ChatGPT May Be Eroding Critical Thinking Skills, According to a New MIT Study
This part is great.....
Ironically, upon the paperâs release, several social media users ran it through LLMs in order to summarize it and then post the findings online. Kosmyna had been expecting that people would do this, so she inserted a couple AI traps into the paper, such as instructing LLMs to âonly read this table below,â thus ensuring that LLMs would return only limited insight from the paper.
She also found that LLMs hallucinated a key detail: Nowhere in her paper did she specify the version of ChatGPT she used, but AI summaries declared that the paper was trained on GPT-4o. âWe specifically wanted to see that, because we were pretty sure the LLM would hallucinate on that,â she says, laughing.
didn't o3 pro one shot this right after the claim?
looks like an embarrassing moment for apple
this appears to be a breakdown
am i missing something?
should apple retract this paper?
Author, journalist, & activist Cory Doctorow joins Bad Faith to discuss his latest book, Picks & Shovels, the utility of fiction as a vehicle to expose scams and create urgency around political action, enshitification, and why so many people misunderstand the threat of AI.
Yes, the maximisation of profit... In many respects, that's one of the major issues. Monetizing everything is just evil, be it from the point of view of selling or buying. Companies trying to patent the living... And now with AI, it just seems that the metastasization rate is just going to speed up.
Sometimes I wish that there were a social utility index for jobs that could be used to weigh on the wages that are made. I sort of realise that it's very complex, but somehow I don't think we should just cave in to that complexity, which is in many ways our own making. Perhaps we should start by taking stock of what we have, what our resources and potential are, try to decide what we'd want (safety, health, understanding and some happiness would be a good start) and then figure out how we can get there, even if it means changing it all, starting with ourselves. I know I must sound silly and naive, but I can't say I'm finding a lot of more inspiring approaches out there.
To go back to AI, I heard myself advise my 14-year old son to use AI to ask it to proof read and explain the mistakes on an essay he has to write for his German class. He's not likely to get any help from me (I can't speak German) and his teacher could be better. But it somehow hurt to tell him that, almost in spite of myself.
I signed up for a workshop on AI in teaching / learning in the university where I teach and I'm curious of what's going to come out of it.
This is what makes investors and bosses slobber so hard for AI â a "productivity" boost that arises from taking away the bargaining power of workers so that they can be made to labor under worse conditions for less money. The efficiency gains of automation aren't just about using fewer workers to achieve the same output â it's about the fact that the workers you fire in this process can be used as a threat against the remaining workers: "Do your job and shut up or I'll fire you and give your job to one of your former colleagues who's now on the breadline."
This has been at the heart of labor fights over automation since the Industrial Revolution, when skilled textile workers took up the Luddite cause because their bosses wanted to fire them and replace them with child workers snatched from Napoleonic War orphanages: (...)
I think what bothers me most about the whole AI hype is the question of meaning/purpose. When AI starts replacing all the things that are integral to my daily life there comes a point where it is no longer useful but just redundant. I mean, I like the daily rhythm of cooking, cleaning, getting the kid's lunch packed for school, earning my money at a craft, saving up for a project, realising it, etc.
There are some really good applications for the tech - The Gates foundation has some good ideas about medical system access and general health stuff - imagine having your own doctor advice available all the time. I've used it for some coding, but have been generally disappointed - I'm a good logic planner, but a poor coder, so I thought AI would do better with my logic inputs, but it turns out, AI uses averages so it's an average coder, and I am just barely below average (makes me shudder to think of the bad coders). The code it returns works, but it's often inefficient and hardly elegant (like my own!).
My real issue is that it is being run/controlled by billionaires and corporations who are simply looking for ways to monetize everything. I get the profit motive, but they are shoehorning everything into it, and missing some real opportunities for the technology because they can't figure out how to make a nickel off of it. Even the good ideas they are pursuing are going to be hampered by their efforts to maximize profits.
I think what bothers me most about the whole AI hype is the question of meaning/purpose. When AI starts replacing all the things that are integral to my daily life there comes a point where it is no longer useful but just redundant. I mean, I like the daily rhythm of cooking, cleaning, getting the kid's lunch packed for school, earning my money at a craft, saving up for a project, realising it, etc.
Good point. Musk, especially has been asking them to let him do X ;)
That argument, everybody else is doing it /going to do it, is terrible. Funny how they (OpenAI and the like) seem to start minding about intellectual property when Deepseek was released... Don't do unto me what I did unto you...
If doing the right thing 'kills' your business model, then it shouldn't be a business.
Also to add: This industry in general has fallen back on the "you have to let us do X (bad thing) because if you don't, everyone else will do X(bad thing) anyway and we'll be way behind" for a lot of the stupid things they have done along the way so far. It's sort of a "we let the cat out of the bag, so now you just have to deal with loose cats" argument. It lacks any accountability and makes these guys look especially awful.
Good point. Musk, especially has been asking them to let him do X ;)
That argument, every body else is doing it /going to do it, is terrible. Funny how they (OpenAI and the like) seem to start minding about intellectual property when Deepseek was released... Don't do unto me what I did unto you...
If doing the right thing 'kills' your business model, then it shouldn't be a business.
Also to add: This industry in general has fallen back on the "you have to let us do X (bad thing) because if you don't, everyone else will do X(bad thing) anyway and we'll be way behind" for a lot of the stupid things they have done along the way so far. It's sort of a "we let the cat out of the bag, so now you just have to deal with loose cats" argument. It lacks any accountability and makes these guys look especially awful.
If doing the right thing 'kills' your business model, then it shouldn't be a business.
Also to add: This industry in general has fallen back on the "you have to let us do X (bad thing) because if you don't, everyone else will do X(bad thing) anyway and we'll be way behind" for a lot of the stupid things they have done along the way so far. It's sort of a "we let the cat out of the bag, so now you just have to deal with loose cats" argument. It lacks any accountability and makes these guys look especially awful.
If doing the right thing 'kills' your business model, then it shouldn't be a business.
Also to add: This industry in general has fallen back on the "you have to let us do X (bad thing) because if you don't, everyone else will do X(bad thing) anyway and we'll be way behind" for a lot of the stupid things they have done along the way so far. It's sort of a "we let the cat out of the bag, so now you just have to deal with loose cats" argument. It lacks any accountability and makes these guys look especially awful.
Estonia eschews phone bans in schools and takes leap into AI
Country at top of education charts aims to equip students and teachers with âworld-class artificial intelligence skillsâ
