Template:Did you know nominations/LaMDA
Appearance
- The following is an archived discussion of the DYK nomination of the article below. Please do not modify this page. Subsequent comments should be made on the appropriate discussion page (such as this nomination's talk page, the article's talk page or Wikipedia talk:Did you know), unless there is consensus to re-open the discussion at this page. No further edits should be made to this page.
The result was: promoted by Theleekycauldron (talk) 09:54, 12 July 2022 (UTC)
DYK toolbox |
---|
LaMDA
- ALT0: ... that Google engineer Blake Lemoine claims that the LaMDA neural language model has become sentient? Source: The Guardian
- ALT1: ... that the LaMDA neural language model was declared to be sentient by Google engineer Blake Lemoine? Source: The Guardian
- Reviewed: 3rd DYK nomination, exempt from QPQ
Moved to mainspace by InfiniteNexus (talk) and Jrincayc (talk). Nominated by InfiniteNexus (talk) at 06:00, 20 June 2022 (UTC).
General: Article is new enough and long enough |
---|
Policy: Article is sourced, neutral, and free of copyright problems |
---|
|
Hook: Hook has been verified by provided inline citation |
---|
|
QPQ: None required. |
Overall: Very interesting article, it seems balanced in favor of the scientific consensus and is well sourced. Hooks are cited and interesting. I'm in favor of ALT0 because it's careful not to endorse the claim that LaMDA is sentient while ALT1 might be more ambiguous. Otherwise, this nom is a-go! BuySomeApples (talk) 04:08, 25 June 2022 (UTC)
- Even so, @BuySomeApples and InfiniteNexus, a Google engineer appears to be an inherently trustworthy source for those who wouldn't deign to click through. It would be like if we said:
- ... that renowned geologist theleekycauldron claims that the earth is actually in the shape of a giant DYK prep set?
- Without mentioning that I would immediately be fired from any prestigious academic posts I held, were I to spout something like that (except this claim is actually somewhat believable). Could we find a way to work in that scientific consensus opposes the claim as stated? theleekycauldron (talk • contribs) (she/they) 11:37, 26 June 2022 (UTC)
- @Theleekycauldron: Hmm. How about:
- ALT2: ... that Google engineer Blake Lemoine claims that the LaMDA neural language model has become sentient, though his claims have largely been rejected by the scientific community? Source: The Guardian
- It's a little less hooky IMO, but it would address your concern. InfiniteNexus (talk) 16:53, 26 June 2022 (UTC)
- Well, how about:
- ALT3:... that Google engineer Blake Lemoine was suspended after he erroneously claimed that the LaMDA neural language model is sentient?
- A little rough around the edges, but we can hedge where we need to. theleekycauldron (talk • contribs) (she/they) 03:43, 28 June 2022 (UTC)
- @Theleekycauldron: I guess that would work too. InfiniteNexus (talk) 00:37, 29 June 2022 (UTC)
- Well, how about:
- @Theleekycauldron and InfiniteNexus: Could we maybe try a different hook, maybe something about how the scientific community reacted to erroneous claims that LaMDA was sentient - or about how that made scientists rethink the Turing Test? BuySomeApples (talk) 21:25, 29 June 2022 (UTC)
- Here are a bunch of different options:
- ALT4: ... that scientists have widely rejected Google engineer Blake Lemoine's claims that the LaMDA neural language model is sentient? Source: The Guardian
- ALT5: ... that Google engineer Blake Lemoine's erroneous claims that the LaMDA neural language model is sentient have generated conversations on the efficacy of the Turing test? Source: Fortune
- ALT6: ... that former New York University psychology professor Gary Marcus denounced claims that Google's LaMDA neural language model is sentient as "nonsense on stilts"? Source: CNN
- ALT7: ... that Google engineer Blake Lemoine was placed on administrative leave after hiring an attorney by neural language model LaMDA's request? Source: Business Insider
- Unfortunately, the source used in ALT5 is locked behind a paywall, and I'm having difficulty finding an alternate RS. I'm not sure whether it would be appropriate to use an archived link to bypass the paywall, should that hook be selected. InfiniteNexus (talk) 22:14, 29 June 2022 (UTC)
- @InfiniteNexus: Using paywalled sources on Wikipedia is fine, and the citation template actually has a special feature to add archived links just in case they become redirects or victims to link rot in the future. The WaPo article [1] already on the page also works as an RS. I think changing the beginning of ALT5 from "Google engineer Blake Lemoine's claims" to "erroneous claims" would also solve @Theleekycauldron:'s concerns about the hook. I'd be ready to approve ALT5 once that change is made. BuySomeApples (talk) 22:41, 29 June 2022 (UTC)
- We aren't using citation templates here, so
|archive-url=
wouldn't help us. The WaPo source is also paywall-locked. Anyway, I've made the change to ALT5, if that's the alt you like best. InfiniteNexus (talk) 23:01, 29 June 2022 (UTC)- @InfiniteNexus: thank you for making that change! Although the citation template isn't used on the nomination page, it can be used in the article. (Technically it's fine anyway per WP:PAYWALL). Weirdly, the WaPo article isn't paywalled for me, but you might have reached your free article limit. Since I can access both articles to verify, I think it's fine. I wonder if we can shorten the hook a little more, right now it's 167 characters and the hard limit is 200. Are you OK with removing the "Google engineer Blake Lemoine" part and replacing it with just "an erroneous claim"? BuySomeApples (talk) 01:49, 2 July 2022 (UTC)
- @BuySomeApples: Yes, I would be fine with that removak. So the final hook would be:
- ALT8: ... that an erroneous claim that the LaMDA neural language model is sentient has generated conversations on the efficacy of the Turing test? Source: Fortune
- Thanks for the note about WaPo, I guess that was probably the case. The archived links to both sources are already in the article, FYI. InfiniteNexus (talk) 03:34, 2 July 2022 (UTC)
- @BuySomeApples: Yes, I would be fine with that removak. So the final hook would be:
- @InfiniteNexus: thank you for making that change! Although the citation template isn't used on the nomination page, it can be used in the article. (Technically it's fine anyway per WP:PAYWALL). Weirdly, the WaPo article isn't paywalled for me, but you might have reached your free article limit. Since I can access both articles to verify, I think it's fine. I wonder if we can shorten the hook a little more, right now it's 167 characters and the hard limit is 200. Are you OK with removing the "Google engineer Blake Lemoine" part and replacing it with just "an erroneous claim"? BuySomeApples (talk) 01:49, 2 July 2022 (UTC)
- We aren't using citation templates here, so
- @InfiniteNexus: Using paywalled sources on Wikipedia is fine, and the citation template actually has a special feature to add archived links just in case they become redirects or victims to link rot in the future. The WaPo article [1] already on the page also works as an RS. I think changing the beginning of ALT5 from "Google engineer Blake Lemoine's claims" to "erroneous claims" would also solve @Theleekycauldron:'s concerns about the hook. I'd be ready to approve ALT5 once that change is made. BuySomeApples (talk) 22:41, 29 June 2022 (UTC)
- Here are a bunch of different options:
- @Theleekycauldron: Hmm. How about:
Thanks @InfiniteNexus: ALT8 is approved! BuySomeApples (talk) 00:07, 7 July 2022 (UTC)