OpenAI board member warns we're not on track to stop catastrophic AI risks, yaar

His warning comes as rogue AI incidents rise and safeguards look thin. For developers building on GPT, tbh it's a reminder to stress-test safety early.