You are viewing a single comment's thread from:

RE: LeoThread 2025-11-18 01-34

in LeoFinance10 days ago

Part 6/19:

Google, OpenAI, and Anthropic are scaling their models, with scaling believed to be directly linked to self-improvement capabilities. Some experts argue that recursive self-improvement loops, especially via code generation, could dramatically accelerate progress toward superintelligence. Others caution that grounding, visual reasoning, and physical interaction might be essential components missing from current models, potentially limiting their ultimate capabilities.

The Business of AI