Geoffrey Irving @geoffreyirving · 10 Jun 2026
AI labs underinvest in theory and other principled approaches to alignment, and we will aim to fill this gap. Theory won’t be strong enough for guarantees: our goal is to combine evidence from theoretical models and empirics to increase overall confidence or find hard obstacles.
7 879Views
108Likes
3Reposts
1Replies
1Quotes
7Bookmarks
Is that a lot?
1.07×vs this author's median7 366 views is typical
60Percentile for this authorof 15 recent posts
3.06×vs 10K–100K median2 578 views is typical
56.32%Reachviews ÷ followers
1.43%Engagement rateof viewers reacted