I know this is a shitpost but I would like to share two examples that shocked me about how bad AI use is.
One, at a young scholars conference everyone was glued to their computers. When panelists gave their talks then got asked questions I watched them all put the questions into AI before answering. What even is the point of talking if you cannot answer about your own research! AI definitely cannot answer for me, it’s insane.
Two, as I was leaving Japan this week I watched a woman in front of me ask gemini how Japanese people can afford to live and why they shop so much, lol. I just thought, you can literally ask any person you have a brief connection with. But rather than learn something in Japan you ask AI? As her husband comes over she then tells him all about what AI said. They seemed to be about 30.
TLDR: I read over peoples’ shoulders 20-30 years ago olds use AI in very disturbing ways, choosing to regurgitate rather than learn.
Speaks a lot about quality and necessity of research when AI already has all the answers.
Also I wonder if these people can be reasonably called scholars if they don’t have the actual knowledge.
Like, I know this doesn’t sit well with many Lemmy folks, but LLMs do have reasonable use in science. For example, they’re amazing at contextual search, they can find what I’m looking for even if it’s buried into walls of text of an obscure 2003 article as a sidenote. Combined with traditional search, this allows me as a junior scientist to better check what’s already been discovered, and identify real gaps of knowledge, not what I personally didn’t find.
They can also be useful when brainstorming concepts from the fields I had little experience with. Finding proper literature, breaking down jargon, and otherwise streamlining the study process. This allows to spend less time researching the widely known things on the subject, and dedicate more time to actual science, and also helps immensely with interdisciplinary research.
But that’s just that - finding well-known, long discovered things that you could miss. Any meaningful steps beyond it require strong grasp on the subject, and this only comes with processing everything through the actual brain. Miss that - and you’re just a machine parrot.
I know this is a shitpost but I would like to share two examples that shocked me about how bad AI use is.
One, at a young scholars conference everyone was glued to their computers. When panelists gave their talks then got asked questions I watched them all put the questions into AI before answering. What even is the point of talking if you cannot answer about your own research! AI definitely cannot answer for me, it’s insane.
Two, as I was leaving Japan this week I watched a woman in front of me ask gemini how Japanese people can afford to live and why they shop so much, lol. I just thought, you can literally ask any person you have a brief connection with. But rather than learn something in Japan you ask AI? As her husband comes over she then tells him all about what AI said. They seemed to be about 30.
TLDR: I read over peoples’ shoulders 20-30 years ago olds use AI in very disturbing ways, choosing to regurgitate rather than learn.
Speaks a lot about quality and necessity of research when AI already has all the answers.
Also I wonder if these people can be reasonably called scholars if they don’t have the actual knowledge.
Like, I know this doesn’t sit well with many Lemmy folks, but LLMs do have reasonable use in science. For example, they’re amazing at contextual search, they can find what I’m looking for even if it’s buried into walls of text of an obscure 2003 article as a sidenote. Combined with traditional search, this allows me as a junior scientist to better check what’s already been discovered, and identify real gaps of knowledge, not what I personally didn’t find.
They can also be useful when brainstorming concepts from the fields I had little experience with. Finding proper literature, breaking down jargon, and otherwise streamlining the study process. This allows to spend less time researching the widely known things on the subject, and dedicate more time to actual science, and also helps immensely with interdisciplinary research.
But that’s just that - finding well-known, long discovered things that you could miss. Any meaningful steps beyond it require strong grasp on the subject, and this only comes with processing everything through the actual brain. Miss that - and you’re just a machine parrot.