Cryptographer Matthew Green: expect AI surprises, models enter new territory by next October
matthew_d_green · x · 2026-10-12
Matthew Green, cryptography professor at Johns Hopkins, argues that people assuming AI progress will be unsurprising haven't been paying attention to this year's surprises. The only certainty is that things will be surprising, and he can't trust anyone confident about what will happen. In a follow-up he adds that we're already in new territory and will be in very new territory by next October, so humility about current model capabilities is warranted.
Related event: Cryptographer: Assume AI Keeps Improving at This Year's Pace(3 posts)→
More from AGI Musings
- willcb: every non-RL algorithm is a proxy — policy gradients are the minimum-variance unbiased estimator — willcb · 2026-10-12
- Is AI Agent just a wrapper? Users don't care about the plumbing, only the chore done — sujingshen · 2026-10-12
- Researcher Revives Sci-Fi Short Story About a Model 'Waking Up' in Pain During RL Training — voooooogel · 2026-10-12
- AI slop now floods every public platform, from X to Reddit — private communities are the last stand — ayushtweetshere · 2026-10-12
- AIandDesign: everyone working 40 hours a week on a computer is 'completely cooked' — AIandDesign · 2026-10-12
- Access to data isn't understanding: why 'AI knows your business' is overblown — Rajxai · 2026-10-12