Why opinion on AI is so divided

Q: Stanford's AI Index reveals massive 50-point perception gap between AI experts (73% positive on jobs

Stanford's AI Index reveals massive 50-point perception gap between AI experts (73% positive on jobs) and general public (23% positive)

Q: TSMC fabricates almost every leading AI chip, creating single-point-of-failure dependency for global

TSMC fabricates almost every leading AI chip, creating single-point-of-failure dependency for global AI hardware supply chain

Q: AI's "jagged frontier" means Claude Code excels at technical tasks while Gemini Deep Think can't rea

AI's "jagged frontier" means Claude Code excels at technical tasks while Gemini Deep Think can't read analog clocks despite Math Olympiad gold

The Download from MIT Technology Review·Monday, April 13, 2026·6 min read

AI/ML Technology

Share𝕏 in

AI Summary

MIT Technology Review's AI newsletter analyzes Stanford's 2026 AI Index, highlighting a massive perception gap between AI experts (73% positive on jobs) and the general public (23% positive). The report reveals AI's "jagged frontier" - models excel at technical tasks like coding but struggle with basic tasks like reading analog clocks, creating divided opinions based on user experience.

Key Facts

✓Stanford's AI Index reveals massive 50-point perception gap between AI experts (73% positive on jobs) and general public (23% positive)

✓TSMC fabricates almost every leading AI chip, creating single-point-of-failure dependency for global AI hardware supply chain

✓AI's "jagged frontier" means Claude Code excels at technical tasks while Gemini Deep Think can't read analog clocks despite Math Olympiad gold

Author Takes

NeutralThe Algorithm from MIT Technology Review

AI capability perception

There are two realities - AI is better than many realize but still bad at stuff people care about

More from The Download from MIT Technology Review

A new US phone network for Christians aims to block porn and gender-related content

MIT Technology Review's The Download covers a new Christian-marketed US phone network launching with mandatory porn blocks and optional gender-content

May 1

Elon Musk and Sam Altman are going to court over OpenAI’s future

Elon Musk and Sam Altman head to trial this week over OpenAI's future, with Musk seeking $134 billion in damages and a reversion to non-profit status

Apr 28

We’re in a new era of AI-driven scams

This MIT Technology Review newsletter covers the rise of AI-powered cybercrime, including phishing, deepfakes, and automated vulnerability scans, whic

Apr 24