More effort spent on early commercialization like keeping ChatGPT running might mean less effort on cutting edge capabilities. Altman was never an AI safety person, so my personal hope is that Anthropic avoids this by having higher quality leadership.