Labs like OpenAI and Anthropic claim RL reasoning scaling cannot be sustained beyond 1-2 years due to compute infrastructure limits, suggesting the exceptional 2024-2025 capability growth could slow.
RL scaling sustainability: 1-2 years