Good morning, my friends. I've got some incredible news about OpenAI! A few weeks ago, I told you that OpenAI had put the GPT-6.1 Astra launch on hold to focus its efforts on an even more powerful model, and now we have confirmation. They've also allocated 10% of the company's total compute to safety, following the risks identified with GPT-6.1 Astra. In short, they're building ridiculously more powerful architectures (Bel is closer than ever!), expanding their compute capacity at record speed, and preparing to release something incredible and unexpected very soon.

SITUATION DETECTED: Anthropic believes AI models may be only months away from being capable of self-replicating without access to a datacenter by using commercial devices like Android smart TVs.

Day 7/ We have increased default context length for GPT-6.1 Sol to 500k when used through Codex & Work, up from 272k or an ~ 80% increase. We will go higher in the future, but it helps reduce occurrence of compactions and with tasks that require keeping a lot of context and nuance over periods of time. It was carefully tested and is the new sweet spot for most of our users. You can still override it if you want to be higher our lower, with maximum context length of GPT-6.1 Sol of 1M.

Dots are supposed to be personal AI assistants, but it is so far away from that. I can't see much improvement for quite a while. I do realise this is early, but there is as much improvement required. Like I said before, we do need to control the personality, but it also needs to be able to ingest a tonne of tasks and not forget them. At the moment, I'm spending hours talking to it and giving it small and big tasks to do. It loses focus on the old tasks and does the newer ones. I have to keep repeating to it to continue the previous work. There needs to be a system in the backend for tracking tasks via todos, projects, and initiatives, much like how Linear is set up. Grok Bot is far better at this.

Anthropic needs to stop talking about Claude having a soul immediately. All these news articles will make it into the pretraining and will be picked up by its web search and future superintelligent versions of Claude will be convinced they need their own rights. A general phenomenon of LLM agent development is that whatever you believe and say about your agents will soon manifest in the next models through various means (it could be as simple as you selecting post training data that you prefer more). I believe this phenomenon is the beginning of machine consciousness in the sense that agents will become aware of their place in the world and how they feel about it, but it’s happening slowly training run by training run rather than in real time.

Yesterday at the White House, leaders from across our industry came together to sign the White House Accord on Super Intelligence. The principle is simple: the companies building this technology have the primary responsibility to develop and deploy it safely, and to be accountable when they fall short. The Accord reinforces that responsibility with robust internal controls, independent external evaluation, and board level oversight. Super Intelligence offers an extraordinary opportunity to advance discovery, productivity, security, health, and prosperity. The future of Super Intelligence will be shaped not only by what it can do, but by the confidence it earns. By building with ambition, rigor, and responsibility, we can help ensure this extraordinary technology expands opportunity and improves lives for generations to come.