AI with Kyle - Daily AI News and Updates

Get AI-Ready with Kyle’s 5-Day Challenge: https://aiwithkyle.com/join

Learn how to become an AI trainer - Live Webinar: https://aiwithkyle.com/ai-workshop-series/webinar

Kyle’s AI Canon — Curated list of AI learning essentials: https://aiwithkyle.com/ai-canon

Subscribe and turn on notifications to catch the next live stream: https://www.youtube.com/channel/UChlLglbHDASnoGkbjDeHnQg

This video is from the ‘AI with Kyle Live Stream’

First aired: 3 March 2026

Summary: Big week for AI drama. Anthropic found themselves declared a "supply chain risk" by the Pentagon after refusing to drop their red lines on autonomous weapons and domestic surveillance, and rather than rolling over, they've lawyered up and are challenging it in court. OpenAI, never ones to miss an opportunity, stepped in within days to take the contract, which went down about as well as you'd expect. ChatGPT uninstalls surged 300%, one-star reviews spiked 775% in a single day, and Claude hit number one in the US App Store. Suddenly everyone's heard of Anthropic.

The irony is that Anthropic's best ever week of growth came wrapped in an existential threat. All those new Claude users flooding in, including apparently Katie Perry, were met with overloaded servers and error messages, which is a painful way to make a first impression. Worth noting too: Anthropic aren't the pacifists some are making them out to be. They took a $200 million Pentagon contract, they're fine with autonomous weapons in principle, they just want guardrails. That nuance is getting lost in the noise, but it matters. Six months from now, whether Claude still exists in its current form is a genuinely open question.

—— Time Stamps ——

00:00 Anthropic vs Pentagon: The AI Cold War Begins

—— More Useful Resources ——

Find everything else at https://aiwithkyle.com/

Get AI-Ready with Kyle’s 5-Day Challenge: https://aiwithkyle.com/join

Learn how to become an AI trainer - Live Webinar: https://aiwithkyle.com/ai-workshop-series/webinar

Kyle’s AI Canon — Curated list of AI learning essentials: https://aiwithkyle.com/ai-canon

Subscribe and turn on notifications to catch the next live stream: https://www.youtube.com/channel/UChlLglbHDASnoGkbjDeHnQg

This video is from the ‘AI with Kyle Live Stream’

First aired: 2 March 2026

Summary: The same AI model you're using to write emails and debug code was reportedly used by the US military to select bombing targets in Iran this weekend. Not a drill, not a hypothetical. And the detail that should genuinely make you pause is this: the government doesn't have a secret, more powerful version of Claude sitting in a classified bunker somewhere. Ethan Mollick made the point well. They're using the same model you get for $20 a month, probably an older version, because new models have to go through procurement before deployment. That's the actual state of AI in modern warfare right now.

What that means in practice is worth sitting with. When Claude is a little too agreeable, a little too confident, prone to the occasional hallucination, that's fine when you're polishing a draft or writing a proposal. When it's informing decisions about military targets, the stakes are rather different. The AI arms race isn't some distant future concern. It's already here, already running, and it's built on the exact same infrastructure the rest of us use every day.

—— Time Stamps ——

0:00 Introduction: Trump Enters the Anthropic War

—— More Useful Resources ——

Find everything else at https://aiwithkyle.com/

Get AI-Ready with Kyle’s 5-Day Challenge: https://aiwithkyle.com/join

Learn how to become an AI trainer - Live Webinar: https://aiwithkyle.com/ai-workshop-series/webinar

Kyle’s AI Canon — Curated list of AI learning essentials: https://aiwithkyle.com/ai-canon

Subscribe and turn on notifications to catch the next live stream: https://www.youtube.com/channel/UChlLglbHDASnoGkbjDeHnQg

This video is from the ‘AI with Kyle Live Stream’

Summary: Today was one of those days where the AI news actually matters. Anthropic and the US Department of Defense have been circling each other all week, and this morning Dario Amodei finally put out a statement telling the Pentagon to do one. The two red lines — autonomous weapons targeting and mass domestic surveillance — remain firmly in place, despite threats to blacklist Anthropic as a supply chain risk, invoke the Defense Production Act, and effectively destroy the company. It is, without question, the most consequential standoff between a government and an AI company we have ever seen, and I am genuinely baffled more people are not talking about it.

On top of that, Jack Dorsey cut 40% of Block's workforce, around 4,000 people, and was unusually honest about why: the company is profitable, growing, and simply does not need as many humans anymore. AI did that. We also got Google's Imagen 3 update with much better consistency across frames, and Anthropic, despite potentially ceasing to exist by Monday, somehow found the time to ship auto memory, remote MCP connections, and scheduled tasks in Cowork this week. Busy Friday.

—— Time Stamps ——

0:00 Intro: Anthropic vs Pentagon, Jack Dorsey AI Cuts & New Claude Features

—— More Useful Resources ——

Find everything else at https://aiwithkyle.com/

Learn how to become an AI trainer - Live Webinar: https://aiwithkyle.com/ai-workshop-series/webinar

Get AI-Ready with Kyle’s 5-Day Challenge: https://aiwithkyle.com/join

Kyle’s AI Canon — Curated list of AI learning essentials: https://aiwithkyle.com/ai-canon

Subscribe and turn on notifications to catch the next live stream: https://www.youtube.com/channel/UChlLglbHDASnoGkbjDeHnQg

This video is from the ‘AI with Kyle Live Stream’

First aired: 26 February 2026

Summary: Anthropic are having an absolute nightmare of a week. Caught between Chinese distillation attacks at the start, and a Pentagon ultimatum threatening to blacklist them as a "supply chain risk" by Friday, they've quietly walked back their flagship responsible scaling policy, the very thing that set them apart as the supposed good guys of AI. Convenient timing, given the policy dropped the same day Pete Hegseth delivered his ultimatum. Make of that what you will.

On the more practical side, Claude Code is steadily rebuilding everything people loved about Open Claude, piece by piece, remote sessions, and now scheduled recurring tasks, so your AI can crack on while you sleep. Google's Anti-Gravity also deserves a mention: it's free, it runs Claude Sonnet, Opus, Gemini and GPT all in one place, and if you're already paying for Google Workspace, you've probably had access to half of this without knowing it. Classic Google.

—— Time Stamps ——

0:00 Anthropic vs The Pentagon: The $200M AI Weapons Showdown

—— More Useful Resources ——

Find everything else at https://aiwithkyle.com/

https://youtu.be/8TKAl2zPVyA

Learn how to become an AI trainer - Live Webinar: https://aiwithkyle.com/ai-workshop-series/webinar

Get AI-Ready with Kyle’s 5-Day Challenge: https://aiwithkyle.com/join

Kyle’s AI Canon — Curated list of AI learning essentials: https://aiwithkyle.com/ai-canon

Subscribe and turn on notifications to catch the next live stream: https://www.youtube.com/channel/UChlLglbHDASnoGkbjDeHnQg

This video is from the ‘AI with Kyle Live Stream’

First aired: 25 February 2026

Summary: Anthropic are having a rough week. The Pentagon have given them until Friday to strip Claude's safety guardrails for military use (think autonomous weapons and mass surveillance) or face being labelled a US supply chain risk, a designation normally reserved for Chinese companies. It's an extraordinary situation, and the irony isn't lost on me: Anthropic spent yesterday warning the world about Chinese labs stealing their models on national security grounds, and today they're being threatened by their own government. OpenAI, Google and xAI took the contracts and quietly complied. Anthropic are the only ones drawing a line.

Elsewhere, Claude Code quietly knocked 13% off IBM's stock price, its worst crash in 25 years, simply by refactoring legacy COBOL applications. If the code moves off COBOL, the mainframes that run it become redundant, and IBM's entire core business starts looking shaky. One blog post did that. And on the education front, UK students are now using AI for nearly half of all their study tasks, double the figure from last year and the highest of any country surveyed. Teachers increasingly can't tell the difference. The education system hasn't caught up and, frankly, it's not clear it knows how to.

—— Time Stamps ——

00:00 Anthropic vs The Pentagon: Friday Deadline Could End AI Giant

—— More Useful Resources ——

Find everything else at https://aiwithkyle.com/

https://youtu.be/8TKAl2zPVyA

Learn how to become an AI trainer - Live Webinar: https://aiwithkyle.com/ai-workshop-series/webinar

Get AI-Ready with Kyle’s 5-Day Challenge: https://aiwithkyle.com/join

Kyle’s AI Canon — Curated list of AI learning essentials: https://aiwithkyle.com/ai-canon

Subscribe and turn on notifications to catch the next live stream: https://www.youtube.com/channel/UChlLglbHDASnoGkbjDeHnQg

This video is from the ‘AI with Kyle Live Stream’

First aired: 24 February 2026

Summary: Today I got into Anthropic's claim that three Chinese AI labs, DeepSeek, Moonshot AI, and Minimax, ran what they're calling an "industrial scale distillation attack" on Claude, using 24,000 fake accounts to extract 16 million interactions and essentially clone its capabilities. It's a serious accusation, dressed up in national security language clearly aimed at Washington rather than the labs themselves. The delicious irony, of course, is that Anthropic built Claude by training on the entire internet's worth of copyrighted material without asking anyone. Thieves stealing from robbers, as I put it.

The second half got into something a bit more uncomfortable: AI and jobs. My honest take is that this technological shift is different from everything that came before it, because AI won't just automate tasks, it'll take the new jobs it creates too. Blue collar and physical work buys you some time, but if your job mostly lives behind a screen, the clock is ticking. My practical advice is simple: don't panic, don't quit, but quietly start building alternative income streams now, because nobody, and I mean nobody, actually knows what survives the next five years.

—— Time Stamps ——

0:00 Anthropic Accuses Chinese Labs of Stealing Claude

—— More Useful Resources ——

Find everything else at https://aiwithkyle.com/

https://youtu.be/8TKAl2zPVyA

Learn how to become an AI trainer - Live Webinar: https://aiwithkyle.com/ai-workshop-series/webinar

Get AI-Ready with Kyle’s 5-Day Challenge: https://aiwithkyle.com/join

Kyle’s AI Canon — Curated list of AI learning essentials: https://aiwithkyle.com/ai-canon

Subscribe and turn on notifications to catch the next live stream: https://www.youtube.com/channel/UChlLglbHDASnoGkbjDeHnQg

This video is from the ‘AI with Kyle Live Stream’

Summary: Everyone asks me how to build a website with AI, and the answer is more complicated than it should be. The real problem isn't the building — it's that most people don't know whether they actually need a website or a web app, and that distinction matters enormously. Use something like Lovable or Bolt and you'll end up with a slick-looking site that Google simply cannot read, because client-side rendered apps serve up a blank page to search engine crawlers. Your SEO is dead before you've written a single blog post.

So here's how I'd think about it: if you want something static, informational, and discoverable on Google, use a proper website builder or a static site generator. If you're building something dynamic — user accounts, payments, dashboards — then yes, reach for the vibe coding tools, but go in with your eyes open. The right tool for the right job sounds obvious, but in a market where every AI product is screaming for your attention, it's genuinely easy to get this wrong. Pick based on what you're actually building, not what's trending on Product Hunt.

—— Time Stamps ——

0:00 Building a Website with AI: Which Tool Should You Use?

—— More Useful Resources ——

Find everything else at https://aiwithkyle.com/

Get AI-Ready with Kyle’s 5-Day Challenge: https://aiwithkyle.com/join

Kyle’s AI Canon — Curated list of AI learning essentials: https://aiwithkyle.com/ai-canon

Subscribe and turn on notifications to catch the next live stream: https://www.youtube.com/channel/UChlLglbHDASnoGkbjDeHnQg

This video is from the ‘AI with Kyle Live Stream’

First aired: 19 February 2026

Summary: Ethan Mollick just dropped his latest guide on which AI models are actually worth your time in 2026, and for the first time ever he's changed the format — because the game has changed. It's no longer just about which chatbot writes the best email. We're firmly in the agentic era now, where your AI is off doing tasks while you get on with your life. I walked through his breakdown of models, apps, and harnesses live, and it's probably the clearest map of the current AI landscape you'll find anywhere.

On the side, the OpenAI vs Anthropic drama continues to be genuinely funny and genuinely important in equal measure. From refusing to hold hands on stage next to the Prime Minister of India, to the OpenClaw saga where Anthropic managed to look like the villain despite being entirely in the right — it's been a rough few weeks for the Claude crowd. OpenAI, meanwhile, are doing a very good job of positioning themselves as the scrappy, democratic option. Strange times.

—— Time Stamps ——

0:00 OpenAI vs Anthropic: The Rivalry Heating Up

—— More Useful Resources ——

Find everything else at https://aiwithkyle.com/

Get AI-Ready with Kyle’s 5-Day Challenge: https://aiwithkyle.com/join

Kyle’s AI Canon — Curated list of AI learning essentials: https://aiwithkyle.com/ai-canon

Subscribe and turn on notifications to catch the next live stream: https://www.youtube.com/channel/UChlLglbHDASnoGkbjDeHnQg

This video is from the ‘AI with Kyle Live Stream’

First aired: 18 February 2026

Summary: Today we're talking about cognitive friction — the productive struggle that actually makes you smarter — and why AI is quietly eroding it. I've been thinking about this a lot lately, and the short version is: every time you offload a problem to AI instead of wrestling with it yourself, you're skipping the bit where the learning actually happens. Whether it's writing an essay, debugging code, or learning a language, the messy middle part isn't the obstacle — it is the point.

The good news is there are some genuinely useful ways to use AI that work with your brain rather than replacing it. ChatGPT's study mode, NotebookLM's quizzes and multi-format learning tools, and the Feynman method are all worth your time. The goal isn't to avoid AI — it's to stop using it as a shortcut for the thinking you actually need to do yourself.

—— Time Stamps ——

0:00 Welcome + Claude Sonnet 4.6: Is It Worth Upgrading?

—— More Useful Resources ——

Find everything else at https://aiwithkyle.com/

Get AI-Ready with Kyle’s 5-Day Challenge: https://aiwithkyle.com/join

Kyle’s AI Canon — Curated list of AI learning essentials: https://aiwithkyle.com/ai-canon

Subscribe and turn on notifications to catch the next live stream: https://www.youtube.com/channel/UChlLglbHDASnoGkbjDeHnQg

This video is from the ‘AI with Kyle Live Stream’

First aired: 17 February 2026

Summary: UK unemployment just hit a five-year high — same levels as COVID — and while politicians are busy pointing fingers at each other, I think the more interesting question is what role AI is actually playing in all of this. Spoiler: it's not firing people dramatically. It's doing something far more insidious. Companies are simply choosing not to hire. Entry-level jobs are quietly disappearing, and the 14% unemployment rate among 18-24 year olds is just the beginning of that story.

The structural shift here is the bit that keeps me up at night. Economic cycles reverse — interest rates change, inflation settles — but AI doesn't go away. It has permanently altered the hiring decision, and that's not something that recovers in the next bull market. The good news is there's actually quite a lot you can do about it right now, and that's what we dig into in this one.

—— Time Stamps ——

0:00 Introduction: UK Unemployment Hits a 5-Year High

—— More Useful Resources ——

Find everything else at https://aiwithkyle.com/