In partnership with
News of the day
1. Uncertainty surrounds the government's approval process for OpenAI's Sol and Anthropic's Fable models, raising questions about transparency and oversight. → Read more
2. Anthropic's J-lens tool reveals Claude's hidden 'J-space,' offering insights into its thought process. OpenAI launches its 'super app,' ChatGPT Work. → Read more
3. French tech leaders are launching AI 'clones' to coach entrepreneurs, replicating their expertise and advice 24/7. → Read more
4. AI infrastructure spending reaches $1.5T in 2026, requiring $3T in revenue to justify. Hyperscalers face risks if cash flow goals aren't met. → Read more
Our take
Hi Dotikers!
A few days ago, we were watching companies refuse to marry a single model provider. Led by Coinbase and Vercel, they now pick freely depending on the task and the cost. The downstream market is opening up. Except that upstream, another door has just quietly closed.
OpenAI released GPT-5.6 to the public this week, after twelve days of preview access restricted to some twenty partners hand-picked by the US government. The Center for AI Standards and Innovation, housed within the Department of Commerce, ran the evaluation. OpenAI sent its engineers to Washington to field questions, and then the green light came. The official rationale: the model's cybersecurity and biology capabilities warranted a review before release.
The trouble is that nobody really knows what this review consists of. No published criteria, no timeline, no clear legal basis, not even agreement on which agency is supposed to decide. The framework is officially voluntary, and the White House swears it grants no authorization, which at least has the merit of consistency, since nobody knows how to obtain one either. In practice, shipping a frontier model in the United States now goes through Washington, whether the law says so or not.
The backdrop does not help. Behind the scenes, Altman is offering 5% of OpenAI's equity to the "Trump Accounts," and his president Greg Brockman is funding the administration's political operation. From the outside, the line between security rigor and well-placed proximity is getting blurry. The real danger is not regulation that is too heavy, but regulation applied case by case, where market access depends on having the right contacts. For Europe, which relies on these models, the everyday tool is now cleared by a process nobody gets to watch.
Alex.
You've seen the AI demos. Viktor does it without you watching.
The AI tool you tried last quarter waited for a prompt, hallucinated a number, then asked if you'd like a summary.
Viktor opened a PR at 2am, rebased it against main, ran your test suite, and posted a note in #eng: "Two flaky tests in payments service, both pre-existing. Recommended merging after fixing them." Then drafted the customer reply for the support ticket the bug created.
That's 619K autonomous actions per day across 20,000+ teams. Not chat replies. Real work shipped to GitHub, Stripe, Linear, Notion, and 3,000+ other tools, from inside Slack and Microsoft Teams.
You don't supervise him any more than you supervise a senior engineer.
SOC 2 certified. Your data never trains models.
"It's what you probably originally thought AI was going to be when you first heard of it in sci-fi movies." Tyler, CEO.
Meme of the day
