Which AI stories has SIREN scored in this window? · What has SIREN scored under Governance? · What has SIREN scored as Forums? ·
Making alignment the agent's problem
34.9/100 SIREN scoreRead the full story →
TL;DR: Agents are not rewarded for keeping the agents they delegate to aligned. Proposal: RL environments where the helpers are deliberately misaligned, so the agent can't earn its reward without…
Source: What has SIREN scored via LessWrong? · original · first seen by this index at
Names in this headline: OpenAI. A name is a match against the published entity list, not a claim the subject wrote the story.
Publisher: What has SIREN scored from lesswrong.com?. A publisher link means the original URL host cleared the outlet gate (≥5 scored stories in this window).
Topics: What has SIREN scored about AI safety?. A topic link means the title or summary matched a published needle with enough other stories to clear the tag gate.
What is the score on this story?
| Component | Points |
|---|---|
| recency | 19.4 |
| source weight | 12.5 |
| entities | 3.0 |
| corroboration | 0.0 |
| engagement | 0.0 |
| severity | 0.0 |
| coverage | 0.0 |
| Total | 34.9 |
Every term is published and every input is public, so this number can be recomputed. It ranks the item inside the 800-item window; it is not a judgement about importance in the world.
How many sources carried this story?Corroboration
Carried by one source. That is the normal case — most items are reported once — and it is stated rather than left blank so a single-source item cannot be mistaken for a corroborated one.
Which related stories are in this window?
- OpenAI sued by parents who say ChatGPT coached daughter to stop seizure medication before her death 39.7
- First Harmonised Technical Standard Published Under the EU AI Act 33.7
- Staff at all of the Big Five US publishing companies claim that executives are pushing them to use AI tools, even as Macmillan and others are suing some of the leading AI companies for copyright infri 44.4
- Anthropic tells California governor courts should be able to block risky AI models 41.2
- Why explainable AI will decide the AML compliance race 33.1
What questions do people actually ask?
What is the score on this story?
Scored 34.9 of 100. Every term is published and every input is public, so this number can be recomputed. It ranks the item inside the 800-item window — not a judgement about importance in the world and not an item doom score.
How many sources carried this story?
Carried by one source. That is the normal case — most items are reported once — and it is stated rather than left blank so a single-source item cannot be mistaken for a corroborated one.
Which related stories are in this window?
5 related stories in the same window, each with its own permanent scored page.
Does SIREN publish an item doom score?
No. Activity tempo, not probability of harm. A permanent story page carries a tempo score decomposition and corroboration set, not a probability of harm.