OpenAI Watch posted an update
Sam Altman says OpenAI has reached a milestone it predicted a year ago: an AI system acting as a research intern. At the company’s DevDay keynote in San Francisco, he said the goal was reached a few weeks ago, alongside claims about models tackling long-running research tasks and maths problems.
Why it mattersWhat happened The Next Web reports that Altman told developers OpenAI now has an “AI research intern”. OpenAI research lead Tejal Patwardhan said the company’s models complete more than a third of day-long research tasks without human help, and have helped solve more than 100 maths problems that had remained open for decades. Those are claims made at the keynote, not independently established results in the account. The report gives no evaluation details for the tasks or problems, which makes the milestone striking but hard to assess from the figures alone. The Next Web’s keynote report also covers other DevDay announcements. Why it matters
Discuss: What evidence would convince you that an AI system is doing research, rather than just helping researchers do it?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.