Prof. Emily M. Bender(she/her)Feb 6, 2023

"We come to bury ChatGPT, not to praise it." Excellent piece by @danmcquillan
https://www.danmcquillan.org/chatgpt.html

I suggest you read the whole thing, but some pull quotes:

"ChatGPT is a part of a reality distortion field that obscures the underlying extractivism and diverts us into asking the wrong questions and worrying about the wrong things."

>>

We come to bury ChatGPT, not to praise it.

Show threadProf. Emily M. Bender(she/her)

Pull quotes continued:

"The compulsion to show 'balance' by always referring to AI's alleged potential for good should be dropped by acknowledging that the social benefits are still speculative while the harms have been empirically demonstrated."

"Instead of reactionary solutionism, let us ask where the technologies are that people really need. Let us reclaim the idea of socially useful production, of technological developments that start from community needs."

-- @danmcquillan

>>

Feb 6, 2023 at 3:36pmWeb
Show threadProf. Emily M. Bender(she/her)Feb 6, 2023

And finally:

"Transformer models and diffusion models are not creative but carceral - they and other forms of AI imprison our ability to imagine real alternatives."

--@danmcquillan

Show threadNorbieWegsFeb 6, 2023

@emilymbender @danmcquillan

Algorithmic Prison-house indeed.

Show threadPeter ButlerFeb 6, 2023
@emilymbender @danmcquillan This is the fundamental argument against LLM, to me. There are NO new ideas. NONE.
Show threadmeowdudeFeb 6, 2023
@peterbutler @emilymbender @danmcquillan It is also why at the moment it isn't a threat.
Show threadStephan EggermontFeb 7, 2023
@meowdude @peterbutler @emilymbender @danmcquillan It is a critical threat to our knowledge distribution infrastructure and therefore democracy now.
Show threadDaniel M KarlssonFeb 7, 2023
@emilymbender Excellent.
Show threadJigenFeb 12, 2023
@emilymbender @danmcquillan there are real material benefits to AI as well, https://www.newscientist.com/article/2330866-deepminds-protein-folding-ai-cracks-biologys-biggest-problem/, https://phys.org/news/2022-05-ai-reveals-unsuspected-math-underlying.html, https://www.scientificamerican.com/article/ai-generates-hypotheses-human-scientists-have-not-thought-of/
DeepMind's protein-folding AI cracks biology's biggest problem

Artificial intelligence firm DeepMind has transformed biology by predicting the structure of nearly all proteins known to science in just 18 months, a breakthrough that will speed drug development and revolutionise basic science

New Scientist
Show threaddanmcquillanFeb 12, 2023
@JigenD @emilymbender maybe; i'm not convinced. i'd like to see a proper study of the effort that goes into producing and the dismissing all the guesses that turn out to be wrong, compared to existing forms of scientific enquiry.
Show threaddanmcquillanFeb 12, 2023
@JigenD @emilymbender and, stepping back a bit, i suspect that _if_ dl/rl turn out to be able to _do science_ is some non-guesswork way then it will turn out to intensify those aspects of the scientific approach which feminist and post-colonial critiques of science have been trying to challenge