<![CDATA[The power of ChatGPT]]>The power of ChatGPT

]]>https://forum.fedi.dk/topic/21efe3c9-8ba7-4760-a486-0f2ab04eca64/the-power-of-chatgptRSS for NodeMon, 06 Apr 2026 14:16:40 GMTMon, 30 Mar 2026 18:36:04 GMT60<![CDATA[Reply to The power of ChatGPT on Wed, 01 Apr 2026 08:53:38 GMT]]>@GossiTheDog This is what techbros are uplifting as that which will replace human workers...They are a fucking joke.

They will never get humans out of the workplace with a bullshitting bundle of stolen data and weighted tokens.

]]>https://forum.fedi.dk/post/https://sunny.garden/users/WanderingInDigitalWorlds/statuses/116328603197710737https://forum.fedi.dk/post/https://sunny.garden/users/WanderingInDigitalWorlds/statuses/116328603197710737Wed, 01 Apr 2026 08:53:38 GMT<![CDATA[Reply to The power of ChatGPT on Wed, 01 Apr 2026 01:28:21 GMT]]>@GossiTheDog this is hilarious and if it had alt text I’d boost it.

]]>https://forum.fedi.dk/post/https://social.coop/users/heybenji/statuses/116326852314564713https://forum.fedi.dk/post/https://social.coop/users/heybenji/statuses/116326852314564713Wed, 01 Apr 2026 01:28:21 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 21:04:00 GMT]]>@GossiTheDog If we assume this is the answer that LLMs are able to give, does this change the argument we are having? As far as I can tell from other sources I've reviewed, this answer seems to be reasonable.

]]>https://forum.fedi.dk/post/https://inuh.net/users/jafo/statuses/116325812803578317https://forum.fedi.dk/post/https://inuh.net/users/jafo/statuses/116325812803578317Tue, 31 Mar 2026 21:04:00 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 20:52:56 GMT]]>@tempusfelix @alice @GossiTheDog
Fwiw describing that image as "chatgpt being wrong" is like describing this one as "a funny tweet". Technically correct but not actually very helpful to a visually impaired person trying to understand the joke.

]]>https://forum.fedi.dk/post/https://chaos.social/users/lexr/statuses/116325769330144262https://forum.fedi.dk/post/https://chaos.social/users/lexr/statuses/116325769330144262Tue, 31 Mar 2026 20:52:56 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 16:01:41 GMT]]>@benjamineskola @GossiTheDog @loucyx @jafo It’s the difference between lying and bullshitting. Lying at least has a regard for what is true. Bullshitting doesn’t care if it is correct or not.

]]>https://forum.fedi.dk/post/https://hcommons.social/users/musiciankate/statuses/116324624068627893https://forum.fedi.dk/post/https://hcommons.social/users/musiciankate/statuses/116324624068627893Tue, 31 Mar 2026 16:01:41 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 15:48:42 GMT]]>@benjamineskola @jafo @GossiTheDog 100% this! If they were always right or always wrong it would be one thing, but the only constant is that they are always confident about their answer (either if it’s right or wrong) which is what makes them dangerously unreliable.

And this isn’t even getting into the whole detrimental effect they have on cognitive analysis and reasoning for LLM consumers.

]]>https://forum.fedi.dk/post/https://mastodon.social/users/loucyx/statuses/116324573007966700https://forum.fedi.dk/post/https://mastodon.social/users/loucyx/statuses/116324573007966700Tue, 31 Mar 2026 15:48:42 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 15:35:43 GMT]]>@jafo @loucyx @GossiTheDog but your mileage should not vary. that's the point.

getting a different answer each time is what makes these tools not fit for purpose. if they return the right answer some of the time but you never know which times, what's the point in them?

]]>https://forum.fedi.dk/post/https://hachyderm.io/users/benjamineskola/statuses/116324521998362806https://forum.fedi.dk/post/https://hachyderm.io/users/benjamineskola/statuses/116324521998362806Tue, 31 Mar 2026 15:35:43 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 14:28:24 GMT]]>@loucyx @GossiTheDog I don't know about you, but I've long ago learned to not just blindly trust tools I use, on the Internet and elsewhere. I use tools understanding the limitations, and check the work. In this case, it seemed like outside sources corroborated the assertions ChatGPT made. I can't speak to Kevin's answer, because no information on WHAT ChatGPT was given; as I said, I used "Thinking-Standard" to get my answer, YMMV if you use other models.

]]>https://forum.fedi.dk/post/https://inuh.net/users/jafo/statuses/116324257260286256https://forum.fedi.dk/post/https://inuh.net/users/jafo/statuses/116324257260286256Tue, 31 Mar 2026 14:28:24 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 14:28:11 GMT]]>@jafo @loucyx @GossiTheDog Elsewhere in this thread, Kenny Everett was claimed to be the first — but the timeline might be wrong for that, depending when he actually came out.

]]>https://forum.fedi.dk/post/https://hachyderm.io/users/benjamineskola/statuses/116324256444714768https://forum.fedi.dk/post/https://hachyderm.io/users/benjamineskola/statuses/116324256444714768Tue, 31 Mar 2026 14:28:11 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 14:21:24 GMT]]>@benjamineskola @loucyx @GossiTheDog What do you consider a correct answer? According to the respective Wikipedia entries for them, the answer I got seems to be correct. The answer ChatGPT gave me linked to citations which also seemed to back up the answer. https://www.theguardian.com/tv-and-radio/2022/aug/25/farewell-scott-mills-bbc-radio-1?utm_source=chatgpt.com https://en.wikipedia.org/wiki/Kevin_Greening?utm_source=chatgpt.com https://en.wikipedia.org/wiki/Scott_Mills?utm_source=chatgpt.com

]]>https://forum.fedi.dk/post/https://inuh.net/users/jafo/statuses/116324229752994881https://forum.fedi.dk/post/https://inuh.net/users/jafo/statuses/116324229752994881Tue, 31 Mar 2026 14:21:24 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 13:59:40 GMT]]>@GossiTheDog no homo gay is still gay in the digital cat fart world lol

]]>https://forum.fedi.dk/post/https://mstdn.ca/users/idren/statuses/116324144270418705https://forum.fedi.dk/post/https://mstdn.ca/users/idren/statuses/116324144270418705Tue, 31 Mar 2026 13:59:40 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 11:02:34 GMT]]>@tempusfelix @alice @GossiTheDog Though Kenny was openly gay by the late Eighties and was certainly one of the first Radio 1 presenters in the Sixties. I don't think he was openly gay at the same time that he was presenting at Radio 1

]]>https://forum.fedi.dk/post/https://techhub.social/users/technicaladept/statuses/116323447879127079https://forum.fedi.dk/post/https://techhub.social/users/technicaladept/statuses/116323447879127079Tue, 31 Mar 2026 11:02:34 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 08:58:45 GMT]]>@loucyx @jafo @GossiTheDog it's also not even correct, so what you've managed to get there is a different wrong answer.

If you think 'confidentaly incorrect' is an improvement over 'obvious gibberish', then yeah, I suppose this is preferable, but it doesn't get you any closer to the truth.

(personally I think 'obviously wrong' is preferable, because then at least you know to ignore it.)

]]>https://forum.fedi.dk/post/https://hachyderm.io/users/benjamineskola/statuses/116322961038662096https://forum.fedi.dk/post/https://hachyderm.io/users/benjamineskola/statuses/116322961038662096Tue, 31 Mar 2026 08:58:45 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 08:34:23 GMT]]>@GossiTheDog decided to experiment with AI by asking which philosopher was run over by a milk float and it did a similar thing.

If you know who it actually was, that would be much appreciated!

]]>https://forum.fedi.dk/post/https://tech.lgbt/users/criticalbackend/statuses/116322865216490795https://forum.fedi.dk/post/https://tech.lgbt/users/criticalbackend/statuses/116322865216490795Tue, 31 Mar 2026 08:34:23 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 08:12:33 GMT]]>@GossiTheDog imagine how much fossil fuel was used to generate that sophisticated answer. Any such energy calculation must include all resources required to build the data sets that are
required for the system to perform the operation.

]]>https://forum.fedi.dk/post/https://toots.nu/users/kofzmann/statuses/116322779378343589https://forum.fedi.dk/post/https://toots.nu/users/kofzmann/statuses/116322779378343589Tue, 31 Mar 2026 08:12:33 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 08:11:40 GMT]]>@GossiTheDog the right answer is probably Kenny Everett

]]>https://forum.fedi.dk/post/https://mastodon.social/users/foobarry/statuses/116322775892787839https://forum.fedi.dk/post/https://mastodon.social/users/foobarry/statuses/116322775892787839Tue, 31 Mar 2026 08:11:40 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 04:43:52 GMT]]>@jafo @GossiTheDog even then, a reliable source of information should be consistent, meaning both Kevin and you should have gotten the same result, but we all know LLMs aren't consistent (even when the same user asks the same question) so if anything, you added more evidence proving we should avoid LLMs 🤷🏻‍♀️

]]>https://forum.fedi.dk/post/https://mastodon.social/users/loucyx/statuses/116321958767092084https://forum.fedi.dk/post/https://mastodon.social/users/loucyx/statuses/116321958767092084Tue, 31 Mar 2026 04:43:52 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 03:49:40 GMT]]>@GossiTheDog reminds me of https://www.youtube.com/watch?v=OMdPj3HXMgQ

]]>https://forum.fedi.dk/post/https://masto.hackers.town/users/cinebox/statuses/116321745675465478https://forum.fedi.dk/post/https://masto.hackers.town/users/cinebox/statuses/116321745675465478Tue, 31 Mar 2026 03:49:40 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 03:29:29 GMT]]>@GossiTheDog You can refer to me as Maddie (not gay)

]]>https://forum.fedi.dk/post/https://masto.hackers.town/users/maddiefuzz/statuses/116321666288886958https://forum.fedi.dk/post/https://masto.hackers.town/users/maddiefuzz/statuses/116321666288886958Tue, 31 Mar 2026 03:29:29 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 03:24:01 GMT]]>@GossiTheDog@cyberplace.social ​​

]]>https://forum.fedi.dk/post/https://hatoya.cafe/notes/akhma856y0sn05jnhttps://forum.fedi.dk/post/https://hatoya.cafe/notes/akhma856y0sn05jnTue, 31 Mar 2026 03:24:01 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 02:30:28 GMT]]>@wotsac @alice @GossiTheDog

ChatGPT must have been trained on the 1954 version of the film Sabrina, where a French chef in a cooking school tells his class that a souffle must be "gay" (different connotation in 1954).

https://www.youtube.com/watch?v=FR7KlvISE2w

]]>https://forum.fedi.dk/post/https://fosstodon.org/users/bzdev/statuses/116321434259878385https://forum.fedi.dk/post/https://fosstodon.org/users/bzdev/statuses/116321434259878385Tue, 31 Mar 2026 02:30:28 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 01:53:21 GMT]]>@GossiTheDog

]]>https://forum.fedi.dk/post/https://mastodon.online/users/PizzaDemon/statuses/116321288292703777https://forum.fedi.dk/post/https://mastodon.online/users/PizzaDemon/statuses/116321288292703777Tue, 31 Mar 2026 01:53:21 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 01:35:55 GMT]]>@simonzerafa @GossiTheDog

Nah, just live with the 20% chance that the next moon landing will bury the astronauts 20 metres underground. Trump will approve of this because it will distract people from the Epstein files for a period of time.

]]>https://forum.fedi.dk/post/https://mstdn.social/ap/users/116148271994190049/statuses/116321219724846721https://forum.fedi.dk/post/https://mstdn.social/ap/users/116148271994190049/statuses/116321219724846721Tue, 31 Mar 2026 01:35:55 GMT<![CDATA[Reply to The power of ChatGPT on Tue, 31 Mar 2026 01:26:08 GMT]]>@lbcp @GossiTheDog

Gays... Guys... the words look the same, so they must be the same.

]]>https://forum.fedi.dk/post/https://mstdn.social/ap/users/116148271994190049/statuses/116321181257638876https://forum.fedi.dk/post/https://mstdn.social/ap/users/116148271994190049/statuses/116321181257638876Tue, 31 Mar 2026 01:26:08 GMT