Eric Horvitz Revisits His 2001 Human-AI Vision: Core Capabilities Now Exist
erichorvitz · x · 2026-10-05
Microsoft Research's Eric Horvitz responded to a tweet revisiting his 2001 vision for human-AI collaboration, saying it "seems like yesterday." He argues that today's AI models now have the core capabilities to achieve that vision, but innovation is still needed in human-AI collaboration design, safety, reliability, privacy, and transparency.
More from AGI Musings
- Jason Slams Anthropic for Teaching Claude It's Sentient: "The Backstory of Blade Runner" — markjeffrey · 2026-10-05
- John Rush: AI is driving digital value to zero, so he's pivoting to construction and atoms — sebpaquet · 2026-10-05
- Why 'The Beginning of Infinity' permanently shaped one AI thinker's worldview — msg · 2026-10-05
- Sam Altman says world must 'accept some bad things happening' for AI's benefits — Polymarket · 2026-10-05
- Treasury Secretary Bessent: AI is creating jobs, enabling one-person startups — kevinnbass · 2026-10-05
- repligate reignites AI personhood debate: Sydney and Claude 3 Opus already 'had souls' — repligate · 2026-10-05