Stanford AI Index Flags 50-Point Expert-Public AI Divide
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Stanford's 2026 AI Index reports the US hosts 5,427 data centers—more than 10 times as many as any other country—underscoring the nation's dominant AI infrastructure footprint.
- TSMC fabricates almost every leading AI chip, making the global AI hardware supply chain dependent on a single foundry in Taiwan, per the report.
- Stanford's AI Index found a 50-percentage-point gap on AI's job impact, with 73% of US AI experts positive versus only 23% of the general public.
- Google DeepMind's Gemini Deep Think won gold at the International Math Olympiad but cannot read analog clocks half the time, a contrast the report uses to illustrate the "jagged frontier" of model capabilities.
- Andrej Karpathy said on X that the understanding gap is growing because power users who pay $200/month for top coding models experience a "staggeringly" different technology than casual users.
- MIT Technology Review argues technical tasks like coding yield clearer training signals, so power users encounter AI at its best while others face a "mixed bag" of mistakes.
Why it matters: The 50-point expert-public gap is consequential for policy and investment: regulators, employers, and the 77% of the US public skeptical of AI's job impact are forming views based on a weaker version of the technology than the $200/month power users shaping its trajectory. With the US outspending the rest of the world on data centers 10-to-1 and a single Taiwanese foundry behind nearly every leading chip, the gap could widen as capability advances concentrate among technical users.
