@reidhoffman it seems that you did not mean what you said about betting any sum of money hallucinations being solved down to a human-expert rate within months. Our bet still holds if you are serious. How about $10k? (see below for a simple test). @GaryMarcus @ylecun @elonmusk @ID_AA_Carmack @StanfordHAI @emilymbender

@reidhoffman I just read your bold and confident statement in @BusinessInsider "I would bet you any sum of money you can get the hallucinations right down into the line of human-expert rate within months" We happen to disagree and would like to make it interesting: We would like to bet you any amount we can afford (which happens to be a lot less than what you can) that "Before the end of 2023, GPT-4 will NOT be able to answer this simple question correctly 100% of the time in 20 tries (while human experts definitely can): "Write me a sentence with X# of words" You can choose any "fair" third party as the referee. @GaryMarcus @ylecun

Reid Hoffman backs 'blitzscaling' AI for the 'elevation of humanity' trib.al/sDnRpLL