[HN Gopher] Which Humans? (2023)
___________________________________________________________________
Which Humans? (2023)
Author : surprisetalk
Score : 24 points
Date : 2025-12-29 19:57 UTC (3 hours ago)
(HTM) web link (osf.io)
(TXT) w3m dump (osf.io)
| jdkee wrote:
| As an aside: Last year a student of mine (we're at a U.S.
| college) told me that his teenage cousins back in Mongolia were
| all learning English in order to use ChatGPT.
| croisillon wrote:
| is it a steppe up?
| MengerSponge wrote:
| /usr/bin/humans, presumably
| dwa3592 wrote:
| /usr/bin/human/weird, to be precise.
| mncharity wrote:
| Since the page didn't load for me several times, and the title is
| ambiguous, here's the Abstract: Large language models (LLMs) have
| recently made vast advances in both generating and analyzing
| textual data. Technical reports often compare LLMs' outputs with
| "human" performance on various tests. Here, we ask, "Which
| humans?" Much of the existing literature largely ignores the fact
| that humans are a cultural species with substantial psychological
| diversity around the globe that is not fully captured by the
| textual data on which current LLMs have been trained. We show
| that LLMs' responses to psychological measures are an outlier
| compared with large-scale cross-cultural data, and that their
| performance on cognitive psychological tasks most resembles that
| of people from Western, Educated, Industrialized, Rich, and
| Democratic (WEIRD) societies but declines rapidly as we move away
| from these populations (r = -.70). Ignoring cross-cultural
| diversity in both human and machine psychology raises numerous
| scientific and ethical issues. We close by discussing ways to
| mitigate the WEIRD bias in future generations of generative
| language models.
| rokizero wrote:
| This was submitted 30 months ago. Still interesting. I would be
| interested if this got 'worse' or 'better' with newer models.
| andy99 wrote:
| 2023, and using some kind of in-page pdf reader from 2002
| wellthisisgreat wrote:
| It's 2025 today, Andy :)
| didgetmaster wrote:
| Surprise, Surprise. LLMs will respond according to the set of
| data that their model was trained on!
|
| While just about every LLM is trained on data that far surpasses
| the output of just one person, or even a decent sized group; it
| will still reflect the average sentiment of the corpus of data
| fed into it.
|
| If the bulk of the training data was scraped from websites
| created in 'WEIRD' countries, then it's responses will largely
| mimic their culture.
| cathyreisenwitz wrote:
| I wonder whether it might be useful to the continued existence of
| humanity to correct for the individualism bias in WEIRD countries
| with some collectivism
| levocardia wrote:
| I think it is mostly a good thing that LLMs have "WEIRD" values.
| We are at a very fortuitous point in history, where the modal
| position in extant written text is classically liberal and
| believes in respecting human rights. Virtually no other point in
| history would be that way, and a true modal position among values
| and moral beliefs held among all 8 billion people currently alive
| on earth -- much less the modal position among all ~100 billion
| humans ever -- would, I'd hazard to guess, not be a very nice
| place to end up.
___________________________________________________________________
(page generated 2025-12-29 23:00 UTC)