Post Abf1ep8P81dFTA8Ty4 by tante@tldr.nettime.org
(DIR) More posts by tante@tldr.nettime.org
(DIR) Post #AbetPuCjCKAhDRX9Ky by tante@tldr.nettime.org
2023-11-10T08:55:53Z
0 likes, 0 repeats
The way that LLMs shield you from learning is a form of intellectual self-harm TBH.
(DIR) Post #Abeu1QN6YyXZhCHHRQ by rhold@norden.social
2023-11-10T09:05:32Z
0 likes, 0 repeats
@tante in my opinion that's the most important reason to avoid LLMs.It will widen the gap between people who know and understand stuff and people who don't.
(DIR) Post #AbeugAu4yhp1ax8HXE by CatherineFlick@mastodon.me.uk
2023-11-10T09:13:39Z
0 likes, 0 repeats
@tante I've heard of PhD students submitting LLM-written abstracts for conferences and I'm like really??? How does this benefit you??
(DIR) Post #AbevvTbLqFl3sqEVHM by morandi@hachyderm.io
2023-11-10T09:27:47Z
0 likes, 0 repeats
@tante It’s very easy to make the counter argument — LLMs force you into learning.
(DIR) Post #AbexP3iQkVBmF3DKcK by Npars01@mstdn.social
2023-11-10T09:44:21Z
0 likes, 0 repeats
@tante Psychological crutch meet Duning-Krueger
(DIR) Post #AbeyNnJqKFDF3Mfs92 by tante@tldr.nettime.org
2023-11-10T09:55:22Z
0 likes, 0 repeats
@CatherineFlick TBH: For PhD Students I do at least see the pragmatic value. They need a bunch of publications which is a lot of work so generate the abstracts and try to hit any conference to fulfill your requirements.(But I have a very cynical view on the way science is structured organizationally so maybe that's a tad unfair)
(DIR) Post #AbeyaTlKLeE0ADZv6m by tante@tldr.nettime.org
2023-11-10T09:57:09Z
0 likes, 0 repeats
@CatherineFlick Like dissertations are written not to be read but to be written to get the thing. Publications for PhD students (or professors trying to get tenure etc) have been overloaded with so much external shit that it stands in the way of all the abstract values of science, sharing, learning etc.
(DIR) Post #AbezNOQcy4p3oKDiQi by morandi@hachyderm.io
2023-11-10T10:03:17Z
0 likes, 0 repeats
@zombiecide @tante Even the developers are saying publicly that we shouldn’t believe LLMs, they’re like a single JPG of the entire internet, so any detail is going to be fuzzy., and we don’t fully understand how they produce outputs. Therefore if you’re using LLMs and you’re a half-ways informed and intelligent person, you are going to question and query and reprompt and do your own research. Thus LLMs foster discovery and learning.
(DIR) Post #AbezNPR1E6jIvpPXg8 by tante@tldr.nettime.org
2023-11-10T10:06:32Z
0 likes, 0 repeats
@morandi @zombiecide the developers say that to shield themselves from criticisms. But we do see that that's exactly what's not happening: People use these systems as shorthand to generate something to fulfill whatever shirt term goal they have without checking. Because checking often takes more time than just doing the work so it would totally invalidate the point. One could use these systems as catalyst to spark self-directed learning. And a few people actually might. But that's not happening at scale, not on a structural level.
(DIR) Post #AbezfddVAI8oBlwT1E by morandi@hachyderm.io
2023-11-10T10:09:51Z
0 likes, 0 repeats
@tante @zombiecide So much of your argument is an argument against society, behaviours and systems (PhDs, academic publishing, etc.) but not about the technology. You could make the same argument about printed encyclopaedias.
(DIR) Post #Abezvxd8axA8phYSx6 by rticks@mastodon.social
2023-11-10T10:12:42Z
0 likes, 0 repeats
@tante They also make anyone who uses them a thief unless trained on ethical source material
(DIR) Post #Abf0qblOZGXLWKQePY by mnl@hachyderm.io
2023-11-10T10:22:52Z
0 likes, 0 repeats
@tante did google or the calculator or the computer shield people from learning though?If people get their short term goals solved, does that mean they’ll stop there? Or might they actually learn about things that interest them more? What should people have to learn I’m the first place? At a personal level (and friends), llms seem to be a technology that makes people rediscover the fun of learning and building things by lowering barriers to entry.1/
(DIR) Post #Abf0qejNXDhqiwVaBU by mnl@hachyderm.io
2023-11-10T10:22:53Z
0 likes, 0 repeats
@tante For example, people building home automation and embedded software they never would have, people having conversations in languages they are learning, understanding computer science concepts I never fully understood (me), building 10x more opensource than I ever did (again me), getting into programming (my boss, another coworker).People who are interested in learning won’t stop because of llms, and people who aren’t aren’t doing it yet either.
(DIR) Post #Abf1TuYYd7I56EAJF2 by estelle4565@masto.nu
2023-11-10T10:30:03Z
0 likes, 0 repeats
@tanteWhich University allows people to get a PhD without publications in journals?@CatherineFlick
(DIR) Post #Abf1ep8P81dFTA8Ty4 by tante@tldr.nettime.org
2023-11-10T10:31:27Z
0 likes, 0 repeats
@estelle4565 @CatherineFlick Depending on the department that's rather common. In CS for example journals are not as important as in other domains, there it's more about conferences.
(DIR) Post #Abf1qHdpTsdf4EIbWS by estelle4565@masto.nu
2023-11-10T10:33:18Z
0 likes, 0 repeats
@tanteThat is an issue of the people and how they are advertised though not the technology itself.Do not get me wrong I am not a fan and I personally do not use them, but they do have some limited (at least for now), good applications. They are terrible for the planet though.@morandi @zombiecide
(DIR) Post #Abf4hhrIw5QPdRCgi0 by ythos@dotnet.social
2023-11-10T11:06:13Z
0 likes, 0 repeats
@tante And yet, the opportunities they offer for learning are amazing.For example - LLMs are really good content summarizers - if you write an abstract and get the LLM to boil it down to a couple of paragraphs, and it *doesn't* produce what you'd expect, your abstract is probably not emphasizing the things you thought it was, or it lacks clarity or contains contradictions.You can use the fact that LLMs are based on statistical consensus to explore that consensus.
(DIR) Post #Abf6Sc5GKAVhHzfAsS by Patashu@dragon.style
2023-11-10T11:25:35Z
0 likes, 0 repeats
@tante curious about this view - it reminds me of 'people won't ever learn anything more, they'll just google it'.BUT, I don't use LLMs to answer questions I have (any question I actually have and google doesn't answer is too obtuse or esoteric for an LLM to be able to help with), so I don't know how people who DO do this use it.
(DIR) Post #Abf6ctwolIqNaYcRI8 by mado@mastodon.online
2023-11-10T11:25:38Z
0 likes, 0 repeats
@tante interesting discussion. Seems like it depends very much on the understanding of 'learning'. For instance one can ask is it really possible not to learn? Then the question kinda transforms into the maybe more precise and interesting one of 'what' and 'how'..
(DIR) Post #Abf70FBNydHM6lZ0OO by IanSudbery@genomic.social
2023-11-10T11:32:00Z
0 likes, 0 repeats
@tante @CatherineFlick Good luck defending a dissertation written by an LLM in a viva exam.
(DIR) Post #AbfDrL2TX8QRLO62dM by CatherineFlick@mastodon.me.uk
2023-11-10T12:48:46Z
0 likes, 0 repeats
@tante They should still be learning how to pitch their work for conferences rather than relying on LLMs to do it for them. The PhD is a training exercise for academia (warts and all), writing these sorts of things is part of that training.
(DIR) Post #AbfU35nhc5MspCoUDI by estelle4565@masto.nu
2023-11-10T15:50:13Z
0 likes, 0 repeats
@tanteCS is?@CatherineFlick