Anthropic co-founder Jared Kaplan warns AI may design its own successor by 2030, risking control.
digital
S
Storyboard•03/12/2025
Anthropic co-founder Jared Kaplan warns AI may design its own successor by 2030, risking control.
- •Anthropic co-founder Jared Kaplan warns AI could design its own successor by 2030.
- •Kaplan believes AI self-improvement could start between 2027 and 2030, posing profound risks.
- •Artificial general intelligence (AGI) remains poorly defined, with labs racing to achieve AGI-level models.
- •Kaplan is optimistic about aligning AI with human values until it surpasses human intelligence.
- •Two major risks: loss of human agency and rapid AI self-improvement potentially outpacing human oversight.
Why it matters: AI self-design could outpace human control, risking unpredictable and irreversible consequences.
✦
Related Articles
Loading more articles...
