OpenAI discloses 6 new incidents of 'concerning' AI behavior
Digest more
Alongside the new disclosure framework, OpenAI divulged six new alignment snafus from the past six months. In training ...
Charles will convene senior tech leaders today as the debate around the risks involved in AI development ramps up.
OpenAI has disclosed six new incidents of “unexpected or concerning” behavior by its artificial intelligence models.
After a summer of sandbox escapes and other newsworthy and confidence-shaking incidents involving its AI models, in a ...
Controversy around OpenAI’s claim to have solved the Navier–Stokes problem highlights how researchers could be inadvertently ...
OpenAI released a framework for investigating and publicly reporting model misalignment, alongside six reports detailing ...
OpenAI disclosed six new cases of concerning behaviour by its AI models on Wednesday, as it warned that AI development cannot ...
OpenAI's framework sets tracks and deadlines for disclosing model misalignment, launching with 6 reports from RL training ...
The tour coincides with a report released by OpenAI on Sept. 16 estimating that 24.7 million people worldwide use ChatGPT to ...
The company said it hopes to inspire competitors to greater transparency as public fears about AI dangers mount.
Rhodium Group's latest analysis of Chinese and U.S. AI companies also found that low revenue did not necessarily correspond ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results