Altman Flags Two Ways AI Progress Could Go Badly Amid Debate

OpenAI CEO Sam Altman has flagged two ways in which AI progress could go badly: humans losing control of increasingly powerful AI and the concentration of power in the hands of one company, person, or country. Altman advocates for a "narrow middle path" of careful development, acknowledging that safety measures will make development slower and more expensive, but justified.