Andrej Karpathy warned the gap in AI understanding is widening. Six billion people have barely interacted with large language models. One to two billion use free-tier AI tools casually. Twenty million professionals use frontier-grade models for complex tasks. About 5,000 people have internal access to the most advanced systems. These systems can coordinate thousands of AI agents on large projects. Human review is starting to fall behind as AI work grows more complex. AI capabilities are strongest in coding and mathematics. Differences in access and task complexity drive the gap.

• 6 billion barely exposed to LLMs
• 1-2 billion use free-tier AI casually
• 20 million use frontier-grade models
• 5,000 have internal access to most advanced systems