3 ms·

Spoiler (but not really): academic paper on ChatGPT written by ChatGPT. See, for example: AI makes plagiarism harder to detect, argue academics – in paper writ

Spoiler (but not really): academic paper on ChatGPT written by ChatGPT. See, for example: AI makes plagiarism harder to detect, argue academics – in paper written by chatbot - https://www.theguardian.com/technology/2023/mar/19/ai-makes-plagiarism-harder-to-detect-argue-academics-in-paper-written-by-chatbot https://www.theguardian.com/technology/2023/mar/19/ai-makes-...

revelio 4y ago
>> institutions admit they are racing to catch up LOL these institutions aren't racing anywhere, they've been publishing papers written by SciGen and other obvious nonsense text generators for years and are still doing it: https://pubpeer.com/search?q=scigen https://pubpeer.com/search?q=scigen If they can't even keep crap generated by Perl scripts out of the literature what hope do they have against ChatGPT? >> the biggest [clue] is that it does not properly understand academic referencing – a vital part of written university work – and often uses “suspect” references So just like non-student papers then? https://retractionwatch.com/2020/09/09/an-isolated-incident-should-reviewers-check-references/ https://retractionwatch.com/2020/09/09/an-isolated-incident-... >> She added: “If you cheat your way to a degree, you might get an initial job, but you won’t do well and your career won’t progress the way you want it to.” Loads of people in super highly placed and powerful positions have been discovered decades later to have plagiarised their phd or masters thesis. Cheating works because nobody is reading most academic output, it's only the piece of paper that people care about. >> She said that students who can’t spot the weaknesses in what the bot is producing may slip up. “In my subject of computer science, AI tools can generate code but it will often contain bugs,” she explained uhhhmm student programs contain bugs too even when not written by AI, that's why they're paying you to teach them. Who is this fool? "Irene Glendinning, head of academic integrity at Coventry University", oh right of course she is. Academia is over, it's just gonna take like a decade for people to fully accept that and another decade or two to do something about it. Cheating is rampant even amongst the professors even before ChatGPT, which is why they're so lazy about tackling it amongst the students. Now there are cheap LLMs academic output, whether of papers (already a mostly write-only medium) or grades, is doomed to become a sick joke that we're all forced to pay for.
- Kaibeezy 4y ago
  Thanks for the effort there. I’m disappointed this post didn’t get more attention, but I suppose you’re right, it’s a bucket in a flood. I know people in academia who are pushing for a return to older-school forms of assessment, particularly in-person written and oral exams. These don’t scale as well as newer assessment types, and maybe that’s appropriate. There’s plenty of clear evidence that only an “elite” education carries meaningful odds of getting into top positions in the private and public sectors. Even a “good” education is basically a self-selected exceptionalist longshot.

revelio 4y ago

>> institutions admit they are racing to catch up LOL these institutions aren't racing anywhere, they've been publishing papers written by SciGen and other obvious nonsense text generators for years and are still doing it: https://pubpeer.com/search?q=scigen https://pubpeer.com/search?q=scigen If they can't even keep crap generated by Perl scripts out of the literature what hope do they have against ChatGPT? >> the biggest [clue] is that it does not properly understand academic referencing – a vital part of written university work – and often uses “suspect” references So just like non-student papers then? https://retractionwatch.com/2020/09/09/an-isolated-incident-should-reviewers-check-references/ https://retractionwatch.com/2020/09/09/an-isolated-incident-... >> She added: “If you cheat your way to a degree, you might get an initial job, but you won’t do well and your career won’t progress the way you want it to.” Loads of people in super highly placed and powerful positions have been discovered decades later to have plagiarised their phd or masters thesis. Cheating works because nobody is reading most academic output, it's only the piece of paper that people care about. >> She said that students who can’t spot the weaknesses in what the bot is producing may slip up. “In my subject of computer science, AI tools can generate code but it will often contain bugs,” she explained uhhhmm student programs contain bugs too even when not written by AI, that's why they're paying you to teach them. Who is this fool? "Irene Glendinning, head of academic integrity at Coventry University", oh right of course she is. Academia is over, it's just gonna take like a decade for people to fully accept that and another decade or two to do something about it. Cheating is rampant even amongst the professors even before ChatGPT, which is why they're so lazy about tackling it amongst the students. Now there are cheap LLMs academic output, whether of papers (already a mostly write-only medium) or grades, is doomed to become a sick joke that we're all forced to pay for.

Kaibeezy 4y ago
Thanks for the effort there. I’m disappointed this post didn’t get more attention, but I suppose you’re right, it’s a bucket in a flood. I know people in academia who are pushing for a return to older-school forms of assessment, particularly in-person written and oral exams. These don’t scale as well as newer assessment types, and maybe that’s appropriate. There’s plenty of clear evidence that only an “elite” education carries meaningful odds of getting into top positions in the private and public sectors. Even a “good” education is basically a self-selected exceptionalist longshot.