It's of my opinion that there is going to be a market for wrangling AI content from a consumer perspective to help maintain human-to-human knowledge transfer. I just have no idea what that looks like.
Honestly the biggest way in which LLMs have changed society is in the desperate, almost pathetic way every every business leader, career influencer, advice guru insists that they must use AI, that you should "learn" AI, that AI is taking over.
Anyway, in terms cultural change, I think the emerging image and video models will be a lot more disruptive. Text has been easy to fake for a while now, and barely gets people's attention anymore.
The author needn't regret not publishing this two years ago, it's a thought that had occurred to pretty much everyone long before then. It's just not clear that anything can be done to stop the snowball from gathering speed.
In the long run, the internet will be so riddled with trash that no one will trust it. Instead people will turn to authorities they trust for the truth the same way they did with encyclopedias and local papers. Information provenance will be a massive market.
The return to that world will be very painful and chaotic however.
> Therefore, increasing proportions of people consuming text online will be unwittingly mind-controlled by LLMs and their handlers.
The "and their handlers" part is the part I find frightening. I would actually be less concerned if the AIs were autonomous.
Reminds me of a random podcast I heard once where someone was asked: "if you woke up in the middle of the night and saw either a random guy or a grey alien in your bedroom, which would scare you more?" The person being interviewed said the dude, and I 100% agree. AI as proxy for oligarchs is much scarier than autonomous alien AI.
I think the solution is to not aim to go online to "consume content". Instead, go online to learn new techniques and investigate well-reasoned opinions.
Generic "content" is that which fills out the space between the advertisements. That's never been good for you, whether written by humans or matrix multiplication.
I have an issue with "inherently superior ... by dopamine output" part. It's the foundation of the whole article/worry but it's not supported by anything (The Matrix quotes don't count), making the whole article hang on a dubious premise of impending doom that is not shown to exist in reality.
for fear of being overly utilitarian here its not really an issue that people are manipulatable but that they are manipulated into doing the wrong things (consumerism, political divide-and-conquer strategies)
and rejecting manipulation from a deontological stance reduces agency and output for doing good in the real world
manipulation = campaigns = advertisements = psyops (all the same, different connotations)
Television and the commercial Internet are optimized to consume as much life as possible so that part of the captured attention can be auctioned to advertisers and other propagandists for pennies a minute. Returning to doing the same thing but Certified With No AI™ doesn't substantially reduce the badness of the thing.
Before LLMs were mainstream, rationalists and EA types would come on Hacker News to convince people that worrying about how "weak" AI would be used was a waste of time, because the real problem was the risk of "strong" AI.
Those arguments looked incredibly weak and stupid when they were making them, and they look even stupider now.
And this isn't even their biggest error, which, in my opinion, was classifying AI as a bigger existential risk than climate change.
An entire generation of putatively intelligent people lost in their own nightmares, who, through their work, have given birth to chaos.
No, I shall not "be worried." Unless you are some spineless blob of agencyless ooze... you should be able to parse reality in such a way that your day not need to be filled with worrying
> Increasing numbers of people who consume content on the Internet will completely sacrifice their ability to think for themselves.
Bless the author's heart.
All the major social media apps have been doing machine learning-driven getNext() for years now. Well before LLMs were even a thing. The Youtube algorithm was doing this a decade ago. This isn't on the horizon, we've already drowned in it.
Most comments seem to agree with the article, and I don't quite understand why.
People have been manipulated since forever, and coerced before that. You used to be burned or hanged if your opinions differed even a little from orthodoxy (and orthodoxy could change in a span of a couple of years!)
AI slop is mostly noise. It doesn't manipulate, it makes thinking a little more difficult. But so did TV.
> What am I personally going to do about this? Well, to start, I’m going to start taking content way less seriously unless it was created before 2022
There's an old fable about this, The Boy Who Cried "Wolf" about people adapting to false claims. They just discount the source, which is what is going to happen with social media once it is dominated by AI slop. Nobody will find it worth anything anymore, and the empires will melt down. I'm not on any of the big social sites, but I'm already watching a lot less on YouTube, basically only watching channels that I know to be real people. My other recommendations are mostly AI garbage now, outside of that.
Because you are I will not be able to tell whether something is machine- or human-generated, and the machine generated stuff will get more clicks than the human generated stuff, it’s likely that the majority of popular online content (and even printed content post-2023) will have been created by AI (and perhaps solely by AI).
Sorry, but when you make claims like this, it just tells me that you are not very familiar with popular culture. Most people hate AI content and at best find it a meme-esque joke. And young people increasingly get their news from individuals on TikTok/YouTube/etc. - who are directly incentivized to be as idiosyncratic and unique (read: not like AI) as possible in order to get followers. Platforms like YouTube do not benefit from their library being entirely composed of AI slop, and so will be implementing ways to filter AI content from "real people" content.
Ultimately AI tools are mostly going to be useful in situations where the author doesn't matter: sports scores, stock headlines, etc. Everything else will likely be out-competed by actual humans being human.
I think humans having a platform to tell the masses to "be worried" is as troublesome these days as AI content. mass media that can be manipulated has been around for 100 years. I don't think AI is unique.
LLMs are the latest progression in decades of technology and social changes that leave people less connected and less capable in exchange for more comfort. I think it's likely that AI technology eclipses humans at least partially by atrophying our own skills and abilities, particularly 1. our ability to endure discomfort in service of a goal and 2. our capacities to make decisions.
I don't really know what to do about it, even with ground rules of engagement, we all still need to participate in a larger culture where it seems like it's a runaway guarantee that LLMs erode more critical skills that leave us with less and a handful of companies who develop this tech with more.
I'm slowly changing my life around what LLMs tell me, but not necessarily in the ways you'd expect:
1. I have a very simple set of rules of engagement for LLMs. For work, I don't let LLMs write code, and I won't let myself touch an LLM before suffering on a project for an hour at least.
2. I am an experienced meditator with a lot of experience in the Buddhist tradition. I've dusted off my Christian roots, and started exploring these ideas with new eyes, partially from a James Hillman-esq / Rob Burbea Soulmaking Dharma look. I've found a lot of meaning in personal fabrication and myth, and my primary practice now is Centering Prayer.
3. I've been working for a little while on a personal edu-tech idea with the goal of using LLM tech as an auxiliary tech to help people re-develop lost metacognitive skills and not use LLMs as a crutch. I don't know if this will ever see the light of day, it is currently more of a research project than anything, and it has a certain kind of iconoclastic frame like Piotr Wozniak's around what education is and what it should look like.
26 comments
[ 3.0 ms ] story [ 48.7 ms ] threadThe people involved in making these decisions deserve to be locked up for life, and I'm sure they will be eventually.
A more immediate notion, perhaps, but definitely not scarier than human extinction.
Anyway, in terms cultural change, I think the emerging image and video models will be a lot more disruptive. Text has been easy to fake for a while now, and barely gets people's attention anymore.
The return to that world will be very painful and chaotic however.
The "and their handlers" part is the part I find frightening. I would actually be less concerned if the AIs were autonomous.
Reminds me of a random podcast I heard once where someone was asked: "if you woke up in the middle of the night and saw either a random guy or a grey alien in your bedroom, which would scare you more?" The person being interviewed said the dude, and I 100% agree. AI as proxy for oligarchs is much scarier than autonomous alien AI.
Generic "content" is that which fills out the space between the advertisements. That's never been good for you, whether written by humans or matrix multiplication.
and rejecting manipulation from a deontological stance reduces agency and output for doing good in the real world
manipulation = campaigns = advertisements = psyops (all the same, different connotations)
Those arguments looked incredibly weak and stupid when they were making them, and they look even stupider now.
And this isn't even their biggest error, which, in my opinion, was classifying AI as a bigger existential risk than climate change.
An entire generation of putatively intelligent people lost in their own nightmares, who, through their work, have given birth to chaos.
The thing to ask yourself: does what I'm reading provide any value to me? If it does, then what difference does it make where it comes from.
Bless the author's heart.
All the major social media apps have been doing machine learning-driven getNext() for years now. Well before LLMs were even a thing. The Youtube algorithm was doing this a decade ago. This isn't on the horizon, we've already drowned in it.
People have been manipulated since forever, and coerced before that. You used to be burned or hanged if your opinions differed even a little from orthodoxy (and orthodoxy could change in a span of a couple of years!)
AI slop is mostly noise. It doesn't manipulate, it makes thinking a little more difficult. But so did TV.
There's an old fable about this, The Boy Who Cried "Wolf" about people adapting to false claims. They just discount the source, which is what is going to happen with social media once it is dominated by AI slop. Nobody will find it worth anything anymore, and the empires will melt down. I'm not on any of the big social sites, but I'm already watching a lot less on YouTube, basically only watching channels that I know to be real people. My other recommendations are mostly AI garbage now, outside of that.
Sorry, but when you make claims like this, it just tells me that you are not very familiar with popular culture. Most people hate AI content and at best find it a meme-esque joke. And young people increasingly get their news from individuals on TikTok/YouTube/etc. - who are directly incentivized to be as idiosyncratic and unique (read: not like AI) as possible in order to get followers. Platforms like YouTube do not benefit from their library being entirely composed of AI slop, and so will be implementing ways to filter AI content from "real people" content.
Ultimately AI tools are mostly going to be useful in situations where the author doesn't matter: sports scores, stock headlines, etc. Everything else will likely be out-competed by actual humans being human.
Really? Because I still see blatantly obvious AI-generated results in web searches all the time.
I don't really know what to do about it, even with ground rules of engagement, we all still need to participate in a larger culture where it seems like it's a runaway guarantee that LLMs erode more critical skills that leave us with less and a handful of companies who develop this tech with more.
I'm slowly changing my life around what LLMs tell me, but not necessarily in the ways you'd expect:
1. I have a very simple set of rules of engagement for LLMs. For work, I don't let LLMs write code, and I won't let myself touch an LLM before suffering on a project for an hour at least.
2. I am an experienced meditator with a lot of experience in the Buddhist tradition. I've dusted off my Christian roots, and started exploring these ideas with new eyes, partially from a James Hillman-esq / Rob Burbea Soulmaking Dharma look. I've found a lot of meaning in personal fabrication and myth, and my primary practice now is Centering Prayer.
3. I've been working for a little while on a personal edu-tech idea with the goal of using LLM tech as an auxiliary tech to help people re-develop lost metacognitive skills and not use LLMs as a crutch. I don't know if this will ever see the light of day, it is currently more of a research project than anything, and it has a certain kind of iconoclastic frame like Piotr Wozniak's around what education is and what it should look like.
The article never actually backs this up.