Skip to main content

mastodon - Link to source
Recently I was at a party of a young friend (<30) and all of his friends used #ChatGPT and other #LLM to „research“. They also use it to write emails in business or at uni but they „curate“. One point was that it saves them time when having to answer in complete sentences while using the right tone.
We discussed a lot and I left the party quite pessimistic.
“#AI“ may have done more damage than I expected, already.
#ai #chatgpt #llm
in reply to chris@strafpla.net

chris@strafpla.net
mastodon - Link to source
chris@strafpla.net
• •
Another first was receiving a message of an acquaintance I lost contact with. He’s in a difficult situation, healthwise, and it seems he’s still brooding about something stupid someone carelessly said to him 3 years ago. His latest message ended with (translated) “Should I write this more concise, emotional or neutral”?
I was about to reply with “Just write like you do, I know you, I will understand. You’re not a robot”.
Then it clicked.
“#AI”
#ai
This entry was edited (22 hours ago)
in reply to chris@strafpla.net

chris@strafpla.net
mastodon - Link to source
chris@strafpla.net
• •
I don’t think that we’re currently losing the capability to write code, only, but I think that we’re losing the capability to WRITE at all and, because a lot of our communication and discourse is in writing, the capability to think coherently.
This entry was edited (1 day ago)
in reply to chris@strafpla.net

chris@strafpla.net
mastodon - Link to source
chris@strafpla.net
• •
I had to think about this message from him for a while - until now - to understand what my problem is.
He may have run out of spoons to work on his message but still felt the pressure to reply “properly”. But just writing “Excuse my lack of spoons, 2025 sucked, here’s a list of vomit in no particular order” would have been more genuine than having a PR-Robot write a reply.
This entry was edited (22 hours ago)
in reply to chris@strafpla.net

Julian Schwarzenbach
mastodon - Link to source
Julian Schwarzenbach
• •
Skills loss/ failure to gain skills will be a significant long term impact of AI. If you have not learnt to think/ analyse/ evaluate and rely on AI to do this for you, then some key parts of your mental toolkit will not be developed. Those who do things the hard way, even if they end up with lower grades, are more likely to succeed in the long run
in reply to Julian Schwarzenbach

chris@strafpla.net
mastodon - Link to source
chris@strafpla.net
• •
@jschwa1 You put down my exact fears into words. I’m not sure if those who succeed will be enough to keep civilization running. The people at the party were mostly students and young academics or PhDs.
@Julian Schwarzenbach
in reply to chris@strafpla.net

Die Katja mit den roten Haaren
mastodon - Link to source
Die Katja mit den roten Haaren
• •
Local German politicians use LLMs, too. One gave a speech about the history of a local feast. Well, it was bullshit. It came from AI, which, of course, had no sources to say anything about such a specific topic and hallucinated a bit.
in reply to Die Katja mit den roten Haaren

chris@strafpla.net
mastodon - Link to source
chris@strafpla.net
• •
@KatjaGausMimO I feel responsible for what I’m saying (and what I’m writing) so this idea is completely absurd to me.
Of course we appreciated speech writers in ancient times, already, but I guess that they understood their job and responsibilities. LLM don’t understand anything.
Is it even writing when there is no thought, or is it still a Rorschach test?
@Die Katja mit den roten Haaren
⇧