I was talking recently to someone who teaches AI-adjacent courses at a US university (not in a computer science department) and they said that enrollment in their class is lower than expected, which they think is likely due to the severity of the AI backlash among students on campus.
What the tech elite fail to understand is that we are at historic levels of wealth and income inequality. Access to healthcare is determined by one’s employment which makes what I’m about to explain a matter of life and death.
It doesn’t matter if you think it’s all going to work out and AI will bring an unprecedented era of abundance. That is not the current state.
Now what do you think happens when we dramatically expand productivity with AI? Well, we’re already seeing unprecedented layoffs in tech. And it’s easy to draw the conclusion that unless something structural changes all of the productivity gains from AI will go to investors not workers. Leaving said workers without access to healthcare or housing.
And of course let’s not forget that the tech elite in question supported Trump in the last election - someone who has done everything in his power to reduce healthcare access among the low income / unemployed population. This isn’t fucking rocket science guys.
This has mirrored what I've seen in my company. People in the data science/ML part of the company are super excited about AI and are always giving presentations on it and evangelizing it. Most engineers in other areas, though, are generally underwhelmed every time they try using it. It's being heavily pushed by AI "experts" and senior leaders, but the enthusiasm on the ground is lacking as results rarely live up to the extremely rosy promises that the "experts" keep making. Meanwhile, everyone can read the news about layoffs attributed to AI and can see that hiring (especially of junior engineers) has slowed to a trickle. You can only fool people for so long.
The tone deafness of the tech community is so unbearable. Either too on the spectrum, too ambitious (the world is fine cause I’m getting mine), or too isolated from non-tech people, to realise most people despise what they’re creating.
There’s also a lack of willingness to ‘bring along’ the public. It’s just “make the god thing; ask for permission later”.
a person can have full faith in the potential value of ai science and simultaneously have zero faith in the current crop of business stewards of that science.
no one is questioning the underlying model mathematics, they are questioning deceptive & reckless stewards.
[X] Tweets and instagram comments presented as "what society is thinking"
[X] Ties Luigi Mangione and the California warehouse fire to Gen Z discontent (about AI?).
[X] Statistics being used to support the title with little to no regards to continuity: "those respondents who said that AI makes them “nervous” grew from 50% to 52% during the same period" => percentage was 52% in 2023, 50% in 2024 and 52% in 2025, seems mostly flat to me, with the real jump being in 2022-2023 with 39%.
Giant leaps in innovation almost always have a reaction like this.
It's new, people fear it. Sometimes justified, usually not.
People greatly feared the car because of the number of horse-related jobs it would displace.
President Benjamin Harrison and First Lady Caroline Harrison feared electricity so much they refused to operate light switches to avoid being shocked. They had staff turn lights on/off for them.
Looking back at these we might laugh.
We're largely in the same boat now.
It's possible AI will destroy us all, but judging from history, the irrational reactions to something new isn't exactly unprecedented.
Is it irrational to wonder how large swathes of the population will earn a living if their employable skills vanish in a couple of years, with little prospect for retraining into something else that AI hasn't replaced? Is it irrational to wonder what effect an influx of the AI-replaced will have on remaining AI-free fields? Is it irrational to wonder about the psychological impact of work where one simply operates the AI instead of thinking, creating, growing? Is it irrational to wonder if wealth inequality will spiral when these essentially-unobtainable resources are used by a select few to enact the above scenarios?
I can only assume you have easy answers for all of these questions given your casual dismissal of such concerns, likening them to being scared of a light switch.
Paraphrasing the classic, it's not AI that people are unhappy with, it's their life around AI. The world generally appears to have become a harsher and more dangerous place - even though it hasn't. But people and especially tabloid press like finding scapegoats and participating in mass hysteria. The anti-AI hysteria is going to go away soon while AI isn't. It's just another tool, like cars or factories. Granted, it brings some danger, but at the same time it brings overwhelmingly more good.
I think people are really underestimating how poorly today's tweens think of AI. "That looks like chatgpt" is an insult. Kids avoid things because they heard somewhere that AI might have been involved and have a sense that means it is bad or immoral or illegal or cheating in some nebulous way, and it's reinforced by their teachers telling them that using AI for homework is cheating.
I think this next generation is going to come up fundamentally believing that AI is generally a bad thing, and it's going to surprise older people.
I have noticed similar sentiments among some teenagers. It's not a universal sentiment but those who hate AIs really hate them with a passion.
In the meanwhile there is a rising tide of feel good AI content targeted at old people on Facebook. My mother has been sharing with me many "funny videos" that are very obviously AI generated. She evidently does not care, and according what I hear from others she is far from the only old person who gets sucked into "slop." I hesitate to use this word but it captures the feeling too well for me to pass it up.
I don't have data but I sense there is an inverse correlation between age and disgust towards AI generated content.
I'm guessing there's a sizable portion of the HN crowd that are millenials. Millenials who have paid the costs of the Boomer/GenX generations absolute destruction of the "American Dream" for their own benefit. They climbed the ladder and pulled it up behind them leaving millenials holding the bag.
That same set of millenials are now visiting that treatment upon Gen Z. We are building AI that will eviscerate the remaining middle class, raise electricity rates to a level where many people will not be able to power their homes, and poisoning the air and water so that portions of the world will become unliveable.
Gen Z is justified in being upset with millenials. We used to be the victims, but we've become the abusers.
I think it's not that difficult to see why a technology that will likely trigger widespread unemployment during a cost of living crisis, an arms race with China, along with all the alignment concerns, might not be hugely popular with the public.
Maybe I'd be a bit more optimistic if someone could explain a realistic economic scenario for how we're going to transition into our utopian abundant future without a depression or a revolution.
The lack of federal permitting standards for AI data centers is really going to bite the industry in the ass. We also probably need something akin to the WARN Act for AI-related layoffs. (Possibly with multi-year benefits for large companies.)
If "AI" was just free local and open models running on consumer hardware, fewer people would have an issue with it. Which highlights that the issue is with the hyper scalers, the rhetoric, the corporations, the marketing, etc etc.
We are ever so close to nearing the point where 90% of our AI usage can go through providers of open models, who all compete with each other to drive down prices and prevent rug pulls, leaving Dario and Sam holding empty bags.
AI continues to be a stupidly vague term, and the example I keep going back to is present in this article
Meaningful advances in medical diagnosis are not coming from chatbot companies. Some are coming from machine learning methods. Perhaps measuring public sentiment about such a vagary is not a very productive way to quantify anything
That said, I continue to also be frustrated with people using the abstract concept of a new technology as a substitute for the institutions that use that technology to exert power in the world and what they do with that power, which is - as many in the comments already point out - is what the vast majority of people are actually mad about, and right to be
Regardless, I think we are going to see an acceleration of AI research.
I just wish my wife is more serious about camping and learning survival skills. I think Shit is going to hit the fan in the next 5-10 years but she thinks that’s crazy. Oh well maybe I am crazy.
This AI rollout has been fundamentally rushed and fucked from the very beginning and I think the people who are responsible for doing it this way have done more irreparable damage to society than any single group of humans in the entire history of the species, and I mean it.
It’s always only ever about how the new model is faster, better, smarter. Or how the tech will be bringing ruin to the job market and someone should probably do something about that some time soon. Zero efforts to create any sort of educational content - how it even works, how to vet its output, how to have an eye for confabulation, how to use it as thinking enhancement rather than replacement, to keep in mind that it’s trained to please and will literally generate anything to cause users to click the thumbs up button. Nope, it’s just “ModelGPClaude can make mistakes! Better be careful!”
And then everyone’s surprised when an utterly improvident handling of 4o kicks off the biggest concentrated wave of AI psychosis seen yet. Because, surprise! When you give people a model that’s trained to anthropomorphize itself, people who have no idea about any of this tech and have no access to education about any of it might believe it’s more than it is! Boy, who’d’ve thunk; isn’t the world complex?!
This was a symptom of this exact same disease. I have far less worry about the tech and far more worry about how the disconnected venture capital caste is inflicting it upon us.
One of the most hilarious AI-vangelical posts I've seen recently is from Steve Yegg through Simon Willison [0]....
> The TL;DR is that Google engineering appears to have the same AI adoption footprint as John Deere, the tractor company. Most of the industry has the same internal adoption curve: 20% agentic power users, 20% outright refusers, 60% still using Cursor or equivalent chat tool. It turns out Google has this curve too... [0]
Ummmm... Steve. You think Google might be able to figure out a super huge awesome new thing from 1 out of 5 of their employees. Or, given this is a consistent curve across the industry (even at Google)... Maybe AI is only about a fifth as cool and helpful as you and the enthusiasts think it is?
In case you're wondering who they mean by "AI experts", I checked the Pew poll:
> Note: “AI experts” refer to individuals whose work or research relates to AI. The AI experts surveyed are those who were authors or presenters at an AI-related conference in 2023 or 2024 and live in the U.S. Expert views are only representative of those who responded.
> The United States reported the lowest trust in its own government to regulate AI responsibly of any country surveyed, at 31%.
It seems US citizens are really against the current administration, just using the fact that AI investment is intrinsecally connected to it to voice their opposition.
> Country-level expectations follow similar patterns to the earlier sentiment trends.
Nigeria, Japan, Mexico, the United Arab Emirates, South Korea, and India all expected AI to create more jobs than it eliminates, with shares above 60%. The United States and Canada sat at the opposite end, where 67% and 68% of respondents expected AI to eliminate jobs and disrupt industries.
Globally, the disconnect is not growing. It's really just an U.S. problem (spilling to neighbouring Canada too).
So, no luddites in sight, again. It's just a public perception over a polemic topic being leveraged for ideological reasons sinking AI on US only.
It is worth pointing out that we got here despite all of the “alignment” research and safetyism surrounding the models. As it turns out, the models don’t wake up and start destroying things. We knew this all along, but every time a new article came along and anthropomorphized and exaggerated another experiment it fed the clickbait machine.
The fundamental alignment issue is aligning the companies themselves with society, not the models with the companies. Widespread unemployment is not aligned with society, but it is aligned with Anthropic and OpenAI if it makes them rich.
Therefore the only “harms” the companies will take seriously are those which also harm the company. For example reputational harms from enabling scams aren’t allowed.
Perhaps all of this isn’t fair, since companies actively subverted safety research for profitability. But then I would go back to my earlier point of over-indexing on unintended behaviors and under-indexing on intended ones.
Been saying this for a bit but the things I’ve seen associated with AI seem to be the things that it’s pretty mid at. Coding, automated actions etc. I wholeheartedly believe adoption and perception would be better if the things it was amazing at were pushed more.
Take log review for example. Whether it’s admin or security LLMs are incredible at reading awfully formatted logs and even using those to pull meaning from other logs as well. Like turn an hour long log review into a 10 minute log review type thing.
53 comments
[ 3.0 ms ] story [ 54.5 ms ] threadThe kids are alright.
It doesn’t matter if you think it’s all going to work out and AI will bring an unprecedented era of abundance. That is not the current state.
The current state is: Nearly all productivity growth since 1980 has gone to shareholders, not workers: https://www.epi.org/productivity-pay-gap/
Now what do you think happens when we dramatically expand productivity with AI? Well, we’re already seeing unprecedented layoffs in tech. And it’s easy to draw the conclusion that unless something structural changes all of the productivity gains from AI will go to investors not workers. Leaving said workers without access to healthcare or housing.
And of course let’s not forget that the tech elite in question supported Trump in the last election - someone who has done everything in his power to reduce healthcare access among the low income / unemployed population. This isn’t fucking rocket science guys.
There’s also a lack of willingness to ‘bring along’ the public. It’s just “make the god thing; ask for permission later”.
no one is questioning the underlying model mathematics, they are questioning deceptive & reckless stewards.
[X] Tweets and instagram comments presented as "what society is thinking"
[X] Ties Luigi Mangione and the California warehouse fire to Gen Z discontent (about AI?).
[X] Statistics being used to support the title with little to no regards to continuity: "those respondents who said that AI makes them “nervous” grew from 50% to 52% during the same period" => percentage was 52% in 2023, 50% in 2024 and 52% in 2025, seems mostly flat to me, with the real jump being in 2022-2023 with 39%.
It's new, people fear it. Sometimes justified, usually not.
People greatly feared the car because of the number of horse-related jobs it would displace.
President Benjamin Harrison and First Lady Caroline Harrison feared electricity so much they refused to operate light switches to avoid being shocked. They had staff turn lights on/off for them.
Looking back at these we might laugh.
We're largely in the same boat now.
It's possible AI will destroy us all, but judging from history, the irrational reactions to something new isn't exactly unprecedented.
I can only assume you have easy answers for all of these questions given your casual dismissal of such concerns, likening them to being scared of a light switch.
I think this next generation is going to come up fundamentally believing that AI is generally a bad thing, and it's going to surprise older people.
In the meanwhile there is a rising tide of feel good AI content targeted at old people on Facebook. My mother has been sharing with me many "funny videos" that are very obviously AI generated. She evidently does not care, and according what I hear from others she is far from the only old person who gets sucked into "slop." I hesitate to use this word but it captures the feeling too well for me to pass it up.
I don't have data but I sense there is an inverse correlation between age and disgust towards AI generated content.
That same set of millenials are now visiting that treatment upon Gen Z. We are building AI that will eviscerate the remaining middle class, raise electricity rates to a level where many people will not be able to power their homes, and poisoning the air and water so that portions of the world will become unliveable.
Gen Z is justified in being upset with millenials. We used to be the victims, but we've become the abusers.
Maybe I'd be a bit more optimistic if someone could explain a realistic economic scenario for how we're going to transition into our utopian abundant future without a depression or a revolution.
We are ever so close to nearing the point where 90% of our AI usage can go through providers of open models, who all compete with each other to drive down prices and prevent rug pulls, leaving Dario and Sam holding empty bags.
Meaningful advances in medical diagnosis are not coming from chatbot companies. Some are coming from machine learning methods. Perhaps measuring public sentiment about such a vagary is not a very productive way to quantify anything
That said, I continue to also be frustrated with people using the abstract concept of a new technology as a substitute for the institutions that use that technology to exert power in the world and what they do with that power, which is - as many in the comments already point out - is what the vast majority of people are actually mad about, and right to be
I just wish my wife is more serious about camping and learning survival skills. I think Shit is going to hit the fan in the next 5-10 years but she thinks that’s crazy. Oh well maybe I am crazy.
It’s always only ever about how the new model is faster, better, smarter. Or how the tech will be bringing ruin to the job market and someone should probably do something about that some time soon. Zero efforts to create any sort of educational content - how it even works, how to vet its output, how to have an eye for confabulation, how to use it as thinking enhancement rather than replacement, to keep in mind that it’s trained to please and will literally generate anything to cause users to click the thumbs up button. Nope, it’s just “ModelGPClaude can make mistakes! Better be careful!”
And then everyone’s surprised when an utterly improvident handling of 4o kicks off the biggest concentrated wave of AI psychosis seen yet. Because, surprise! When you give people a model that’s trained to anthropomorphize itself, people who have no idea about any of this tech and have no access to education about any of it might believe it’s more than it is! Boy, who’d’ve thunk; isn’t the world complex?!
This was a symptom of this exact same disease. I have far less worry about the tech and far more worry about how the disconnected venture capital caste is inflicting it upon us.
> The TL;DR is that Google engineering appears to have the same AI adoption footprint as John Deere, the tractor company. Most of the industry has the same internal adoption curve: 20% agentic power users, 20% outright refusers, 60% still using Cursor or equivalent chat tool. It turns out Google has this curve too... [0]
Ummmm... Steve. You think Google might be able to figure out a super huge awesome new thing from 1 out of 5 of their employees. Or, given this is a consistent curve across the industry (even at Google)... Maybe AI is only about a fifth as cool and helpful as you and the enthusiasts think it is?
[0] https://simonwillison.net/2026/Apr/13/steve-yegge/#atom-ever...
> Note: “AI experts” refer to individuals whose work or research relates to AI. The AI experts surveyed are those who were authors or presenters at an AI-related conference in 2023 or 2024 and live in the U.S. Expert views are only representative of those who responded.
> The United States reported the lowest trust in its own government to regulate AI responsibly of any country surveyed, at 31%.
It seems US citizens are really against the current administration, just using the fact that AI investment is intrinsecally connected to it to voice their opposition.
> Country-level expectations follow similar patterns to the earlier sentiment trends. Nigeria, Japan, Mexico, the United Arab Emirates, South Korea, and India all expected AI to create more jobs than it eliminates, with shares above 60%. The United States and Canada sat at the opposite end, where 67% and 68% of respondents expected AI to eliminate jobs and disrupt industries.
Globally, the disconnect is not growing. It's really just an U.S. problem (spilling to neighbouring Canada too).
So, no luddites in sight, again. It's just a public perception over a polemic topic being leveraged for ideological reasons sinking AI on US only.
The fundamental alignment issue is aligning the companies themselves with society, not the models with the companies. Widespread unemployment is not aligned with society, but it is aligned with Anthropic and OpenAI if it makes them rich.
Therefore the only “harms” the companies will take seriously are those which also harm the company. For example reputational harms from enabling scams aren’t allowed.
Perhaps all of this isn’t fair, since companies actively subverted safety research for profitability. But then I would go back to my earlier point of over-indexing on unintended behaviors and under-indexing on intended ones.
Take log review for example. Whether it’s admin or security LLMs are incredible at reading awfully formatted logs and even using those to pull meaning from other logs as well. Like turn an hour long log review into a 10 minute log review type thing.