The conversation around the evidence
What does the research add, miss or leave unresolved? Share an experience, evidence link or question on any article. No account is required; every note is reviewed before it appears here.
Reader views are their own. Publication does not verify a claim or mean the newsroom endorses it. Disagreement is welcome; personal attacks, spam and confidential information are not.
The conversation is just starting. There are no approved reader notes yet; choose an article below to contribute.
Government & Policy
England and Wales' judiciary says approved AI may support tasks such as summarisation, transcription, redaction and document analysis, but every output used in judicial work must be reviewed and approved by the responsible judge. The position sets governance boundaries; it does not publish pilot results or authorise autonomous decisions.
Read and comment →AI Risks & Safety
A Meta-led preprint found harmful outcomes in 192 of 1,000 simulated agent scenarios generated by its new red-teaming method. The result exposes cross-service risk, but it comes from one synthetic environment and model-based judges—not live consumer accounts.
Read and comment →Security & Defence
A China-based preprint tested 15 frontier models on 30 mutation-design scenarios against 187,695 existing laboratory measurements. Tool-assisted iteration shifted selected variants further into high-risk assay regions, but the benchmark covers single substitutions—not end-to-end pathogen creation.
Read and comment →Education
A Hong Kong preprint found that a structured tutoring harness improved evidence retrieval by 13.2 percentage points and learner diagnosis by 11.6 points on average across nine model configurations. It did not test real students, retention or learning gains.
Read and comment →Society & Media
Not reliably. On 200 native Ukrainian memes, the best micro hierarchical F1 was 0.384 and macro F1 stayed below 0.20. Models often over-predicted persuasion, while explanations could agree on the same words and still assign the wrong rhetorical label.
Read and comment →Health & Life Sciences
Not reliably. In a peer-reviewed benchmark of 161 image-based clinical cases, the strongest reasoning configuration reached 46.0% diagnostic accuracy. Adding web search cut the simplest configuration from 38.5% to 26.7%, while agent and deep-research modes also trailed that baseline.
Read and comment →