Sakana AI didn't train another giant model with Fugu. It trained a 7-billion-parameter conductor that routes tasks across a swappable pool of frontier models. CEO David Ha calls it the next stage — 'beyond bigger models'.
Unconventional AI just dropped its Un-0 model series. The pitch: run AI on tiny oscillators instead of traditional chips. Sounds wild, but Jeff Bezos put $475 million behind it.
Anthropic's new Economic Index report now samples Claude usage by the hour — and reads the rhythms of real life in it. Recipes at 6 p.m., sleep advice at 5 a.m., a tax spike around the US deadline. But the truly interesting part is the survey behind it.
The releases up to 2.1.195 are classic polish: you can now turn off mouse clicks in fullscreen, the voice-dictation quirks on macOS and for Japanese/Chinese are gone, and background agents no longer vanish from the overview.
Anthropic is hardening remote control for Claude Code. With Trusted Devices, admins can require a device to be verified before it can steer local sessions on the go — via Face ID or a passkey, and without any biometrics ever leaving the device.
A court ruled that Google's AI Overviews are Google's own words — making the company liable for false answers. Bruce Schneier draws a clear line from it, and Simon Willison picks it up: an AI agent acts for whoever deploys it. And that party is on the hook.
GPT-5.6 under government oversight, Gemini 3.5 Pro delayed, Grok 5 nowhere to be seen: June 2026 is the first month where no frontier model got a normal public release.
Google has confirmed Gemini 3.5 Pro won't ship in June. It's the second consecutive time a model announced at I/O has missed its target window.