franco

Using Gemini 4 makes you realize how many other models must be distilling Claude/GPT (or using a similar training pipeline). It feels quite different than other models.

I spent a decent amount of time with Gemini 4 before launch. Overall, Gemini 4 is a big jump over previous Gemini models and it is the first non-OpenAI/Anthropic model to pass the Fable/Mythos/Astra bar. It is also much more pleasant and more agentic then previous Gemini models, without the weird personality issues that you often found in Gemini 3 series models. In fact, a big advantage is that has a very different personality and sense of taste than Claude or Astra, making it a good addition to the AI world. On the usefulness side, in some areas it is at least at par with OpenAI and Anthropic (it is quite creative and a good writer), but in others it lags, including at complex multi-agent work. It has a tendency to be too certain about its conclusions and a little lazy, issues not helped by the Antigravity harness not being quite as good as the competition. Still, I expect to see rapid improvements in model abilities using Gemini 4 as a base. Here is a fun example: "create the ultimate brutalist city builder. you should have lots of controls over buildings in ways that are delightful and inuitive to people who just want to play, but are sophisticated and deep for people who know something about architecture. you should be able to play a lot with design and approach. there should be simulation modes and different ways to view or experience the city. make it stunning. i want to host this on netlify in the end" Astra Max in Codex: formwork-city.netlify.app/Fa… 5.1 Max in Code: brut-city.netlify.app/ Argon in Antigravity: rainbow-panda-689e6b.netlify…

A worthwhile (and unnerving) reminder from Anthropic that we should be paying attention to AI ability thresholds, not just steady climbs In this case, the threshold is the capability needed for a self-replicating AI worm on IoT devices like off-brand TVs anthropic.com/research/self-…

Yes, and there is a case to be made that OpenAI and Anthropic will do the best job of figuring out how to optimize AI cost/intelligence with combinations of agents of different ability levels that have been specifically trained to work well together in their own harnesses.

A mathematician makes an Opus 5.5 music video in response to OpenAI’s proofs. Pretty great.

Not only did Astra beat Montezuma's Revenge, but so did Opus 5 But one of the criteria to resolve this AGI bet is for AI to win a discontinued prize that was sort of like a weak Turing Test. So Metaculus has decided to recreate the test to confirm the criteria is resolved. Neat!

Leaving aside Anthropic's incentives for publishing this research, there is no doubt that open weights models will soon create the same security threats that closed source models have been demonstrating, except without guardrails. We are close. Probably good to plan accordingly.

Hey, Claude: "I asked an earlier Claude to Remove the Squid from All Quiet on the Western Front. Now you can make movies and such. I need you to show how far you have come" & I pasted in the tweet below There is some clever stuff, models have a real sense of humor at this point

Since people liked this, and the weak point was the text to speech, I asked Opus to update with an ElevenLabs voice. This is the same video, but with much better narration.

I was reflecting on how many random contingencies went into creating the current accelerating AI moment and had Opus 5.5 put together a video, inspired by old science show Connections, to try and point out how a series of initially unrelated ideas led to the moment we are in now.