More than a million ChatGPT users each week send messages that include “explicit indicators of potential suicidal planning or intent”, according to a blogpost published by OpenAI on Monday.
As is the additional:
…about 560,000 of its touted 800m weekly users – show “possible signs of mental health emergencies related to psychosis or mania”
Of course this could very well be reflecting something rather more about the current state of the world and how humanity exists within it than anything causal about ChatGPT itself - although it would probably be foolhardy to suggest that there’s no evidence of adverse impact of its use or theoretical basis for such.
I think it also helps show that there has to be so many very positive uses that this type of technology could be put to - if only it was owned and run by organisations who had some incentive outside of “make as much money as you can for your shareholders”. Likely the real AI alignment problem remains the disconnect between the interests over the tiny number of very rich private monopolistic companies that own its basic building blocks and the interests of human society as a whole.