Jacob Cockson, a researcher formerly with Anthropic, has stepped away from the AI sector, citing safety apprehensions. On Tuesday, he revealed that while his expertise lay in training novel AI models using vast datasets, he was reluctant to engage in the industry's competitive sprint to create AI systems capable of self-improvement. Cockson expressed concerns that such systems might spiral out of control, posing a threat to humanity. This development underscores the escalating safety worries within prominent AI firms.
