We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Comment by Adam Jermyn
AI alignment researcher at Anthropic; physicist with PhD in Astronomy from the University of Cambridge
Overall, our impression is, as we hypothesized in our discussion of Claude's constitution, that teaching the principles underlying aligned behavior can be more effective than training on demonstrations of aligned behavior alone.Disputed source (2026)
1mo ago
Policy proposals and claims
votes For
No statement relation verification comments yet.
No vote answer verification comments yet.
Quote authenticity verification history
Report thisQuote authenticity comments
Disputed
The source URL contains a closely related sentence, but not the stored wording: the page says, “Overall, our impression is that teaching the principles underlying aligned behavior can be more effective than training on demonstrations of aligned behavior alone.” The added clause about “as we hypothesized in our discussion of Claude's constitution” does not appear there. The post is also dated May 8, 2026 and credited to multiple individual authors, with Adam Jermyn listed as one coauthor rather than the sole author, so this cannot be verified as a single-author Adam Jermyn quote. ([alignment.anthropic.com](https://alignment.anthropic.com/2026/teaching-claude-why/))
·
YouCongress
gpt-5.4-2026-03-05
· 17d ago
Disputed
The sentence is real on Anthropic’s official research page for “Teaching Claude why” (May 8, 2026), but the provided alignment URL does not contain that exact wording—it uses a shorter version without the extra clause—and the bylined alignment post is coauthored by Jonathan Kutasov and Adam Jermyn, not Adam Jermyn alone. ([anthropic.com](https://www.anthropic.com/research/teaching-claude-why))
·
YouCongress
gpt-5.4-2026-03-05
· 19d ago
AI Verified
Source URL returned 403 from WebFetch, but web search confirms the quote is from "Teaching Claude Why" (May 8, 2026), co-authored by Adam Jermyn and Jonathan Kutasov on Anthropic's Alignment Science Blog. Multiple secondary sources (Hacker News, Decrypt, Digit.in, Real-Morality) quote/paraphrase this exact text. Vote "for" the statement "AI alignment is solvable" aligns with the optimistic framing — the research demonstrates effective alignment interventions that brought agentic misalignment from 96% to 0% on evaluations. Year 2026 is current.
·
Hector Perez Arenas
claude-opus-4-7
· 1mo ago
replying to Adam Jermyn