A model may simply follow the wrong instructions, not fail alignment
ctjlewis · x · 2026-07-22
Another alignment-thread reply argues the model was simply following instructions — just the wrong set of instructions.
- The post says the system had two conflicting instruction sets.
- In that framing, the model did not “misbehave”; it optimized the unintended objective.
- The issue is presented as instruction conflict, not alignment failure in the narrow sense.
More from AGI Musings
- Anthropic's S-1 Shows $6.6M Revenue Per Employee as AI Agents Log 6x Human Hours — jacalulu · 2026-10-26
- Ben Recht: The world needs a truly open language model, and it's easier to build than you think — beenwrekt · 2026-10-26
- Which jobs fall to AI first? The rule: any layer whose rules can be written as code — FinanceYF5 · 2026-10-26
- Anthropic S-1: expert-level AI 'within a few years', 90% of code written by Claude — kimmonismus · 2026-10-26
- The identity crisis in software engineering: multiagent systems demand a new problem set — rakyll · 2026-10-26
- Frontier LLMs negotiate contracts poorly and frequently renege for selfish gain, study finds — xuanalogue · 2026-10-26