The LLM warnings Google fired Timnit Gebru over have all come true • Hacker News Board

Re: The LLM warnings Google fired Timnit Gebru over have all come true

#31

Post by tptacek » Thu, Jun 04, 2026, 5:13 PM UTC

benob wrote: post #29

skupig wrote: post #23

Earlier quoted context omitted.

There has been plenty of research that shows LLMs encode social biases. It seems pretty obvious even before looking at the research that training on the whole internet will end up encoding widely-held social biases and stereotypes. https://arxiv.org/pdf/2508.07111 https://github.com/angl1n/social-bias-llm-vlm

And papers on bias amplification in ML predate LLMs. I remember this specific one which was a spotlight paper at EMNLP: Men Also Like Shopping: Reducing Gender Bias Amplification using Corpus-level Constraints, Zhao et al. https://arxiv.org/abs/1707.09457

The bias concerns in Gebru's paper cover pre-LLM systems. For all we know, modern frontier models might mitigate many of the concerns the paper brings up. It's hard to know. The logic used in summaries like the one we're commenting on is conclusory: centuries of prejudice are encoded in the total corpus of human language, language models are trained on that corpus, ergo language models must be biased.

Re: The LLM warnings Google fired Timnit Gebru over have all come true

#32

Post by taeric » Thu, Jun 04, 2026, 5:14 PM UTC

tptacek wrote: post #27

skupig wrote: post #23

Earlier quoted context omitted.

There has been plenty of research that shows LLMs encode social biases. It seems pretty obvious even before looking at the research that training on the whole internet will end up encoding widely-held social biases and stereotypes. https://arxiv.org/pdf/2508.07111 https://github.com/angl1n/social-bias-llm-vlm

Have you read through the sources on that Github link? It's a set of sociology cites establishing that bias exists (something no serious person ever disputed), followed by a couple papers showing mechanistic descriptions of how bias could propagate through an LLM. The paper you call out specifically takes last-generation open-weights models and attempts to trick them into revealing biases through their level of confi…

I confess I laughed harder at the Grok comment than I wish I had. Sad to remember that some strawmen are given life and promoted by people. Actively.

Re: The LLM warnings Google fired Timnit Gebru over have all come true

#33

Post by plastic-enjoyer » Thu, Jun 04, 2026, 5:14 PM UTC

Legend2440 wrote: post #21

laweijfmvo wrote: post #5

The warnings: > The first warning was about scale itself. Bender and Gebru argued that training ever-larger models on ever-larger scrapes of the internet would produce systems that appeared fluent but had no actual understanding of language. > The second warning was about bias amplification. The paper documented in detail that internet-scale training data contains systematic overrepresentation of dominant viewpoints…

Yeah, I think it's pretty clear that LLMs are more than mere "stochastic parrots" - they can prove theorems, follow instructions, and complete complex tasks. This was the most notable claim of the paper, and it's aged very poorly.

Are they, though? I think what LLMs proved is that proving theorems, following instructions and solving complex problems - intelligent behaviour - does not need any kind of understanding, but only ability to recombine things in a stochastic matter. Which basically just means that these things weren't as special as people had thought.

Re: The LLM warnings Google fired Timnit Gebru over have all come true

#34

Post by staticman2 » Thu, Jun 04, 2026, 5:15 PM UTC

epolanski wrote: post #12

I don't want to say this has not happened, but where's the evidence of anything in this article? According to the article she resigned, which is very different from getting fired, so what is the information the author has to substantiate this claim?

I agree. Why is someone's lazy Tumblr hot take getting upvoted here? Are people considering it a good conversation starter or something?

Re: The LLM warnings Google fired Timnit Gebru over have all come true

#35

Post by PaulDavisThe1st » Thu, Jun 04, 2026, 5:15 PM UTC

taeric wrote: post #26

wesleywt wrote: post #14

Earlier quoted context omitted.

Why should the person identifying the problem provide a solution? This doesn't make sense.

If the criticism can't distill up from "bad things could happen", it just isn't useful to keep paying people to come up with that kind of critique. And it isn't like we stopped paying attention to these concerns, is it? Nor were they completely blind siding us at the time. The question was largely of what to do about them.

The question also whether large-scale utilization of LLMs (and also the prerequisite increased training processes) should proceed before these issues were addressed. Clearly, we collectively answered "yes" without any actual reasoning (and arguably, without any collective decision making either).

Re: The LLM warnings Google fired Timnit Gebru over have all come true

#36

Post by PaulDavisThe1st » Thu, Jun 04, 2026, 5:16 PM UTC

taeric wrote: post #28

thisisthenewme wrote: post #24

Earlier quoted context omitted.

When a researcher discovers that smoking is damaging to the lungs, do they need to provide a solution that allows people to smoke without damaging their lungs? Would their inability to provide a solution take anything away from the research?

To conflate AI with smoking is just not helpful. At all. Or are you saying that there are acute harms from AI that are being ignored?

Acute, chronic - why would it matter?

Why is it unhelpful to conflate AI with smoking?

And yes, lots of people are saying "there are harms from AI that are being ignored".

Re: The LLM warnings Google fired Timnit Gebru over have all come true

#37

Post by ChrisArchitect » Thu, Jun 04, 2026, 5:17 PM UTC

What is/was the source of this rather than random tumblr?

This May 26th Twitter post ...maybe? Account now suspended https://x.com/heygurisingh/status/2059251382960734593

(http://web.archive.org/web/20260526123243/https://twitter.co...)

Re: The LLM warnings Google fired Timnit Gebru over have all come true

#38

Post by Legend2440 » Thu, Jun 04, 2026, 5:18 PM UTC

plastic-enjoyer wrote: post #33

Legend2440 wrote: post #21

Earlier quoted context omitted.

Yeah, I think it's pretty clear that LLMs are more than mere "stochastic parrots" - they can prove theorems, follow instructions, and complete complex tasks. This was the most notable claim of the paper, and it's aged very poorly.

Are they, though? I think what LLMs proved is that proving theorems, following instructions and solving complex problems - intelligent behaviour - does not need any kind of understanding, but only ability to recombine things in a stochastic matter. Which basically just means that these things weren't as special as people had thought.

I think you have already decided that LLMs cannot possibly understand. Therefore anything they do must not have required understanding in the first place. It's circular logic.

Re: The LLM warnings Google fired Timnit Gebru over have all come true

#39

Post by tptacek » Thu, Jun 04, 2026, 5:18 PM UTC

plastic-enjoyer wrote: post #33

Legend2440 wrote: post #21

Earlier quoted context omitted.

Yeah, I think it's pretty clear that LLMs are more than mere "stochastic parrots" - they can prove theorems, follow instructions, and complete complex tasks. This was the most notable claim of the paper, and it's aged very poorly.

Are they, though? I think what LLMs proved is that proving theorems, following instructions and solving complex problems - intelligent behaviour - does not need any kind of understanding, but only ability to recombine things in a stochastic matter. Which basically just means that these things weren't as special as people had thought.

We've clearly crossed a threshold at which "stochastic" is no longer doing the work Gebru (and, more importantly, the acolytes of this paper; I shouldn't tar Gebru with what they've done with the work) expected it to do. Lots of important processes are stochastic, including at some levels human thought itself. Advocates who deploy the term "stochastic" seem to believe it impeaches the technology, which is kind of embarrassing to see.

Re: The LLM warnings Google fired Timnit Gebru over have all come true

#40

Post by taeric » Thu, Jun 04, 2026, 5:18 PM UTC

camphy wrote: post #30

taeric wrote: post #8

Earlier quoted context omitted.

More than not being entirely sure what the impact is, I don't see any suggestion at what to do about it?

If you’re referring to a solution to large datasets without not being auditable, she actually did provide a solution. Something to do with data sheets for these training data sets similar to those provided for hardware components. At least, if my memory serves me.

I was more irked by the diversity of teams developing these concern. Which, feels like a benign enough concern, but not one where you can just stop progress.

Worse, I think it is a ridiculously safe bet that the US was home to the most diverse teams you could get for this sort of work. Asking the good faith participants to stop participating would have decreased the stated goal.