"Tomorrow's not promised. But make plans anyway." —@Lin_Manuel Miranda

Is the AI gold rush built on shaky ground? Intelligent Machines, with @jeffjarvis, @parismartineau, & @jasonhiner, goes inside Anthropic’s blockbuster IPO and its huge compute bets, plus why some insiders say the industry’s math doesn’t work.

2026 has proven that AI is only amplifying that human connections are more meaningful and valuable than ever. See my photos from this year's events over on LinkedIn: lnkd.in/p/eQV_uWVV

The iPhone Duo is here. Don't worry about FOMO yet. Here's what I wrote on LinkedIn: lnkd.in/p/gJSe_5_n1

The Decisions API is now generally available! To celebrate, I had Codex reimagine a very important app… 🌭

Could diffusion LLMs upend the AI industry? I talked to @_inception_ai CEO @StefanoErmon about Mercury 3, which delivers faster, cheaper AI inference using a radically different architecture. The implications go well beyond speed. thedeepview.com/articles/why…

Between this and our prior commitments, roughly 20% of all Anthropic equity will be dedicated to the public good, and that commitment can grow as Anthropic grows. We plan to use these funds to ensure more people can share in the benefits of AI, and to increase our ability to provide funding to deal with its risks.

Inception launches Mercury 3, a diffusion-based LLM it claims is "comparable to Luna and Haiku models" but 4x to 8x faster, generating ~1,000 tokens per second (@jasonhiner / The Deep View) (Visit Techmeme dot com for the link and full context!)

Super Intelligence must be contained... Today, this is an engineering and governance challenge. And it isn't new. We've done it with planes, cars, nuclear materials, food safety, medicines... and everything else. Let's get to work!

Introducing Beam: a highly efficient agentic open model with 501B total parameters and 23B active. - Frontier reasoning efficiency - Advances the Western open frontier on coding & agentic tasks - Trained end-to-end from scratch Full weights release this month. Learn more about Beam: reflection.ai/beam

Mistral releases Mistral Large 4, dubbed "le Chonk", a 1T-parameter open-weight model for general agentic capabilities, trained on 4,000 Grace Blackwell GPUs (@sabrinaa_ortiz / The Deep View) (Visit Techmeme dot com for the link and full context!)

"This study talked to young people who are in romantic relationships with AI, and a quarter of them said something that indicated they thought that this relationship might be able to move out of the computer realm, so either they meet up in real life, or maybe someday they'll see them in the afterlife, or something like that." —Dr. Marie Bragg of @nyuniversity at a panel that addressed child protections in chatbots at the @Stanford Trust & Safety Research Conference (#tsrconf) today

I’m at @Stanford today and tomorrow learning from researchers from all over the world about what they are learning about AI safety, ethics, and governance. #tsrconf @StanfordTIP

"Here's the reality: AI agents are being integrated into people's lives before anyone has built the infrastructure to truly hold them accountable. There's growing gap between what these systems can do and what people affected by them are able to understand, contest, or even see." —Anya Svan, @UCBerkeley, at the @Stanford Trust & Safety Research Conference (#tsrconf)

"Instead of asking only whether a model is aligned, we should also ask how controllable is it for the actor responsible for deploying it." —Renata Barreto, @eBay safety team, at the @Stanford Trust & Safety Research Conference (#trsconf)

My panel on AI transformation at The @AIconference delivered real-world tips for a great audience on the main stage yesterday. Big thanks to Claire Southey, Chief AI Officer of Rokt, and Todd @Earwood, CEO of Betraworks and an AI investor for 11 years. Of all the people I've talked to about AI in 2026, I hand-picked these two for the panel because I knew they had practical advice that AI folks would love to learn from. They delivered and this was confirmed by how many people were excited to chat with them after our panel. I want to share two of the most helpful insights: 1.) CLAIRE: Don't get fixated on only using AI to increase productivity because that is essentially just making yesterday's tasks more efficient. Be on the lookout for the second-order effects that AI can help you drive. Those are hard to see until they start to emerge. For example, when the dot com boom initially unfolded, no one foresaw social media would come from it, and when the smartphone launched, no one realized ride-sharing would come soon. Same applies on a small scale to your organization. When you implement AI, keep an eye out for the second-order effects that could emerge and create new opportunities. 2.) TODD: When you start looking for places where AI could make an immediate impact in your organization, don't just go to the executives and top leaders. They often don't fully understand how the work gets done at the company. Instead, take a bottom-up approach and involve the individual contributors at the grassroots and ask them about their biggest pain points. Then you can brainstorm and consult about where AI could have the biggest and fastest impact.

So @OpenAI #DevDay2026 happened today and we had questions (at the press conference with @thsottiaux, @sama, and @embirico). Check out tomorrow morning’s send of @TheDeepView for our perspective.

Over the past 5 weeks, the team completely rewrote ChatGPT Web on top of the desktop app codebase. This was only possible with Astra, and it’s why Dots, Space, Codex Cloud, and so many more of our launches will come to web and desktop same day. Era of the megaproject begins.

And to celebrate the launch of Space, we're dropping one more icon choice for your Dock. Introducing the Spaceblossom. Appearing shortly in a Settings section near you.

There is a groundswell of AI happening in the real world that never makes the daily headlines. But that's where some of the most interesting stuff is unfolding and that's what we're exploring in a session I've put together for @TheAIConference tomorrow on AI transformation. I talk to lots of people about AI every day and I'm constantly meeting new leaders, builders, and companies. For this panel, I brought together two people who are doing some of the best work I've seen in 2026: Claire Southey of @ROKT and Todd @Earwood of HalfCourt Ventures. Claire is working in ecommerce and her team is using AI to fast-track their work and supercharge what they can do for customers. Todd is training leaders and enterprises on how to effectively implement AI to drive immediate impact and ROI. I can't wait for the audience at The AI Conference in San Francisco to learn from these two, and I'll share the insights afterward on LinkedIn.