People who build AI

Timnit Gebru

founder of the Distributed AI Research Institute

2 records

Records

Why, then, is building AGI often framed as an unquestioned goal in the field of AI? In this paper, we argue that the normative framework that motivates much of this goal is rooted in the Anglo-American eugenics tradition of the twentieth century.

Exact quote, beginning

Timnit Gebru

founder of the Distributed AI Research Institute

In a 2024 paper, Timnit Gebru and Émile P. Torres wrote that efforts to build artificial general intelligence harm marginalized groups, centralize power, and use safety language to evade accountability.

Coauthor of The TESCREAL bundlePublished April 14, 2024 · date of remarks not establishedThe TESCREAL bundle: Eugenics and the promise of utopia through artificial general intelligence

Source Archived

Context and checks

Surrounding words

We argue that, unlike systems with specific applications which can be evaluated following standard engineering principles, undefined systems like “AGI” cannot be appropriately tested for safety. Why, then, is building AGI often framed as an unquestioned goal in the field of AI? In this paper, we argue that the normative framework that motivates much of this goal is rooted in the Anglo-American eugenics tradition of the twentieth century. As a result, many of the very same discriminatory attitudes that animated eugenicists in the past (e.g., racism, xenophobia, classism, ableism, and sexism) remain widespread within the movement to build AGI, resulting in systems that harm marginalized groups and centralize power, while using the language of “safety” and “benefiting humanity” to evade accountability. We conclude by urging researchers to work on defined tasks for which we can develop safety protocols, rather than attempting to build a presumably all-knowing system such as AGI.

What this quote does not say

The passage is jointly authored with Émile P. Torres, so it establishes Gebru's authorship but not which coauthor drafted the sentence.
The passage concerns present and structural harms from AGI development, not a probability estimate or endorsement of extinction-risk claims.

How this was checked

Quote matched character-for-character against the captured page.

Reviewed by

an AI reviewer that read the whole source · an automatic character-by-character check

When people ask what regulations need to be in place to safeguard us from the unsafe uses of AI we’ve been seeing, I always start with labor protections and antitrust measures. I can tell that some people find that answer disappointing – perhaps because they expect me to mention regulations specific to the technology itself.

Exact quote, beginning

Timnit Gebru

founder of the Distributed AI Research Institute

In a 2021 Guardian op-ed, Timnit Gebru wrote that labor protections and antitrust should be the starting point for AI regulation and that the top safeguard was curbing company power and empowering people who speak up about harm.

Founder and executive director of the Distributed AI Research Institute (DAIR)Published December 6, 2021 · date of remarks not establishedFor truly ethical AI, its research must be independent from big tech

Source

Context and checks

Surrounding words

Twitter and especially TikTok, Telegram and Clubhouse have the same issues but are discussed much less. When I wrote a paper outlining the harms posed by models trained using data from these platforms, I was fired by Google. When people ask what regulations need to be in place to safeguard us from the unsafe uses of AI we’ve been seeing, I always start with labor protections and antitrust measures. I can tell that some people find that answer disappointing – perhaps because they expect me to mention regulations specific to the technology itself. While those are important, the #1 thing that would safeguard us from unsafe uses of AI is curbing the power of the companies who develop it and increasing the power of those who speak up against the harms of AI and these companies’ practices. Thanks to the hard work of Ifeoma Ozoma and her collaborators, California recently passed the Silenced No More Act, making it illegal to silence workers from speaking out about racism, harassment and other forms of abuse in the workplace.

What this quote does not say

The passage does not specify statutory text, jurisdiction, enforcement agencies, or implementation thresholds.
The phrase '#1 thing' ranks a package involving both reduced company power and increased worker or critic power; it should not be reduced to antitrust alone.

How this was checked

Quote matched character-for-character against the captured page.

Reviewed by

an AI reviewer that read the whole source · an automatic character-by-character check