Alexa+ offers more expressive voices, access to world knowledge similar to other AI assistants, and AI agents that perform tasks on behalf of the customer — like calling an Uber or ordering food. The company has been steadily rolling out access to the AI platform, with more than 1 million Alexa customers gaining access by last June, and now, “tens of millions” can opt in to upgrade to the AI assistant. Amazon hasn’t given an exact timeline for when it will be open to everyone…right now they are focused on making it available to all Prime members.
Here’s a concept…a coding AI made for ‘regular people.’ That’s what Anthropic, maker of the Claude AI says they are putting out. Engadget.com says Anthropic is calling the feature Claude Cowork, and they claim it is “a simpler way for anyone — not just developers — to work with Claude.” After you give the system access to a folder on your computer, it can read, edit or create new files in that folder on your behalf. As examples of what it might do for you, they say you could ask Claude to organize your downloads folder, telling it to rename the files contained within to something that’s easier to parse at a glance. Another example: you could use Claude to turn screenshots of receipts and invoices into a spreadsheet for tracking expenses. Cowork can also navigate websites — provided you install Claude’s Chrome plugin. Anthropic touts the fact that “Cowork is designed to make using Claude for new work as simple as possible. You don’t need to keep manually providing context or converting Claude’s outputs into the right format,” the company said. “Nor do you have to wait for Claude to finish before offering further ideas or feedback: you can queue up tasks and let Claude work through them in parallel.”
Meta is going to whack its Reality Labs division, cutting about 10% of the some 15,000 staffers there. Geekwire.com notes that this is the division that Mark Zuckerberg dumped a fortune into in creating the Metaverse…a virtual, imaginary space where your avatar could live and work. That has been a big dud, but the Reality Labs folks do also work with the VR headsets, so there could be some effects felt there by the layoffs. Meta employs about 78,000 worldwide.
I’m Clark Reid and you’re ‘Technified’ for now.
Trump Says Nvidia Can Sell H200 Chips to China; Apple Silicon Chief Staying for Now; Claude Code Coming to Slack; Google Project Aura- Hopefully Not AI for Glassholes
Posted: December 9, 2025 Filed under: Uncategorized | Tags: AI, Artificial Intelligence, chatgpt, llm, technology Leave a commentDonald Trump has now said he will allow Nvidia to sell the H200 AI chip to China. Gizmodo.com reports that Nvidia will still be barred from selling its more advanced Blackwell chips to China, but it’s still considered a win for the tech company since the lower quality H200 chip had been sidelined by the Chinese government for supposedly not being powerful enough. There are still some conditions surrounding the ability to sell the chips, and in addition China has just said it may not allow purchase of them anyway…as it tries to get chipmakers there to make their own AI chips that are competitive with US silicon.
There has been an exodus from Apple unlike any in recent memory lately. A number of top players have jumped ship for higher paying jobs or what they see as opportunities to move ahead. One who isn’t leaving…for now…is Apple’s silicon chief Johny Srouji. He had been ‘seriously considering’ leaving, but has decided to stay. This is great news for Cupertino, after losing their AI chief, their environmental chief, and General Counsel…in addition to a top software designer.
Anthropic is launching Claude Code in Slack…allowing developers to delegate coding tasks directly from chat threads. Techcrunch.com says it is available now as a beta, and builds on Anthropic’s existing Slack integration by adding full workflow automation. This signals that AI coding assistants are moving out of integrated development environments and morphing into collaboration tools that will live where teams already work. For Slack’s part, it can position itself as an ‘agentic hub’ that could shape how software teams work.
Google just held a livestream this week on its Android YouTube channel, and they unveiled new Galaxy XR capabilities, AND teased Android XR smart glasses that should be available next year. Bgr.com reports that Project Aura will be glasses that Google didn’t preview last spring. We already knew they were working on some screen less glasses and some with a display. Project Aura is a different animal….they are wired XR glasses. The glasses connect to a smartphone like puck that you keep in your pocket or on a desk. The puck supports touch input, much like a mouse. The glasses will effectively let you run a virtual Android XR computer anywhere you are. You can see the world around you while operating a private computing experience. This is quite a volley across the bow of Apple, which has a great…but big and heavy VR headset. As with their other glasses, Google will be partnering with Warby Parker and Gentle Monster. The first AR glasses will launch in 2026…and Google says even the Project Aura glasses will be available next year!
I’m Clark Reid and you’re ‘Technified’ for now.
Google Bows Gemini 3 AI & Antigravity Agent; iPhone 17 Powers Apple to 25% of China’s Phone Sales; Cloudflare-Latent Bug Caused Huge Outage; Anthropic Now Valued at $350 Billion
Posted: November 18, 2025 Filed under: Uncategorized | Tags: AI, Artificial Intelligence, chatgpt, llm, technology Leave a commentGoogle has released its Gemini 3 AI, and a new agent called Antigravity. Arstechnica.com reports that the release of Gemini 3 Pro is just available in a limited form today, though. Google is touting more immersive, visual outputs and fewer lies, Google says. The company also says Gemini 3 sets a new high-water mark for vibe coding. In addition to Gemini 3, Google has also unveiled a new AI-first integrated development environment…or IDE…named Antigravity. Google says 3 has improved simulated reasoning abilities, and better understanding of text, images, and video. Google says math and coding are the primary focus of this release. Google says people should think of Antigravity as a “mission control” for creating and monitoring multiple development agents. The AI in Antigravity can operate autonomously across the editor, terminal, and browser to create and modify projects, but everything they do is relayed to the user in the form of “Artifacts.”
The lower priced iPhone 17 has helped Apple to pick up 25% of China’s smartphone sales. According to bgr.com, the base model 17 made up 80% of Apple’s iPhone sales in China. Apple sold 37% more iPhones than last year in China. The top line iPhone 17 Pro Max is experiencing shipping delays in China, but just 2-5 days. The premium phone segment is still hotly contested…The Huawei Mate 80 flagship phone is coming out, and that is expected to cut into Apple’s market share.
Cloudflare had a major outage this Tuesday morning that took down or snarled a significant piece of the internet. TechCrunch.com says the outage hit ChatGPT, Claude, Spotify, X, and others. In a post on X, Cloudflare’s chief technology officer Dane Knecht said a latent bug was responsible, and he apologized for the outage. The problem occurred when Cloudflare’s bot mitigation system started to crash after a routine configuration change. A few users were still having some issues logging onto their Cloudflare dashboards by late morning.
As part of new strategic partnerships with both Nvidia and Anthropic, Microsoft will pour up to $5 billion into Anthropic, while Nvidia will dump $10 billion into the AI startup. Cnbc.com notes that this will put the valuation of Anthropic…maker of the Claude AI model…at around $350 billion, up from $183 billion in September. Anthropic has committed to purchasing $30 billion of Azure compute capacity from Microsoft and has contracted for additional compute capacity up to 1 gigawatt, according to a blog post. Anthropic has also committed to purchase up to 1 gigawatt of compute capacity with Nvidia’s Grace Blackwell and Vera Rubin systems. Amazon Web Services continue to be Anthropic’s primary cloud provider, however.
I’m Clark Reid and you’re ‘Technified’ for now.
Snap-Light Consumer AR Glasses in ’26; Threads is Getting DMs; EVs with 3,000 Mile Range on the Horizon; Apple White Paper- Power of New Generation AI Wildly Oversold
Posted: June 10, 2025 Filed under: Uncategorized | Tags: AI, Artificial Intelligence, chatgpt, ev-range, llm, technology, threads Leave a commentSnap is preparing to sell lightweight, consumer AR glasses in 2026. That’s the word from techcrunch.com. They will be called Specs. Snap’s Specs will feature many of the same augmented reality and artificial intelligence capabilities that are available on the company’s developer-facing smart glasses, the Spectacles 5. However, the company says the Specs will be smaller and lighter — ideally making them more innocuous to wear in public than their extremely large predecessors. So far, no word from Snap on pricing, or on how they plan to sell the glasses. If they are lightweight, work well, and don’t look goofy…Snap will have really pulled off something great in the smart glasses race.
Meta CEO Mark Zuckerberg has announced that Threads will start testing its own direct messaging this week…which you will be able to use without leaving the platform. According to theverge.com, the testing will start in Hong Kong, Thailand, and Argentina, then expand to other nations. Users will see a separate inbox for Threads DMs, without having to connect to their linked Instagram account. At this point you will still have to have an Instagram account to use Threads. Maybe one day, they will split the baby, so to speak.
In a huge leap forward, a couple of researchers in South Korea have come up with a way to reduce silicon swelling in traditional EV battery designs. Bgr.com says that the better tech may take us from the 200-300 miles per charge to somewhere in the area of 3,000 mile range! That’s not all…using graphene in the batteries, they have shown that you can fully charge in something like 75 seconds…with no degradation in capacity for over 1,000 recharges. If this tech scales up, we may see truly revolutionary range in electric vehicles…and smartphones you won’t ever stress about running down the battery and being out of contact.
Apple has put out a research paper that some are nodding in the affirmative over…like myself, while others are stunned. Theguardian.com notes that the paper all but eviscerating the popular notion that large language models (LLMs, and their newest variant, LRMs, large reasoning models) are able to reason reliably. well-known venture capitalist Josh Wolfe went so far as to post on X that “Apple [had] just GaryMarcus’d LLM reasoning ability” – coining a new verb (and a compliment to me), referring to “the act of critically exposing or debunking the overhyped capabilities of artificial intelligence … by highlighting their limitations in reasoning, understanding, or general intelligence.” What Apple did was show that the leading models like ChatGPT, Claude, and Deepseek may “look smart – but when complexity rises, they collapse”. In short, these models are very good at a kind of pattern recognition, but often fail when they encounter novelty that forces them beyond the limits of their training, despite being, as the paper notes, “explicitly designed for reasoning tasks.” The Cliff’s Notes takeaway for you…relax…Artificial Intelligence and robots won’t be taking over…at least not yet.
I’m Clark Reid and you’re ‘Technified’ for now.
Recent Comments