Karpathy Says the LLM Capability Gap Is Becoming a Funnel

What happened

Andrej Karpathy argues the gap in shared understanding of LLM capability is widening into a sharp funnel: roughly 6B people have barely touched LLMs, 1-2B use free-tier chatbots casually, about 20M use frontier models for math and code, and roughly 5,000 have internal access to frontier systems. He attributes the funnel to problem ambition, capability jaggedness in verifiable domains, and access tiers.

Reported by 2 sources

The story so far

Andrej Karpathy argued that people’s understanding of LLM capabilities is diverging into a “sharp funnel,” and Ethan Mollick responded that even many people with internal access at AI labs have not used advanced models to their full extent. Their exchange describes differences in both access to powerful systems and the range of problems people use them to tackle. Karpathy estimated that roughly 6 billion people have barely encountered LLMs, 1–2 billion casually use free-tier ChatGPT-like products, about 20 million see frontier-model capabilities in math and code, and roughly 5,000 have internal access to frontier systems. He attributed the divide to the ambition of users’ problems, uneven capabilities across tasks, and access tiers. Mollick challenged the idea that access alone means broad use: he said many of those roughly 5,000 people at labs still have not used advanced models to their full extent, with use concentrated in cybersecurity, math, and AI research and development. He argued that many other fields and problems remain to be addressed.

Keep reading

Trending now

1. 1Meta Releases Muse Spark Open With Mac and Cloud Support, Topping US Open-Weight Index
2. 2Google Launches Gemini 4 Argon Globally, Expands Paid Access as Benchmark Results Vary
3. 3OpenAI Updates GPT-6 Astra Pro in ChatGPT With Parallel Agents and Faster Responses
4. 4Meta’s New Business Terms Raise Dispute Over Website Crawling and AI Training Rights