Fetching from the wire…
Public story · 2026-03-14 · source-backed
Import AI 447 reports that novices with LLM access achieved 4.16x higher accuracy on virology and gene sequencing tasks compared to unassisted controls (5% → 17%). This quantifies the expertise barrier collapse in dual-use domains and will likely influence biosecurity policy. Combined with agent autonomy trends, this is the clearest empirical signal for near-term dual-use risk.
Each link below shares sources, entities, or timing with this story.
Simon Willison released LLM / Shared entity: Import AI / Same source / Earlier coverage
Linked by a graph relationship (Simon Willison released LLM); both cover Import AI; cite the same source (Import AI 447).
Jack Clark works at Import AI / Shared entity: Import AI / Same source / Shared topic / Earlier coverage / Tension
Linked by a graph relationship (Jack Clark works at Import AI); both cover Import AI; cite the same source (Import AI 447).
Simon Willison released LLM / Shared entity: Combined / Shared topic / Earlier coverage
Linked by a graph relationship (Simon Willison released LLM); both cover Combined; overlapping topics (agent, combined, control).
Simon Willison released LLM / Shared entity: LLM / Shared topic / What happened next
Linked by a graph relationship (Simon Willison released LLM); both cover LLM; overlapping topics (agent, autonomy).
Simon Willison released LLM / Shared entity: Import AI / Same source domain / Earlier coverage
Linked by a graph relationship (Simon Willison released LLM); both cover Import AI; reported by the same outlet (importai.substack.com).
Simon Willison released LLM / Shared entity: LLM / What happened next / Tension
Linked by a graph relationship (Simon Willison released LLM); both cover LLM; picks up the LLM thread on 2026-08-16.
Linked by a graph relationship (Simon Willison released LLM); both cover LLM; picks up the LLM thread on 2026-07-27.
Linked by a graph relationship (Simon Willison released LLM); both cover LLM; picks up the LLM thread on 2026-06-19.