📆 ThursdAI - Jan 15 - Agent Skills Deep Dive, GPT 5.2 Codex Builds a Browser, Claude Cowork for the Masses, and the Era of Personalized AI!

Hey %%first_name%%, Welcome to what cybersecurity experts are calling “Black October” - after a series of critical vulnerabilities that allow remote execution were found in PHP and Node.js, several companies were hacked, and a new supply chain attack on the very popular LiteLLM library was acknowledged. The internet is freaking out, and we’ve tried to make sense of it all, with a security expert Josh Long on the show. Also this week, Anthropic stepped into the personal agent category with Claude Duet, their direct competitor to Dots from OpenAI, Muse from Meta etc’. Duet is proactive and always on, and can do work while you’re asleep! I got access this morning so this isn’t the full review yet. Also DeepSeek is back with a new model and their inference engine! Let’s dive in: ThursdAI - Highest signal weekly AI news show is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Hey, it’s Alex 👋 What a freaking week! I spent 3 days at AI Engineer in New York, and we did the show from a real studio in Manhattan (the internet did NOT cooperate at first, but we got there 😅). And while I was running around the Sheraton, the labs kept shipping. Meta became a frontier lab, for real this time. Muse Wonder is #2 on Arena, and then, in the last minutes of the show, Meta ALSO dropped Muse Spark Open, an 816B open weights model. Google’s big Gemini is finally back with Gemini 4 Argon, and Omar Sanseviero and Geneviève Huskens from Google DeepMind came on to show it off. Anthropic gave us 4 stories plus a rumor, and Nathan Benaich came by to talk State of AI 2026. My theme this week: AI engineers are sleeping on personal agents. More on that below, including how my Muse found this studio and got me coffee on the way here. With me: Wolfram Ravenwolf, Peter Gostev (Arena), Nisten, Yam Peleg and LDJ (who brought the breaking news, as always). Let’s dive in (all links at the end as always!) Meta is BACK at the frontier Muse Wonder - Meta’s first frontier model is #2 on Arena, and it smells BIG (X, X, Arena, Blog) Folks, we told you after Llama 4 was a disaster, we told you when Zuck started hiring people with soup: do not, under any circumstances, bet against Zuck (and now Alex Wang). Muse Wonder is #2 on Arena Text at 1513, 12 points behind Argon, and it definitely smells like a big model. There are 2 of them. Muse Wonder is in the API at $4 in / $20 out per million tokens, basically Opus 5.5 pricing. Muse Wonder Full, the better one, is only for Muse Code subscribers at $50 or $200 a month. Did you see what they did there? Everybody calls their top setting “Max”, Meta called it Full, so it’s... Wonder Full. Wonderful 😂 No open weights for either, Meta says they’re too big and maybe too risky. Meta’s own table beats Opus, Astra and Fable on a bunch of evals... but Personal Superintelligence Index, GDP.pdf, JobBench, MemoryBench (which Meta built, and then won)? I’ve never seen most of these. Every launch now is the iPhone 18 moment: numbers bigger, best one yet. On evals we know, Artificial Analysis has Wonder Full at 56 and Wonder at 52, Opus 5.5 on top at 58. And on OSWorld, the Wonder we can use sits below Fable and Opus, and computer use is the number one thing I need from an assistant. Peter’s Arena view: #2 on text means people like talking to it, but it’s 12th on Agent Arena, and he tried building websites with it, “it was kind of awful.” His theory: Theo loved Wonder Full and thought the API Wonder was bad, so maybe Meta is talking about a different model than the one we got. Not ideal. He also found the half-life of a benchmark quoted in a launch is about ONE release: quote it once because the number looks good, never see it again. Wolfram says it’s excellent in German. Nisten joined not knowing what Muse Wonder was (”No idea”), sometimes our co-hosts come to get their news from the show. My Muse got upgraded with a click Here’s where I disagree with the timeline. “It can’t do 3D scenes, so it sucks” misses the point. Meta is tuning Wonder for Muse, the personal agent. Officially, Wonder will “soon” power more of Muse, subscribers first. I’m on the paid plan, and my Muse (Wilfred) flipped to Wonder the moment I subscribed. Matrix moment: “I know kung fu.” Then I kept shoving tasks at it. Muse found the podcast studio I recorded from in NY and emailed Ezra at Virgo Studios with my requirements, got me coffee on the way here, and designed and ordered the wristbands for our AIE sauna meetup. The failures weren’t logic, they were the real world: a site that’s blank because Cloudflare doesn’t like agents, a Stripe link that wouldn’t connect, and businesses that hung up because they didn’t want to talk to a personal assistant 😂 Peter’s read: that’s the harness, not the model, and it’s exactly why this stuff is so hard to benchmark. Muse Spark Open - Llama is SO back (X, X, Blog, HF) In the last minutes of the sho

Hey folks, a quick bonus episode! At CoreWeave's Fully Connected in San Francisco, a few minutes after she walked off the keynote stage, I sat down with Chen Goldberg, EVP of Product & Engineering at CoreWeave. We talked about Forge, the loop every AI engineer is running now (run, observe, curate, improve, evaluate), what carries over from the Kubernetes days to agents, and what breaks first when agents hit your infrastructure. Video below, and the audio is right here in your podcast feed. Enjoy!

Hey %%first_name%%, another banger AI week, let me catch you up! We say this often but this week we all felt the acceleration! Just look at one Tuesday. OpenAI dropped 722 math manuscripts. Meta( With Stripe, Shopify and Walmart) announced a new agent protocol. OpenAI opened up the Decisions API. Mistral came back with Le Chonk (Mistral 4), Claude moved into Google Docs, and we at CoreWeave shipped RL Rollouts. That was ONE day. Then the rest of the week happened 😂 Haiku 5.5, D1, and tons more! I had 48 topics on my list, so I asked Claude to build me a Tinder for AI news: I swipe on each story, and it stack-ranks what makes the show this week. I hope we did good (lmk in comments if we missed a major news story) With me: LDJ, Peter, Yam, Nisten and friend of the pod Maxime Labonne from Liquid AI joined us to talk decision models. Let’s dive in, all links at the end as always! OpenAI solves Math!? OpenAI drops 722 math manuscripts from a model nobody can use (X, GitHub, Blog, Fable’s tally) Remember when ONE Navier-Stokes result was the whole show? On Tuesday night, OpenAI quietly pushed 722 math manuscripts to GitHub, grouped into 372 families of results. No hype video, no exploding-head emojis, just a very thin blog post. My favorite new term for this is a “slop grenade”: somebody throws a mountain of output at you, and now you have to shovel through it. An internal model nobody outside OpenAI can use was pointed at about 4,000 open problems, at about 3 hours of ChatGPT Pro-level thinking per result. Will Depue asked Fable to measure the drop in Navier-Stokes units. The answer: roughly 5. Roughly 5 “Navier Stokes” size solutions dropped all at once! The math is so advanced that a mathematician in one family of results often can’t follow the family next door without an LLM explaining it. I find that fascinating and scary at the same time. Peter’s tried this before! For weeks he threw GPT-6, Fable and Opus at one problem, Hadwiger-Nelson. He thinks he burned 300 to 400 BILLION tokens, and in his words, “I did not discover a single bloody thing.” And this new OpenAI’s model spent about 3 hours on it and narrowed the bounds from 5-to-7 down to 6-or-7 😅 Then LDJ dropped a stat that I made him repeat slowly. Fable and Astra had put together a list of the 500 most important open problems in math ever. On Tuesday, OpenAI dropped solutions to 92 of them. There’s also a list of the 100 most significant problems from the last 12 months, and over 80% of those got solutions in this drop. Absolutely insane, folks. Yam’s reaction was a very loud “F***ing go.” Yam’s favorite is the Riemann one. It doesn’t prove the hypothesis, it bounds a region for the zeros, which “was never done before, and many, many, many people have tried.” And to the “it’s just brute force” crowd, Yam says: “Let’s brute force everything.” Yes please, room temperature superconductors next! LDJ’s mathematician friends think that on some of these, the model used fewer tokens and less time than a human would. So who’s brute forcing whom? 🤔 The missing crypto results (speculation!) If you hold any crypto, this part is for you. The manuscripts are numbered, and some numbers are missing. There’s a #44 and a #46, but no #45, and LDJ counted 4 or 5 gaps like that. He also looked at which topics made it in, and cryptography is almost absent. So here’s the theory going around. Maybe OpenAI found something big in cryptography and held it back, either by its own choice or because someone asked them to. LDJ said it himself: it sounds conspiratorial. But the US government really can stop cryptography research from being published on national security grounds. To be clear, this is SPECULATION. Nobody outside OpenAI knows what’s in #45. But Bitcoin’s security rests on elliptic curve math. Imagine a paper that shows a way into the 25,000 old Satoshi wallets, each holding 50 BTC. Not great for the price, not great for the world, and not great for encryption in general. “Are you saying y

Hey %%first_name%%, it’s Alex 👋 What a freaking week! This one was special. We came to you LIVE from the middle of the show floor at Moscone, from CoreWeave’s Fully Connected, with robots walking behind us and a Vera Rubin rack a few feet away. ThursdAI is only possible because of CoreWeave, and this is their biggest event of the year, so we packed up the mics and did the show right there (at 11am, which confused a bunch of you, sorry!). It’s October. The last quarter of 2026. And nobody’s pacing! There’s no pacing the frontier. Three big models this week (GPT-6.1 Sol, Claude Sonnet 5.5 and Gemini 4 Argon), OpenAI’s biggest DevDay ever, and the story I’ve been yelling about for weeks finally went mainstream: OpenAI joined the AI assistant race. With me: Wolfram Ravenwolf in person on the floor, Peter Gostev from Arena (who was at DevDay with me), Nisten and LDJ remote. In hour two we interviewed four folks from CoreWeave, on physical AI, Forge, sandboxes, and one piece of breaking news I’m still excited about. Let’s dive in (all links at the end as always!) OpenAI DevDay 2026 - 20 launches and a new assistant Dots - OpenAI joins the AI assistant race (X, Blog) Meta has Muse. Grok has Grokbot. Anthropic has Claude, but it’s not really an assistant. And this week OpenAI stepped in with something called Dots. I was in the room at Fort Mason when Sam talked about it, and it was the headline of their fourth DevDay, which honestly felt like OpenAI’s WWDC: 20 releases, the biggest DevDay ever. Funny thing, the original ChatGPT system prompt was literally “you are an AI assistant.” But it wasn’t proactive, it never pinged you, it just sat there. So OpenAI looked at their competitors, looked at Peter Steinberger from OpenClaw (who they hired like 8 months ago), and built this. A dot is an always-on agent in ChatGPT, powered by GPT-6 Astra, with its own computer and its own browser in the cloud, connected to 4,000+ apps people already built for ChatGPT. It works 24/7, you can talk to it in Slack or Teams, and there’s an interface that looks like a phone call. I think that’s going to be great for my mom. Sam literally said he uses dots to run OpenAI. And my favorite DevDay moment: Romain Huet’s live voice demo didn’t work, because somebody was deploying in the middle. But Thibault’s dot had already pinged him that the demo might break. The dot knew before the humans did 😂 Wolfram (him and Amy are a known duo) loved that Sam uses it for actual work. Peter was more honest: underwhelmed at onboarding (”you connect and it’s like... then what?”), but once you’re past that you’re talking to Astra, and it’s great at juggling threads through one bot. Same shape as Muse. Wolfram’s version: one executive assistant, and a lot of sub-agent employees working for you. Dots is Pro only for now (including the $100 plan), while Meta gives Muse away free. With 1.2 billion weekly ChatGPT users, this goes to everyone eventually. This is the race now. GPT-6.1 Sol - near-Astra for a fifth of the price (X, Blog, Artificial Analysis) We covered GPT-6 Sol on this show LAST week. Five days later it’s already replaced. GPT-6.1 Sol is the same price, $2 in and $10 out, but cached input is now 95% off, 10 cents a million tokens. OpenAI’s pitch is near-Astra intelligence for a fifth of the price. And the independent numbers kind of back that up. Artificial Analysis has it 1 point below Astra at 72 cents a task vs over 3 dollars. On Deep SWE it actually beats Astra at about a sixth of the cost. It’s the default in Codex now, and folks are calling it the workhorse. LDJ said it has “less of the big model smell” but loves the token efficiency, Sonnet and Opus 5.5 burn a LOT of tokens. Peter has it 4th on Code Arena, above Fable, and says it’s a bit “artistic,” it goes into its hole and comes back with the thing. And his verdict is the one I had to repeat straight to the camera: It does not make sense to use Fable or Astra as your daily driver right now. Opus 5.5 is enoug