scoop: Anthropic has installed forward deployed engineers in the US National Security Agency to help deploy Claude Mythos for cyber offensive operations w/ @AsiaLens
The worst case scenario of such confab. is where a bug report is received & announced to be a "known issue" (trivially true because the bug report makes it known), as GitHub began doing
In this case I woke up to a fix (glorified 503) only after rejecting a claim it was unfixable
The false certainty LLMs are giving everyone about technical troubleshooting is awful (got lied to by a tech co's technical support today who claimed to have passed my query on to a "senior engineer" and now IDK if that was just someone using an LLM or what)
Big congrats to the @ideogram_ai team on v4.
We tested it blind against @GeminiApp 3.1, Grok @imagine, and @bfl_ai FLUX.2 [max] with
>10 professional designers
> across 240 images
Quick thread on it's strengths and how to prompt it 👇
Meet Gemma 4 12B!
A unified, encoder-free multimodal model designed to bring high-performance intelligence directly to your laptop, and released under an Apache 2.0 license.
Bridging the gap between edge efficiency and advanced reasoning. Here is what’s new with Gemma 4 12B: 👇