AI Might Start Selling Your Choices Before You Make Them, Study Warns (courthousenews.com) 36
AI ethicists are cautioning that the rise of AI may bring with it the commodification of even one's motivations. From a report: Researchers from the University of Cambridge's Leverhulme Center for the Future of Intelligence say -- in a paper published Monday in the Harvard Data Science Review journal -- the rise of generative AI, such as chatbots and virtual assistants, comes with the increasing opportunity for persuasive technologies to gain a strong foothold.
"Tremendous resources are being expended to position AI assistants in every area of life, which should raise the question of whose interests and purposes these so-called assistants are designed to serve," Yaqub Chaudhary, a visiting scholar at the Center for Future of Intelligence, said in a statement. When interacting even causally with AI chatbots -- which can range from digital tutors to assistants to even romantic partners -- users share intimate information that gives the technology access to personal "intentions" like psychological and behavioral data, the researcher said.
"What people say when conversing, how they say it, and the type of inferences that can be made in real-time as a result, are far more intimate than just records of online interactions," Chaudhary added. In fact, AI is already subtly manipulating and influencing motivations by mimicking the way a user talks or anticipating the way they are likely to respond, the authors argue. Those conversations, as innocuous as they may seem, leave the door open for the technology to forecast and influence decisions before they are made. "We caution that AI tools are already being developed to elicit, infer, collect, record, understand, forecast, and ultimately manipulate and commodify human plans and purposes," Chaudhary said.
"Tremendous resources are being expended to position AI assistants in every area of life, which should raise the question of whose interests and purposes these so-called assistants are designed to serve," Yaqub Chaudhary, a visiting scholar at the Center for Future of Intelligence, said in a statement. When interacting even causally with AI chatbots -- which can range from digital tutors to assistants to even romantic partners -- users share intimate information that gives the technology access to personal "intentions" like psychological and behavioral data, the researcher said.
"What people say when conversing, how they say it, and the type of inferences that can be made in real-time as a result, are far more intimate than just records of online interactions," Chaudhary added. In fact, AI is already subtly manipulating and influencing motivations by mimicking the way a user talks or anticipating the way they are likely to respond, the authors argue. Those conversations, as innocuous as they may seem, leave the door open for the technology to forecast and influence decisions before they are made. "We caution that AI tools are already being developed to elicit, infer, collect, record, understand, forecast, and ultimately manipulate and commodify human plans and purposes," Chaudhary said.
Comedians predicted this (Score:2)
So, an AI version of hover-to-click? (Score:1)
Gaslight (1944) (Score:3)
Best way to poison AI? (Score:5, Insightful)
It occurs to me that there might be a way to feed LLMs enough false and contradictory data to make them both less reliable, and less effective at anticipating and manipulating people's responses and actions. We need to be engaging in this kind of guerilla warfare if we're going to hinder the corporatocracy's efforts to make the general population more cattle-like than we already are. Sadly, I don't see that happening.
Getting old kinda sucks, but there are times when I envy those who are even older than I am. They may miss a lot of the brutally dystopian shit that I foresee happening in the next decade or two. And a lot of that shit will be excreted by that thing - whatever it is - that we miscall "AI".
Re: (Score:2)
Poison the data pool. Start now and be relentless. I wish I knew enough to write a "How To" book on the subject.
Re: (Score:1)
Re: (Score:2)
I want an AI assistant that talks to corporate AI for me.
Maybe I could tell it that I want some product for under X Euro, and it can keep negotiating day and night until it gets me one.
Re: (Score:2)
Here is a start [amazon.com] to poisoning your Amazon history. Oh wait, this is slashdot, you've already hit that link.
Might want to consider just what your poison contains.
Re: (Score:2)
Here is a start [amazon.com] to poisoning your Amazon history. Oh wait, this is slashdot, you've already hit that link.
Might want to consider just what your poison contains.
No, I didn't click on the link - I hover over any link I'm interested in and read the URL first. I have no clue what "Passion Body Glide Natural Water" is, and hope never to find out. I did notice that you scrubbed the tracking crap from the end of the URL though - thanks. I always delete tracking from URLs whenever I can.
You've raised an interesting point. I spend a fair amount of time on 'Zon, and I could start clicking on random shit just to skew my profile. I could do the same on YouTube - I spend way t
Re: (Score:2)
It occurs to me that there might be a way to feed LLMs enough false and contradictory data to make them both less reliable, and less effective at anticipating and manipulating people's responses and actions. We need to be engaging in this kind of guerilla warfare if we're going to hinder the corporatocracy's efforts to make the general population more cattle-like than we already are. Sadly, I don't see that happening.
Getting old kinda sucks, but there are times when I envy those who are even older than I am. They may miss a lot of the brutally dystopian shit that I foresee happening in the next decade or two. And a lot of that shit will be excreted by that thing - whatever it is - that we miscall "AI".
I'm kinda with you on the getting old points. In a lot of ways, I wish I were checking out sooner just to avoid watching society fall further and further into dystopia. We've really let corporatism take over in the west, and have placed such a high value of it that we're letting everything else fall away in pursuit of it. The push for AI everywhere is one of the more brutally stupid aspects of it. It's not helpful to us, the 'users.' It's only helpful to the data aggregators, and the people who want to use
Re: (Score:2)
Won't they be disappointed when they make the wrong prediction? And then what will they do with the sunk cost on AI bullshit?
Citizen: You have been found in violation of the AI predictive algorithm. Report to the reeducation center immediately. You're lack of compliance is deemed harmful to the AI congregate and must be dealt with swiftly. Do not disclose this breach to anyone on your way to reeducation, or they will be tainted as well. Have a nice day.
Maybe OK (Score:2)
Re: (Score:3)
So what (Score:1)
Commercial and government entities are already guessing my future decisions and conditioning my prospects for loans, educational opportunities, and ability to purchase firearms (for example) on their estimates of my choices.
This process is already opaque and only sort-of-right. Why should I be up in arms over it becoming slightly more sort-of-right and just as opaque?
Re: (Score:3)
AI monitoring us and influencing us is small potatoes compared to the tech that is coming in the future. Cognitive-enhancing implants will make us smarter in various ways, and people will get them because they need them to be competitive in the labor market. They will come with assurances of safety and privacy, which will be lies, and these devices will be used by the elite shadow government to control us. By the time anyone figures this out, it will be far too late.
Then AI will literally be directing ou
Fold, Spindle, Mutilate. (Score:2)
Some of you may be old enough to remember the words "Do not Fold, Spindle or Mutilate" on punchcards.
No, I didn't program on punchcard -- i'm old, but not that old. Missed it by 10 or 15 years
BUT -- when I was little, pre-1980's, we got our power bill in PR in the form of an IBM punchcard, and I remember the words.
My most depraved fantasy right now is to take the concept of "AI" and fold it, spindle it, and mutilate it. As it surely will do to us. Terminator does come to mind.
Augmented Idiocy is the name
I worry that those machines teach our children now (Score:4, Insightful)
AI is going to subtly but strongly promote Big Tech to them, lesson after lesson, year after year, and turn them into slave who will forever accept any abuse from tech monopolies, and go with whatever they say without a second thought. The same way McDonald's subtly created generations of adults addicted to fast food - and will never be held accountable for the damage they did and the lives they ruined.
Why wouldn't they? It's not like AI teachers are run by philanthropic non-profits. This is Big Tech's golden occasion to turn the next generation into servile obedient zombies.
If you were of the opinion that human teachers pushed "leftist" ideas into their heads, personally I'd rather kids got that bias into their heads than becoming subserviant to tech monopolies.
Re: (Score:2, Flamebait)
Cannot tell if you are serious, or what really saying. Let me get help ...
Key Points:
- Criticism of AI in Education: The poster is worried that AI-driven teaching tools will subtly indoctrinate children to trust and depend on Big Tech, leading to long-term societal and individual harm.
- Comparison to McDonald's: They draw a parallel between Big Tech's influence and McDonald's alleged role in fostering unhealthy eating habits and societal dependence on fast food. This analogy suggests skepticism about corpo
Re:I worry that those machines teach our children (Score:4, Insightful)
And here you've demonstrated exactly what I'm talking about:
I assume you obtained that text vomit from an AI analyzing my post - which was VERY obviously not ironic, sarcastic or jocular in any way.
When the next generation of school children is drilled to always take the lazy route and go by the pre-chewed AI interpretation of everything, they'll read your vomit and they'll partly dismiss my critique of Big Tech as possibly a joke. And they'll probably dismiss Slashdot as a forum that's not to be taken too seriously too.
And if I was whichever tech monopoly is running that particular AI teacher, it would be in my corporate interest to steer school kids to dismiss posts and forums that are critical of Big Tech.
Thank you for the demonstration.
Re: (Score:1)
OK then, here is my N.I take. You have started with a kernel of truth, genuine issues such as obesity, and wrapped conspiracy theories around it.
Primitive man, looking at the marvels of the world, invented a maker, an all-powerful entity in their own image who designed and planned every mountain and insect. We now know that these result from geological and evolutionary processes, with no knowledge of where they are going.
Similarly, observing the problems of our society, you invent a po
Re: (Score:3)
People eat more because food is tasty and easily available
You are quite the shill.
People eat more when they're hooked young to addictive food that's high in fat and sugar. And hooking them young has been a cornerstone [researchgate.net] of McDonald's corporate strategy for decades.
Companies in our system prosper when they sell people what they want, quickly and conveniently.
Most people who try crack cocaine really REALLY want crack cocaine as quickly and conveniently as possible. And yet it's banned because it's bad for people. Just like junk food and easy, lazy tech products laced with subtle pro-corporate messages.
Re: (Score:1)
Ah, the ad homenim is the standard response of the conspiracy theorist.
Surveillance Capitalism (Score:2)
This is just the logical next step for Surveillance Capitalism [wikipedia.org]
TLDR: Surveillance Capitalism is the amassing by Big Tech companies of vast quantities of individual behavioral data (usually without the individual's knowledge or consent) and, by data mining it, predicting and even directing future behavior for profit. This involves web tracking, app tracking, social media usage, data from smart devices, and more.
People's AI usage will just be another data stream for the prediction and influencing algorithms.
This should be illegal. (Score:4, Interesting)
Combine this with custom per-user for online shopping for disastrous results. With digital price tags and user profiling, an aisle in store can change it's prices in order extract the most from every customer. People will find prices rising for everything they are about to want or need, just they need it.
When you don't need it, bottled water will be pennies. When you're thirsty, the prices will jump to $10. Hoarding can also be predicted, so the prices will be low when you're broke, and prices will rise on payday for anything you might want to buy in advance. There will be no winning against AI.
Re: (Score:3)
A great example was a picture a friend sent me from her new phone. A cat swimming with a fish it had caught. Lovely photo that on the face of it was real. The cat looked perfect, with the just the head sticking out of the water and water had the ripples with the expected reflections. At first glance just a normal pretty photo. One catch, the cat's fur was dry, right down the water line. Anyone who as seen a real cat in water know
Look at Windows to see how this will be accepted (Score:4, Interesting)
Likewise when Big Tech offers you an AI assistant you should not act surprised if it is tuned to server the needs of Big Tech ahead of your own needs. The solution is the same, an open source AI that is trained to be an actual assistant. Of course to use such an AI assistant people will have to seek it out and set it up themselves, so you know the masses will go with the Big Tech's solution of "stop thinking for yourself, give us some money and we will take care of everything for you".
In short "no news here, move along". AI will start selling your choices before you make them because it makes money for companies and the average person simply doesn't care.
Do I have to watch the ads? (Score:2)
Re: (Score:2)
That is somewhat of a nice point you have there.
And as the answer to any question in a headline is to be answered with 'NO!'...
In any case, It would reduce or even eliminate the need to make any commercial video, or billboards, or neon/led-signs for any corporation or store or restaurant or whatever.
As your AI "assistant" will lead your carcass to the correct places at the time it chooses for you, from the day you are born to the day future generations will die.
From there it is not too difficult for the AI
guy man. eepa (Score:1)
The Force can have a strong influence... (Score:2)
...on the weak minded.
This is not unique to AI, nor is it a potential future. It is the past. The problem is people have been encouraged not to have a mind of their own for years, by other people. That campaign (through technology) predates AI by at least a decade, is gradually getting worse, would be and is happening without AI, and is the intent and practice of all social media platforms. These are, in fact, psi-ops programs with psychologists on staff trying to drive addiction, increase self-doubt in ord
No Funny (Score:2)
Disappointed. As usual.
Now I know (Score:2)