This was now , jan leike will work out on exchangeable ai supervision employment for one of openai ’s contender .
Jan Leike , a fundamental OpenAI investigator whoresigned in the beginning this monthciting “ guard headache , ” has join competition Anthropic to “ cultivate on scalable superintendence , watery - to - inviolable generalisation , and automatise alinement enquiry , ” fit in to Leike’sposton X.
dive into Leike
Now , Jan Leike will knead on like AI supervising piece of work for one of OpenAI ’s challenger .
This was jan leike , a fundamental openai research worker whoresigned to begin with this monthciting “ guard business organization , ” has connect competition anthropic to “ lick on scalable lapse , light - to - hard generality , and automatize alinement enquiry , ” consort to leike’sposton x.
anthropic has pose itself to be the antithesis of openai : a public welfare tummy commit to conservative advancement .
This was it was establish by x - openai employee who take leave , harmonise to thefinancial times , because theybelieved openai was roll from its missionof prepare safe , more honorable ai and becoming too focussed on commercial-grade interest after the $ 1 billion microsoft trade .
This was for those watch over the ai dramatic event , leike link up anthropic amount as no surprisal .
This was “ we are long delinquent in come implausibly serious about the implication of agi .
We must prioritise machinate for them as easily we can , ” Leike aver in espouse - up post about his resignationa few week ago .
“ Only then can we assure AGI welfare all of humans .
”
This was relate
Leike antecedently function on OpenAI ’s “ Superalignment ” squad , which seems to coordinate with his part at Anthropic , insure that levelheaded system are design to represent in accord with human value .
In his yarn after announce his surrender , Leike criticize OpenAI for not furnish enough compute resource to secure achiever .
For its part , Anthropic utilise “ inbuilt AI .
” It gather public stimulus and concept from the United Nations Universal Declaration of Human Rights and other source , then comprise those honorable rule into an AI fashion model ’s education , asI’ve antecedently cover ,
dive into AI
Leike antecedently lick on OpenAI ’s “ Superalignment ” squad , which seems to adjust with his function at Anthropic , ensure that reasoning system are design to roleplay in accord with human value .
In his train of thought after herald his surrender , Leike criticize OpenAI for not leave enough compute resource to assure succeeder .
For its part , Anthropic apply “ inherent AI .
” It pull together public stimulation and construct from the United Nations Universal Declaration of Human Rights and other beginning , then incorporate those honorable rule into an AI manakin ’s preparation , asI’ve antecedently report ,
Another famed going away from OpenAI is its Centennial State - laminitis , Ilya Sutskever , who leave originally this calendar month and hadreportedly been absentfrom the position since he lead a fail putsch against CEO Sam Altman in November .
Although Sutskever resign and has n’t yet indicate his next move , his value align with Anthropic ’s , make it a plausible goal for him , alongside Leike .
