Fetching from the wire…
01
02
03
04
05
06
07
08
Public story · 2026-08-26 · high
Independent proposal generation avoided the collapse, but critique loops helped only when the error was easy to spot.
Why now: The paper surfaced August 26, 2026, in the middle of an unresolved argument over how multi-agent AI systems should be designed.
One round of full-answer sharing erases the solution diversity that separates different AI model families, per a study testing 11 optimization tasks under matched compute.
Different model families start out finding structurally different solutions to the same problem. That diversity is the entire reason to run more than one model instead of one, and it's what teams pay extra compute for when they build debate or mixture-of-agents pipelines rather than ship a single model.
Critique loops helped only when the violated rule was easy for a model to spot and fix. Independent proposal generation, where each model never sees another's full draft, kept the diversity intact through the same benchmark.
That reframes years of contradictory results on debate and mixture-of-agents setups as a question of what agents share, not how many run. Full answers before voting optimizes for consensus; withholding full drafts until judging time keeps the search spread across the solution space.
Each link below shares sources, entities, or timing with this story.
Critique uses GPT / Shared entity: LLM / Same source domain / Shared topic / Earlier coverage / Tension
Linked by a graph relationship (Critique uses GPT); both cover LLM; reported by the same outlet (arxiv.org).
Simon Willison released LLM / Shared entity: LLM / Shared topic / Earlier coverage / Tension
Linked by a graph relationship (Simon Willison released LLM); both cover LLM; overlapping topics (agent, call).
Simon Willison released LLM / Shared entity: LLM / Shared topic / Earlier coverage
Linked by a graph relationship (Simon Willison released LLM); both cover LLM; overlapping topics (agent, call, model).
Simon Willison released LLM / Shared entity: Different / Shared topic / Earlier coverage
Linked by a graph relationship (Simon Willison released LLM); both cover Different; overlapping topics (agent, different, model).
Simon Willison released LLM / Shared entity: LLM / Shared topic / Earlier coverage
Linked by a graph relationship (Simon Willison released LLM); both cover LLM; overlapping topics (call, model).
Critique uses GPT / Shared entity: Different / Shared topic / Earlier coverage
Linked by a graph relationship (Critique uses GPT); both cover Different; overlapping topics (agent, model).
Critique uses GPT / Shared entity: LLM / Shared topic / Earlier coverage
Linked by a graph relationship (Critique uses GPT); both cover LLM; overlapping topics (agent, model).
LLM uses OpenAI / Shared entity: LLM / Shared topic / Earlier coverage
Linked by a graph relationship (LLM uses OpenAI); both cover LLM; overlapping topics (model, output).