The graph error OpenAI made has nothing do with AI but with internal organisation and quality controls. In my own work it would be unthinkable that an error of this magnitude would slip through to the live presentation. How can/could this happen, is OpenAI a very strange / atypical organisation?
And if they can't control the quality here what means that in a general sense?
#OpenAI #GPT5 #qualitycontrol