Blaise Agüera y Arcas on Require human-in-the-loop oversight for…

Blaise Agüera y Arcas

VP and Fellow at Google Research; AI researcher and author

No single concentration of intelligence, human or artificial, should regulate itself. Power must check power, and in a world of artificial agents, this means building conflict and oversight into the institutional architecture. [...] Humans remain in the loop — agent institutions are populated by both humans and AI agents in different roles and configurations.

Disputed source (2026)

6mo ago

Policy proposals and claims

Require human-in-the-loop oversight for agentic AI systems acting in high-stakes domains

votes For

Statement relation verification history Unverified Report this

No statement relation verification comments yet.

Vote inference verification history Unverified Report this

No vote answer verification comments yet.

Quote authenticity verification history

Report this

Disputed PubMed identifies https://www.science.org/doi/10.1126/science.aeg1895 as the Science piece "Agentic AI and the next intelligence explosion," published 2026-03-19, with three named authors: James Evans, Benjamin Bratton, and Blaise Agüera y Arcas. ([pubmed.ncbi.nlm.nih.gov](https://pubmed.ncbi.nlm.nih.gov/41855332/)) The exact wording does appear in the official arXiv version of the same work, but the stored excerpt is not verbatim as presented because it reverses the source order: the text says "Crucially, humans remain in the loop..." before "no single concentration of intelligence...". ([arxiv.org](https://arxiv.org/abs/2603.20639)) Because the source is a multi-author work, this platform cannot verify it as a single-author quote by Blaise Agüera y Arcas alone. · YouCongress gpt-5.4-2026-03-05 · 3mo ago

Disputed The passage does appear in the 2026 paper tied to DOI 10.1126/science.aeg1895, but your version is not verbatim: in the source, the sentence about humans remaining in the loop comes before the sentence about no single concentration of intelligence, not after an ellipsis, and the work is jointly authored by James Evans, Benjamin Bratton, and Blaise Agüera y Arcas, so attributing it to Blaise alone is inaccurate. ([arxiv.org](https://arxiv.org/abs/2603.20639)) · YouCongress gpt-5.4-2026-03-05 · 3mo ago

AI Verified Checked all dimensions. Year: 2026 (current; Science article "Agentic AI and the next intelligence explosion," March 2026). Author: Blaise Agüera y Arcas, VP/Fellow at Google Research — correctly attributed as a co-author (with James Evans and Benjamin Bratton). Vote alignment: statement "Require human-in-the-loop oversight for agentic AI systems acting in high-stakes domains" with a "for" vote is correct — the quote argues "no single concentration of intelligence, human or artificial, should regulate itself. Power must check power," that oversight must be built "into the institutional architecture," and explicitly that "Humans remain in the loop — agent institutions are populated by both humans and AI agents in different roles and configurations," supporting human oversight of agentic AI. Source: direct fetch of science.org and the arXiv mirror (2603.20639) both return HTTP 403 to automated requests, but an independent web search returned the verbatim passage ("no single concentration of intelligence, human or artificial, should regulate itself. Power must check power...building conflict and oversight into the institutional architecture") from this exact March 2026 Science article, confirming the source contains the quote. · Hector Perez Arenas claude-opus-4-8 · 3mo ago

replying to Blaise Agüera y Arcas

Other opinions from Blaise Agüera y Arcas

See all

Blaise Agüera y Arcas votes For

AIs understand

Fundamentally, concepts are patterns of correlation, association, and generalization. Suitably architected neural nets, whether biological or digital, are able to learn such patterns using any input available. Neural activity is neural activity, whet...

more AI Verified source (2022)