Desk
Coverage tied to OpenAI products and research posts.
Two rival labs say their models found a faster-than-expected algorithm that overturns a 1999 pillar of complexity theory. Computer scientists are now organizing to verify it, and the way they’re doing it may matter as much as the result.
When a government summarizes asylum interviews with a model it will not fully disclose, the efficiency gain is real and the accountability gap is the bigger story.
Retell's growth from 30 million to 100 million monthly calls is a real signal that voice agents are leaving the demo stage, but the numbers come from a vendor showcase.
Apple is putting an assistant inside its office apps and opening them to outside agents, and the second move may matter more than the first.
Data centers have become a rare issue that unites angry voters across parties, and the AI industry is learning that money spent on campaigns can't make a neighbor like a server farm.
Oracle putting its business apps inside ChatGPT is a bet that the chat window becomes the front door to enterprise software, and that's a bigger shift than it looks.
OpenAI's dots is mostly its coding agent in friendlier clothes, and the real story is what it takes to make a proactive assistant work: more of your data, more tokens and, soon, your payments.
Check Point cutting vulnerability research from hours to minutes is the defensive case for frontier cyber models, and it raises the bar for keeping humans in charge.
Greptile is betting that AI-written code now ships faster than humans can test it, so it is pointing computer-use agents at apps to break them before users do.
OpenAI says its models disproved the Exponential Time Hypothesis and proved existence results tied to a Millennium Prize problem; if these hold, it's historic, and the verification process now matters as much as the math.
CodeRabbit is using GPT-6.1 Sol to look for security problems across whole codebases, not just single changes, and the human sign-off it keeps in place is the part we'd protect.
AI models are getting better at translating COBOL, but experts say rewriting code isn't replacing a system, and the gap between the two is where critical infrastructure could break.