Test scores across OECD countries peaked around 2012 and have declined since. IQ scores in many developed countries appear to be falling after rising throughout the twentieth century. Nataliya Kosmyna at MIT’s Media Lab began noticing changes around two years ago when strangers started emailing her to ask if using ChatGPT could alter their brains. She posted a study in June tracking brain activity in 54 students writing essays. Those using ChatGPT showed significantly less activity in networks tied to cognitive processing and attention compared to students who wrote without digital help or used only internet search engines. Almost none could recall what they had written immediately after submitting their work. She received more than 4,000 emails afterward. Many came from teachers who reported students producing passable assignments without understanding the material. A British survey found that 92% of university students now use AI and roughly 20% have used it to write all or part of an assignment. Independent research has found that more screen time in schools correlates with worse results. Technology companies have designed products to be frictionless, removing the cognitive challenges brains need to learn. AI now allows users to outsource thinking itself.
Looking around at my family, neighbors and coworkers getting hoodwinked by AI left and right… yeah that tracks.
Chatgpt helped me build my garden beds, build a lamp out of an old telephone, learn Japanese grammar quirks, improve lacklustre recipes, look after my plants and stop my dog from barking. That’s this month.
I could have gone to Reddit or blogs and scrounged around for hours/days to find that info, but chatgpt had already aggregated it and neatly presented it.
The Luddites who refuse to use AI are simply that. You either don’t understand it’s weaknesses and learn to work around them, or you’re becoming your techphobic grandparents.
The resources that the LLM you used are pulling from will cease to exist and the LLM will continue to degrade in usefulness as a result the longer people like you do not engage the actual content being collated by the LLM. This is a zero sum game. The LLM is not sustainable long term for various reasons, up to and including how it’s funded and the fact that its data set is dependent on the websites and resources of other sites where humans engage with each other and the LLM is actively choking off those resources.
Fair enough, so why not use it while it’s there?
Because once you recognize that using it is helping along the death of the internet, you start to maybe sort of not want to contribute to the sites you like visiting being scraped for information that when presented to you by a fancy statistics algorithm may not even be correct.
But also because I like learning things and the process of gathering information and vetting it is both easier and better in quality than can be received from and LLM. Or if it isn’t, a simple search engine query will suffice.
I’m also going to point out that this isn’t even the only thing wrong with LLM’s.
When this bubble bursts going to wipe out the 401K’s and IRA’s of Americans saving for retirement (the ones that are lucky enough to be able to save) and tank the economy. The largest companies in the world are playing a stock market shell game with other people’s money and it’s gonna end badly.
I’ve also used it successfully for those kinds of special cases - particularly translating complicated medical documents back and forth to Japanese due to my wife’s treatment.
But I think the caution here is overreliance. Using it in a university setting, where you feed it everything you were supposed to read and understand, and having it write down all the analysis that you were meant to analyze, and what have you personally gained as a result? The article cites students who couldn’t even recall what they’d “written” after submitting an assignment.
You can use it as a tool, or you can use it as a crutch. If you outsource your whole thought process to a computer, I can see the detriment.
This is the same with literally any tool. People were complaining about computer use decades ago for fear we would forget how to write by hand. Same thing with writing in general and memory.
You’re proving the point, though. People’s ability to write by hand has indeed deteriorated. Literacy has indeed reduced the need for and intensity of memorisation - and having stuff memorised is useful. What skill will AI cause to atrophy? Is that skill merely like handwriting, or something more?
There’s a difference between being able to think for yourself and being able to write something by hand no?
Even If I became completely unable to write with my hands due to some injury I would still have my own thoughts and ideas about the world; I know people that have effectively stopped having their own opinions because they just ask ChatGPT or another LLM to have an opinion for them. We literally have people using something owned by corporations deciding for them how they should feel about different topics. If you don’t understand why that’s an issue I don’t think we live in the same reality.
Except I’m watching people literally use LLMs as a replacement for their critical thinking skills. There is absolutely a difference.
There is no difference. You are just reading about it now. This shit has been happening in one way or another for 40 some odd years that ive been watching and for centuries before and will continue to happen. Everyone just likes to believe that whatever they notice must be new or novel. Its not.
If you’re using AI as a more powerful search engine, more power to you, that’s IMO how it should be used.
The problem is too many people use it to avoid learning and critical thinking, because it’s much easier.
We are living in the golden age of the MAGAt.
