Laut einer neuen MIT -Studie kann ChatGPT kritische Denkfähigkeiten untergraben
https://time.com/7295195/ai-chatgpt-google-learning-school/
Laut einer neuen MIT -Studie kann ChatGPT kritische Denkfähigkeiten untergraben
https://time.com/7295195/ai-chatgpt-google-learning-school/
12 Kommentare
**Submission Statement**
In this new study from MIT’s Media Lab (not yet peer-reviewed & small sample size), 54 subjects—18 to 39 year-olds from the Boston area—were placed into three groups, and asked them to write several SAT essays using OpenAI’s ChatGPT, Google’s search engine, and nothing at all. ChatGPT users had the lowest brain engagement and „consistently underperformed at neural, linguistic, and behavioral levels.“ With ChatGPT and other generative artificial intelligence chatbots becoming more and more mainstream, this article brings attention to the effects that this technology might be having how cognitive development especially in younger people.
Let’s be honest, that was in a steep decline long before ChatGPT…
Seems like an obvious response to me.
Not saying it’s a good or bad thing, but it’s like saying people stopped walking long distances as much once transportation became common place.
It bugs me that this study keeps getting posted while failing to highlight the most important finding:
„The paper has not yet been peer reviewed, and its sample size is relatively small.“
If you use it as something to replace you thinking then yes. If you use it a tool, should make it better because the amount of ways you can solve problems, enhances.
I think it’s just a catalyst for what was already happening. The difference in the way I (my generation) was taught to parse information from a passage or a story, and the way my daughter (her generation) was taught are a large piece of this puzzle. She’s getting a lot of real life experience now at work, but schools are failing children with basic reading and language comprehension skills. I’m talking about the most basic building blocks, like phonics, spelling, vocabulary, etc. I blame a lot of it on AR and whole language.
AR (accelerated reading) turns reading into a chore, and doesn’t really foster any understanding of the material. I am an avid reader and tried to help her to learn to enjoy it as best I could. We read every night before bed for years, she saw me reading often, and I tried to foster a love for it. If your child already doesn’t enjoy reading for fun (she’s mathy) AR just makes them hate reading even more. I understand that teachers simply do not have the time for individualized instruction but these types of programs are not helping.
I also made sure she had a basic understanding of phonics, because the schools were REALLY into whole language at that point. I feel like whole language as an ADD ON to basic instruction is fine, but it was the primary model for too long. It strips out the structure of learning the language, which is vital to understanding it. Can you tell it still annoys me twenty years later?
I’m optimistic. Humankind is going through a phase of learning to deal with this new technology, and although there are some spams and hiccups along the road, we will come out of it smarter and more aware what constitutes true intelligence, that is NOT ai. Humans will adapt faster than AI will manage to make progress.
Whenever you see articles like this about some shocking new finding, it’s nearly always a preprint, i.e. it hasn’t been peer reviewed yet.
This is easily confirmed by talking to anyone who regularly uses ChatGPT.
The monetary incentives for information are out of whack. Before chat gpt came out, there had already been an explosion of click bait and unproductive content being generated at a mind boggling pace.
I would love to see a reorientation of search algorithms to high quality information, analysis etc.
It would be nice for not every person to need to be an internet ninja to find good content and sources for the projects they are working on and to feed their intellectual curiosity.
There will always be lots of crap on the internet, but the volume problem is threatening to drown us. Furthermore, social media platforms, Amazon, google and other large aggregators have abdicated their responsibility towards promoting quality.
If we were to reorient expectations and incentives towards quality, ChatGPT and llms could be pulling from a smaller number of better sources.
I don’t think AI is an insurmountable issue. I do think it’s very unfortunate that it has come out at the height of inshitification of the internet and when we haven’t yet dealt with the existing problems of tech monopolies serving us plates of garbage.
my maga dad doesn’t like trump anymore after discovering chat gpt so
It’ll be replaced by something else. They said the same shit when I stopped combing through the Encyclopedia of Britannica in the library and started using Google.