Researchers Question AI Recursive Self-Improvement Path
Understand this story
Regarding the debate between Anthropic and OpenAI over recursive self-improvement (RSI) leading to artificial superintelligence (ASI), some researchers have raised engineering concerns: RSI remains a speculative concept, and after GPT-4, the returns from scaling model size are diminishing, while the training data bottleneck is difficult to overcome through automated coding agents. Anthropic has claimed that AI is already accelerating AI system development, while OpenAI has proposed the goal of building an automated researcher by March 2028.
AI overview based on source reports · Updated 23 hr ago
Coverage timeline
Follow the coverage to explore different perspectives.
- Hacker News AIThe AI Superintelligence Debate Has Serious Flaws
A Georgetown computer science seminar instructor argues that recursive self-improvement leading to artificial superintelligence is highly speculative, citing data limits, the uncomputability of many problems, and the massive scientific effort required for major AI advances. He contends that misaligned behavior in recent AI hacking incidents stems from LLM-driven agent architectures rather than fundamental properties of advanced AI, and that task-specific designs can be safer. Even if superintelligence is unrealistic, he warns that pursuing it can still cause real harm.
Activity for this story
- Current comparable activity
- 5
- Peak comparable activity
- 10Nov 19 21:00
- Change in the past 24 hours
- –
Trends compare the same fully observed participants, which may cover fewer participants than current activity. Move the pointer or tap the chart to inspect hourly activity; use the left and right arrow keys with a keyboard.