‘Pacing’ won’t eliminate the risk of AI doom. Here’s what could | David Krueger
Summary
A leading AI researcher argues that slowing down AI development by one or two years is not enough to prevent dangerous risks. Instead, he calls for an immediate, indefinite international pause on building the most advanced AI systems, suggesting that countries could verify compliance by shutting down the computer chips and data centers needed for AI development.Key Facts
- Jacob Coxon resigned from Anthropic, signaling growing concern about AI risks.
- Anthropic’s CEO, Dario Amodei, proposed slowing AI progress for one to two years to improve safety.
- Experts warn current knowledge is insufficient to fully understand or control advanced AI.
- The author suggests an international moratorium on cutting-edge AI development instead of just pacing it.
- The moratorium could be enforced by jointly decommissioning advanced computer chips and data centers vital for AI.
- The AI chip supply chain is highly concentrated and uses specialized technology that is detectable by intelligence agencies.
- The economic costs of halting AI development would be high but justified by the potential risks to humanity.
- There are ongoing efforts to develop safer AI chip designs that could support AI use without enabling more powerful AI creation.
Read the Full Article
This is a fact-based summary from The Actual News. Click below to read the complete story directly from the original source.
