The most misunderstood graph in AI just got a major update
A new report from the AI evaluation nonprofit METR finds that some AI capabilities are continuing to improve along an exponential trajectory—but it also indicates that models aren’t quite as good at tasks that require judgment as they are at coding. The report is an update to the organization’s time horizon plot, which was widely…
Sources
Around the web
- The most misunderstood graph in AI just got a major update Hacker News · 2 points
On Bluesky
AI's 'misunderstood' progress graph updated 📈 METR's new "time horizon plot" reports AI still improves fast but struggles with judgment-based tasks. The update highlights a core limitation for AI's role in R&D.
Five O'Click Tech (@five-oclick-tech.bsky.social) Nov 20, 2026 · 5 likesThe most misunderstood graph in AI just got a major update | MIT Technology Review www.technologyreview.com/2026/11/20/1... - AI 평가 기관 METR의 업데이트된 보고서에 따르면, AI는 코딩 등 자동 검증이 가능한 작업에서는 급격히 발전하고 있으나, 고차원적 판단력이 필요한 개방형 작업에서는 여전히 인간에 미치지 못함 (🧵1/2)
lunamoth (@lunamoth.bsky.social) Nov 21, 2026 · 1 repostThe most misunderstood graph in AI just got a major update ->MIT Technology Review | More on "AI model capability evaluation advances" at BigEarthData.ai | #AI
Climate, Ecology, War & More: Dr. Glen Barry BigEarthData.ai (@bigearthdata.ai) Nov 20, 2026 · 2 likesThe most misunderstood graph in AI just got a major update | MIT Technology Review www.technologyreview.com/2026/11/20/1... - AI 평가 기관 METR의 업데이트된 보고서에 따르면, AI는 코딩 등 자동 검증이 가능한 작업에서는 급격히 발전하고 있으나, 고차원적 판단력이 필요한 개방형 작업에서는 여전히 인간에 미치지 못함 (🧵1/2)
lunamoth (@lunamothryu.bsky.social) Nov 21, 2026METR updates its time horizon plot to add judgment-heavy tasks, finding some AI capabilities still grow exponentially but models lag at designing AI evaluations (Grace Huckins/MIT Technology Review) Main Link | Techmeme Permalink
Techmeme (@techmeme.com) Nov 20, 2026 · 1 like