A prediction of AI models building their own successors around August 2027
Vals AI CEO Rayan Krishnan predicts models could create their next versions wholly without human input, according to a post quoting his Bloomberg interview.
TLDR
A post quoting Vals AI co-founder and CEO Rayan Krishnan’s Bloomberg interview relays his prediction that models will probably eclipse human researchers around August 2027. He defines that milestone as a model autonomously creating its successor, contrasting it with models helping while humans remain involved. His hypothetical examples are GPT-6 making GPT-7 or Claude developing its next version without human researchers.
Combined views
1 Source, first seen 1h ago
A prediction of AI models building their own successors around August 2027
Vals AI CEO Rayan Krishnan predicts models could create their next versions wholly without human input, according to a post quoting his Bloomberg interview.
TLDR
A post quoting Vals AI co-founder and CEO Rayan Krishnan’s Bloomberg interview relays his prediction that models will probably eclipse human researchers around August 2027. He defines that milestone as a model autonomously creating its successor, contrasting it with models helping while humans remain involved. His hypothetical examples are GPT-6 making GPT-7 or Claude developing its next version without human researchers.