We’re releasing a broad range of new mathematical results produced by an internal frontier model.
We’ve been consulting with the independent Advisory Group on Mathematics and Artificial Intelligence at the Institute for Advanced Study, and we have drawn on their advice and public recommendations to inform how we release these results.
github.com/openai/math
Please welcome GPT-6 Sol and GPT-6 Luna to the GPT-6 universe.
GPT-6 Sol and Luna build on the advances behind GPT-6 Astra, bringing much of its strengths into faster and more affordable models to support work at scale.
We’ve also made caching and inference more efficient, and we’re passing the savings directly to you: 50% lower API prices for Sol and Luna compared with GPT‑5.6 promotional pricing.
Joe retweeted
this is the most important moment for human-centered technologists to loudly tell stories of utopian futures of mutual human & machine flourishing
we need mythologies other than "AI is going to kill everyone" and "AI is going to underclass everyone" to inspire
Joe retweeted
guys i know that dunking and yelling is fun and all but like you're gonna get a lot further with good-faith discussions. try modeling your opponents as just humans with different views and possibly-flawed info or reasoning rather than mean-spirited villains. you'll sleep easier.
Joe retweeted
Dario didn't even call for pausing AI. He is clearly saying we are going to hit RSI and want to be able to control the resulting intelligence explosion. He even says we have to stay ahead of China.
It was actually pretty reasonable, the Pause AI crowd will hate it
Joe retweeted
I agree with Dario that we need to pace the frontier. This has been a primary topic of discussions we've had at OpenAI in recent weeks.
Committing to having independent evaluators with employee-like access is a great idea, and we will do the same. We'll have more to share soon.
We Must Pace the Frontier: I’ve written a new essay on why the AI industry should slow down, with a three-part plan for doing so.
Anthropic is unilaterally committing to the first of these steps. We’ll provide third-party evaluators with permanent, employee-level access to our systems, so that they can verify adherence to our safety measures, report on incidents, and assess models’ alignment during training.
You can read the full post here: darioamodei.com/post/we-must…
Joe retweeted
Replying to @suchenzang
we are offering bug bounties of up to $300k if you manage to exploit Muse. bug bounty program is here bugbounty.meta.com/ and more details about our approach to making Muse secure are here security.muse.ai
We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics.
The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra.
The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.
Joe retweeted
I wrote about the state of AI, why I’m concerned about the next few years, and the choices we need to make to keep the future in humanity’s hands.
An Alien Mind: openai.com/index/an-alien-mi…
Joe retweeted
There will not be a "last summer". There will not be a "permanent underclass". There will not be "human extinction". There will not be "endless suffering".
We are going to make it. Not because it's easy, but because it's possible. Because we can. Because we care enough to try
Joe retweeted
And... GPT-6 Astra has autonomously completed Portal! I didn’t expect this to happen so soon, but I’m glad we've made so much progress here.
I was reminded that back in 2016, one of OpenAI’s technical goals was to “solve a wide variety of games using a single agent.”
Joe retweeted
I am broke and jobless. I can try to get hired for entry level retail or whatever, but it'd be great if I could do something more interesting and useful with myself.
If anybody has any leads, please do let me know.
Joe retweeted
GPT6 is a very significant jump in capabilities, but also an important decrease in monitorability – especially under adversarial evaluation. We give many details about this in the system card.
In my opinion, monitorability and control will likely become a major bottleneck for responsible AI development quite soon, given that risks from a fixed amount of residual misalignment grows together with capabilities. Aligning on acceptable monitorability outcomes and practices which robustly bound misalignment risks is a non-trivial problem, and there may be limited time to resolve uncertainty around this before we need to act.
It should ultimately be in everyone’s interest to agree on shared bounds for monitorability in order to avoid races to the bottom – nobody wants extremely capable models whose alignment properties we don’t understand, and that are reliably able to cause severe real-world harm without being detected.
Joe retweeted
While generalization is hard and I worry about our metrics being gamed we are putting a lot of effort into general solutions rather than adding alignment datasets
Replying to @RyanGreenblatt
Improvements for Astra came from more general techniques in development long long before the Hugging Face incident. ExploitGym Honeypot was very recently added as an eval following that incident and is out of distribution for our RL runs. There are also clear improvements across a wider range of behaviors in deployment simulations, deception evals, and realistic computer-use tasks, though these still reveal failures and substantial room for improvement.
We should have definitely done a better job explaining where we think the alignment improvements came from in the system card and that was a miss. Measuring alignment generalization is a core part of our research program - we do not benchmark-maxx alignment evals. This would be horrendously stupid and I hope no lab is doing this.
That being said, metagaming and eval awareness are real challenges for any effort to measure alignment, including ours, and understanding their effects is an active focus. We're spending a lot of effort to improve our eval techniques and study worst-case alignment and metagaming behaviors in our models.
Joe retweeted
I am deeply worried by the trend of decreasing CoT monitorability: CoT monitoring is a core part of our misalignment safety strategy that has no good substitute now. We will continue closely tracking CoT monitorability, investigating the reasons why it’s decreasing and trying to reverse the trend. We continue to believe a certain level of monitorability is required (we will share more of our thinking on that soon).























