Fetching from the wire…
01
02
03
04
05
06
07
08
Markets2026-08-12 · source-backed
Material Discovery Bench measures LLM progress on discovering thermally conductive dielectrics for 3D chip packaging, built on the AI Security Institute's open-source Inspect framework, with models given web search, Python/bash sandboxes, and property-computation ML tools (Discovered Materials). Seven models, runs of 30–100M tokens each, 500+ previously unknown materials computationally discovered, one with a plausible synthesis pathway worth attempting. GPT-5.6 Sol led at 4.0 favorable materials per run. Best anti-hype datapoint of the week: generation volume and validated output are separated by roughly three orders of magnitude. Anyone building a vertical R&D agent should put this number on a wall.
Each link below shares sources, entities, or timing with this story.
Simon Willison released LLM / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (Simon Willison released LLM); both cover GPT, LLM; overlapping topics (agent, model).
Linked by a graph relationship (Simon Willison released LLM); both cover GPT, Python; overlapping topics (agent, anyone).
Simon Willison released LLM / Shared entity: GPT / Shared topic / Earlier coverage / Tension
Linked by a graph relationship (Simon Willison released LLM); both cover GPT; overlapping topics (agent, model).
GPT competes with Claude / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (GPT competes with Claude); both cover Best, GPT; overlapping topics (agent, model).
Simon Willison released LLM / Shared entity: GPT / Shared topic / Earlier coverage
Linked by a graph relationship (Simon Willison released LLM); both cover GPT; overlapping topics (agent, benchmark, model).
GPT competes with Claude / Shared entity: GPT / Shared topic / Earlier coverage / Tension
Linked by a graph relationship (GPT competes with Claude); both cover GPT; overlapping topics (agent, benchmark, best).
GPT competes with Claude / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (GPT competes with Claude); both cover GPT, LLM; overlapping topics (agent, model).
Simon Willison released LLM / Shared entities / Earlier coverage
Linked by a graph relationship (Simon Willison released LLM); both cover LLM, Python; earlier LLM coverage from 2026-06-22.