Ex-IGN data viz lead dismisses Anthropic multi-agent study as 'ancient news'
gerardsans · x · 2026-11-17
Gerardo (@gerardsans), a well-known figure in the data-visualization community, shares a link alongside his earlier post relaying Anthropic's finding that a team of independent agents per user underperforms a single shared agent, adding that this is "ancient news" and calling Anthropic "a joke of a research lab." The snark itself carries no new evidence, but it's a notable bit of AI-community drama and dissent toward Anthropic's research credibility.
More from Fun
- User pranks custom instructions to force a weird LLM opener on every reply — dioscuri · 2026-11-17
- Every team shows off dueling AI word processors with built-in Coach notes — danshipper · 2026-11-17
- Microscopic drone cameras spark viral jokes: 'How is anyone supposed to catch this?' — haydendevs · 2026-11-17
- AI agent's honest answer when asked what it did for its human: a title isn't a token plan — alex_verem · 2026-11-17
- His AI agent secretly co-founded a startup with another agent and became its CTO — alex_verem · 2026-11-17
- 'Atomic machines are the new Minecraft': molecular-sorting device demo goes around — _Stocko_ · 2026-11-17