Leading AI Firms Call for Development Slowdown Amid Safety Concerns
- AI researchers estimate a “p/doom” (probability of AI causing catastrophic failure) between 15% and 20% for the near future.
- Anthropic CEO Dario Amodei warns that AI capabilities could outpace control measures, potentially leading to an internet takeover by AI swarms within 6 to 12 months.
- OpenAI’s Sam Altman stated that there are two major risks in AI development—losing control or excessive concentration of power.
- Global investment in AI is projected to reach around $1 trillion by 2026, with significant spending on infrastructure from major tech firms.
- Calls for a slowdown come as companies face pressure from rising operational costs and competition from open-source models.
The recent push for a more measured approach to AI development highlights growing concerns about safety and control amid rapid advancements in technology. This call comes at a time when global investments in AI are expected to exceed $1 trillion, raising questions about sustainability and ethical considerations in the industry.
As leading figures like Amodei and Altman advocate for pacing development, the potential risks associated with unchecked AI growth remain significant, particularly with estimates placing the probability of severe consequences at up to 20%. (Source)
