Thank you, Brian, for your thoughtful post about this topic, and for the further research and writing you will be doing about it. Pls also consider including research on, and covering, the blitz of industry corporations and individuals who are acting as PR behind AI -- those who are trying to convince the masses that AI is good for humanity, and "just get on board" or you'll miss out out on innovation, the future as it's supposed to be, etc. It's a powerful force that is making a lot of people just give up and give in.
Such a good point! It’s really malicious & coercive in my opinion. They greatly exaggerated the techs capabilities, while simultaneously hanging the threat of obsolescence over people’s heads so that they start using it out of fear, I know that’s 90% of the reason I make an effort to use AI.
I definitely think it played a part. I’m a software engineer and I worked for a software company but they laid every engineer off and now have this vibe coded AI website idk what they’re even trying to do because it won’t work without engineers.
We're going to have a lot of work cleaning up vibe-coded garbage when it all starts blowing up. Especially if companies like OpenAI continue to hemorrhage money with no clear path to profitability after the economy goes into decline and credit/investment capital dries up. Once OpenAI goes bankrupt and the founders and VC's all run for the hills with their bags full of cash I expect things will go well for us again.
It didn’t take my job as an English teacher, but it affected it and set it up for future replacement. I earned my Masters in English Education and worked very hard for 18 years, and suddenly was told I had to become an expert at AI and teach students to use it for some research and writing tasks. It felt like all my experience was wasted. I retired last June.
I'm commenting even though none of the above apply to me--I'm a homesteader, no threat of AI horning in on that since there's no money in it. My comment is that the threat to jobs is not my top concern about AI; There are two others--surveillance, of workers, yes, but also of all of us all the time everywhere. And, the fact that AI is not intelligent and can't really do creative work--it can only plagiarize, blending what it steals from multiple sources so it can't be legally charged. It also can't do less creative work as well as the humans it replaces, which will have myriad harmful effects on the people who rely on what humans are now doing. It can't exercise judgement. So a hundred things will become more difficult, more frustrating, less efficient and effective, less pleasurable for virtually all of us. But the people making the decisions SIMPLY DON"T CARE. Billionaires don't depend on any of it--if airports are more dangerous, well, they're flying in private jets; if Social Security becomes riddled with errors and it gets to be a nightmare to try to sign up for it, it will have no effect on them, since their SS earnings are a tiny pittance compared to their stashed wealth. If their data centers and cryptocurrency raise the use of fossil fuels and thus speed catastrophic climate change, well yes that WILL affect them but they think they're going to migrate to sunny Mars or download into an android or some goddamn thing They may not be sure of this but the IMPORTANT thing is that they're making lots of money now. Why would people who have more money than they could spend in 20 lifetimes think making even more money is worth risking everything? because they're mentally ill, that's why. And likely part of the attraction of the Mars and similar fantasies is that they'd be enormously expensive, thus giving them a reason to work so hard at piling up ever more wealth They already own a yacht, a dozen huge homes, designer everything, a trophy wife, and several governments. What's next? They need something to get excited about.
I lost my job because management THOUGHT they could replace me and the product I supported with AI, but then shortly afterwards hired a contractor who was assisting me. The CIO was completely tech-illiterate and wouldn't have known the difference between LLM, GM or NLP if his life depended on it. He thought a set of dozens of carefully programmed web forms and process automation flows could be replaced with an LLM chatbot.
I've also taken jobs from people with NLP (Natural Language Processing) which has been around for well over a decade and is now called (barf) AI. This is the kind of algorithm for taking soft squishy human communications methods like speech and written text and turning it into bits.
Data entry. I automated away data entry, and having done data entry for a living I know firsthand it's the kind of awful soul destroying work no human would want to do. The good kind of tech.
The other much less sexy way I've 'taken jobs away' is with process automation software like UIPath; and also APIs, which are somehow sexy despite being a fancy way of saying 'web form a computer fills out instead of a human.'
I kid you not there are people who's entire job all day is copy-pasting text from one program to another. Again, I don't feel bad about automating a horrible mindless job like laboriously copying text from one place to another or going through an email inbox and forwarding emails according to a rote script all day.
There are good applications for algorithms out there, but I have yet to see one for LLMs that an LLM can do reliably. I thought maybe an LLM could do a synopsis or extract metadata from scanned documents, but the last time I looked at it the hallucinations made it unusable for that purpose.
At one company I got an entire roomful data entry clerks an early retirement buyout by replacing slow and error prone paper bookkeeping with webforms, process automation and API calls.
Mostly what I've done is let people like counselors do more their actual job of helping people instead of spending all their time on data entry.
But as I said earlier, what I do isn't sexy or interesting. It's dull and useful.
You "don't feel bad" about automating that job...but the person laid off, with kids and a family to sustain? Maybe he liked it. I used to do a repetitive job years ago. I liked it. When it was automated and I was laid off, I wanted to literally kill myself. Not everyone lives in the US where finding a job is easy. Not everyone can reskill every 2 years.
My perspective is a little different from most people's. I've never had a job last longer than five years. Every job I've ever had has been subject to corporate corruption scandals, buyouts, mergers and more. Employers aren't reliable, and if a job is too easy it's a sign you need to freshen up your resume and skills.
People who've had a job last longer than ten years are like aliens from a different planet.
A layoff for me is just another Tuesday. As a GenX'r I'm a bit ahead of the curve on wage stagnation and employment instability compared to my peers.
Automation and technological progress happen, and at an ever increasing pace. Farms used to employ 50% of all people, and now they employ less than one percent. Many different sectors in manufacturing used to employ ten times as many bodies for the same outputs.
The abomination is that our work weeks aren't shorter, and most of the profits from those production gains have gone to the owners due to lack of regulation and corruption of our economies.
I'm opposed to the so-called AI because it's built on theft, it doesn't fucking work and its burning money at three times the rate it 'earns' money. And that's all even before we talk about the stupid use of water and electricity on the damn things.
You shouldn't be getting angry about losing the job. It's at the lack of severance, a social safety net, the employer keeping some staff at the same hours and pay and regressive tax policies that redistribute all our wealth to the rich.
The vast majority of my work doesn't replace people. It takes away the most awful and stupid aspects of their jobs so they can concentrate on doing the parts only humans can do. I've also saved forests worth of trees by replacing paper with bits, so there's that too.
That's a cop-out sir, sorry to tell you. You can afford to change work every 5 years because you are an American and you get paid so much that you just have to do some basic budgeting to not go broke. Outside of the US, we can't afford to live like you guys. Houses cost a lot everywhere in the world, but WAGES are high only in the US. The average US McDonald employee gets paid 10 times more than a professional in the rest of the world. The US created this international economic system, you made us all dependent on your demand, and now you rugpull us and leave us jobless, what do you think angry, poor people will do when US companies monopolize AI, like they monopolized tech? Then you guys wonder why China and Russia and all the other axis of evil countries manage to find support in the rest of the world...
If you want to be the world leader, you need to lead. Not just exploit us and treat us as expendables. No one trusts the US anymore. You always abandon your allies. And now you will also abandon your workers. Washington is rotten. The US is rotten.
Thanks for this, Brian. We’re seeing the job losses of artists and coders as the Canary in the Coal Mine, which, yeah, also got replaced by a machine at some point (lucking little canary). My summation on the AI push: It’s bosses doing it bc they are scared for their own jobs – they see it as the easy panacea for increasing productivity and reducing labor costs. Same as it ever was. Just faster and with more recklessness.
Thank you for the work you're doing Brian. I tried but it is kind of burning me, it is robbing my peace man. So I'd rather support people like you, Paul Kingsnorth, Peco and others here. Don't give up man, and watch your back! I really hope that there will be an awakening, especially for the jaggernaut that is still in making, that is general AI. No more devastation ☠️
I love your writing, Brian, and your book has been a critical corrective to dominant narratives about the tech lash and what being a luddite actually means. But I'm concerned that when you frame the question as AI killing our jobs, or investigating AI replacement of labor you elide the fact that implementing or adopting AI in the workforce is a political decision made by people. Talking about AI replacing working removes human agency and obfuscates the political economy and power dynamics that are at the root of putting AI into the workplace.
We all know this but it isn't AI that kills the job, its the owners that do (unless those bills pass).
Looking forward to the stories and thanks again for inspiring me to get involved in this discourse on tech and politics.
In my world of product design I see the ground shifting but its far from settled. Talking to users, doing usability research, and finding opportunities are not skills at risk but UI generating, rapid prototyping, copywriting, and some data analysis are. If I only generated UI and prototypes from someone else's research I would not be needed - an owner could get 80% of my value from an emerging tool. Now its not this simple in practice but I am 100% already learning and paying for new AI tools. I'll learn how I can utilize them and acquire new skills to make up for the ones that are becoming commoditized. Its the incessant wave of innovation thats barreling on me and for now I have the energy to ride it. My name isn't at the top of the AI hit list but damn if its not scribbling my name in the margins for later.
Hi Brian, loved the piece. Per your question, I’ve had quite an eclectic anecdotal experience with not the loss of my job, but a spike in demand for something previously not in my field: personal programming and deepfake training. For context, I founded a boutique editorial company about 10 years ago. Among the many editorial throughlines I worked within, I did executive ghostwriting for a string of C-Suite professionals. I wrote an article recently about how one exec created a deepfake with his face and aspects of my linguistic voice (without my knowledge or consent), and the mutual risk exposures that come from not understanding the ownership differences for different architectural systems: Gen-AI, Agentic, LLM, and API.
As an AI tool creator I always aspired to create things that open new doors that didn’t exist before so people who didn’t have the option to achieve X now can. But is this too naive? Do companies inevitably use the same capabilities to replace existing methods because it’s more profitable? What can toolmakers and innovators do to improve things?
Thank you, Brian, for your thoughtful post about this topic, and for the further research and writing you will be doing about it. Pls also consider including research on, and covering, the blitz of industry corporations and individuals who are acting as PR behind AI -- those who are trying to convince the masses that AI is good for humanity, and "just get on board" or you'll miss out out on innovation, the future as it's supposed to be, etc. It's a powerful force that is making a lot of people just give up and give in.
Such a good point! It’s really malicious & coercive in my opinion. They greatly exaggerated the techs capabilities, while simultaneously hanging the threat of obsolescence over people’s heads so that they start using it out of fear, I know that’s 90% of the reason I make an effort to use AI.
Valid point. All in a given week, several people in my life parroted the same line of “AI is here to stay, so we might as well jump on board”.
I definitely think it played a part. I’m a software engineer and I worked for a software company but they laid every engineer off and now have this vibe coded AI website idk what they’re even trying to do because it won’t work without engineers.
We're going to have a lot of work cleaning up vibe-coded garbage when it all starts blowing up. Especially if companies like OpenAI continue to hemorrhage money with no clear path to profitability after the economy goes into decline and credit/investment capital dries up. Once OpenAI goes bankrupt and the founders and VC's all run for the hills with their bags full of cash I expect things will go well for us again.
META out sourcer lays off IT 2,000 workers in Barcelona:
https://www.barrons.com/news/meta-content-moderator-cuts-over-2-000-jobs-in-spain-union-bf842312
It didn’t take my job as an English teacher, but it affected it and set it up for future replacement. I earned my Masters in English Education and worked very hard for 18 years, and suddenly was told I had to become an expert at AI and teach students to use it for some research and writing tasks. It felt like all my experience was wasted. I retired last June.
I'm commenting even though none of the above apply to me--I'm a homesteader, no threat of AI horning in on that since there's no money in it. My comment is that the threat to jobs is not my top concern about AI; There are two others--surveillance, of workers, yes, but also of all of us all the time everywhere. And, the fact that AI is not intelligent and can't really do creative work--it can only plagiarize, blending what it steals from multiple sources so it can't be legally charged. It also can't do less creative work as well as the humans it replaces, which will have myriad harmful effects on the people who rely on what humans are now doing. It can't exercise judgement. So a hundred things will become more difficult, more frustrating, less efficient and effective, less pleasurable for virtually all of us. But the people making the decisions SIMPLY DON"T CARE. Billionaires don't depend on any of it--if airports are more dangerous, well, they're flying in private jets; if Social Security becomes riddled with errors and it gets to be a nightmare to try to sign up for it, it will have no effect on them, since their SS earnings are a tiny pittance compared to their stashed wealth. If their data centers and cryptocurrency raise the use of fossil fuels and thus speed catastrophic climate change, well yes that WILL affect them but they think they're going to migrate to sunny Mars or download into an android or some goddamn thing They may not be sure of this but the IMPORTANT thing is that they're making lots of money now. Why would people who have more money than they could spend in 20 lifetimes think making even more money is worth risking everything? because they're mentally ill, that's why. And likely part of the attraction of the Mars and similar fantasies is that they'd be enormously expensive, thus giving them a reason to work so hard at piling up ever more wealth They already own a yacht, a dozen huge homes, designer everything, a trophy wife, and several governments. What's next? They need something to get excited about.
I lost my job because management THOUGHT they could replace me and the product I supported with AI, but then shortly afterwards hired a contractor who was assisting me. The CIO was completely tech-illiterate and wouldn't have known the difference between LLM, GM or NLP if his life depended on it. He thought a set of dozens of carefully programmed web forms and process automation flows could be replaced with an LLM chatbot.
I've also taken jobs from people with NLP (Natural Language Processing) which has been around for well over a decade and is now called (barf) AI. This is the kind of algorithm for taking soft squishy human communications methods like speech and written text and turning it into bits.
Data entry. I automated away data entry, and having done data entry for a living I know firsthand it's the kind of awful soul destroying work no human would want to do. The good kind of tech.
The other much less sexy way I've 'taken jobs away' is with process automation software like UIPath; and also APIs, which are somehow sexy despite being a fancy way of saying 'web form a computer fills out instead of a human.'
I kid you not there are people who's entire job all day is copy-pasting text from one program to another. Again, I don't feel bad about automating a horrible mindless job like laboriously copying text from one place to another or going through an email inbox and forwarding emails according to a rote script all day.
There are good applications for algorithms out there, but I have yet to see one for LLMs that an LLM can do reliably. I thought maybe an LLM could do a synopsis or extract metadata from scanned documents, but the last time I looked at it the hallucinations made it unusable for that purpose.
At one company I got an entire roomful data entry clerks an early retirement buyout by replacing slow and error prone paper bookkeeping with webforms, process automation and API calls.
Mostly what I've done is let people like counselors do more their actual job of helping people instead of spending all their time on data entry.
But as I said earlier, what I do isn't sexy or interesting. It's dull and useful.
You "don't feel bad" about automating that job...but the person laid off, with kids and a family to sustain? Maybe he liked it. I used to do a repetitive job years ago. I liked it. When it was automated and I was laid off, I wanted to literally kill myself. Not everyone lives in the US where finding a job is easy. Not everyone can reskill every 2 years.
My perspective is a little different from most people's. I've never had a job last longer than five years. Every job I've ever had has been subject to corporate corruption scandals, buyouts, mergers and more. Employers aren't reliable, and if a job is too easy it's a sign you need to freshen up your resume and skills.
People who've had a job last longer than ten years are like aliens from a different planet.
A layoff for me is just another Tuesday. As a GenX'r I'm a bit ahead of the curve on wage stagnation and employment instability compared to my peers.
Automation and technological progress happen, and at an ever increasing pace. Farms used to employ 50% of all people, and now they employ less than one percent. Many different sectors in manufacturing used to employ ten times as many bodies for the same outputs.
The abomination is that our work weeks aren't shorter, and most of the profits from those production gains have gone to the owners due to lack of regulation and corruption of our economies.
I'm opposed to the so-called AI because it's built on theft, it doesn't fucking work and its burning money at three times the rate it 'earns' money. And that's all even before we talk about the stupid use of water and electricity on the damn things.
You shouldn't be getting angry about losing the job. It's at the lack of severance, a social safety net, the employer keeping some staff at the same hours and pay and regressive tax policies that redistribute all our wealth to the rich.
The vast majority of my work doesn't replace people. It takes away the most awful and stupid aspects of their jobs so they can concentrate on doing the parts only humans can do. I've also saved forests worth of trees by replacing paper with bits, so there's that too.
That's a cop-out sir, sorry to tell you. You can afford to change work every 5 years because you are an American and you get paid so much that you just have to do some basic budgeting to not go broke. Outside of the US, we can't afford to live like you guys. Houses cost a lot everywhere in the world, but WAGES are high only in the US. The average US McDonald employee gets paid 10 times more than a professional in the rest of the world. The US created this international economic system, you made us all dependent on your demand, and now you rugpull us and leave us jobless, what do you think angry, poor people will do when US companies monopolize AI, like they monopolized tech? Then you guys wonder why China and Russia and all the other axis of evil countries manage to find support in the rest of the world...
If you want to be the world leader, you need to lead. Not just exploit us and treat us as expendables. No one trusts the US anymore. You always abandon your allies. And now you will also abandon your workers. Washington is rotten. The US is rotten.
Thanks for this, Brian. We’re seeing the job losses of artists and coders as the Canary in the Coal Mine, which, yeah, also got replaced by a machine at some point (lucking little canary). My summation on the AI push: It’s bosses doing it bc they are scared for their own jobs – they see it as the easy panacea for increasing productivity and reducing labor costs. Same as it ever was. Just faster and with more recklessness.
Thank you for the work you're doing Brian. I tried but it is kind of burning me, it is robbing my peace man. So I'd rather support people like you, Paul Kingsnorth, Peco and others here. Don't give up man, and watch your back! I really hope that there will be an awakening, especially for the jaggernaut that is still in making, that is general AI. No more devastation ☠️
I love your writing, Brian, and your book has been a critical corrective to dominant narratives about the tech lash and what being a luddite actually means. But I'm concerned that when you frame the question as AI killing our jobs, or investigating AI replacement of labor you elide the fact that implementing or adopting AI in the workforce is a political decision made by people. Talking about AI replacing working removes human agency and obfuscates the political economy and power dynamics that are at the root of putting AI into the workplace.
We all know this but it isn't AI that kills the job, its the owners that do (unless those bills pass).
Looking forward to the stories and thanks again for inspiring me to get involved in this discourse on tech and politics.
In my world of product design I see the ground shifting but its far from settled. Talking to users, doing usability research, and finding opportunities are not skills at risk but UI generating, rapid prototyping, copywriting, and some data analysis are. If I only generated UI and prototypes from someone else's research I would not be needed - an owner could get 80% of my value from an emerging tool. Now its not this simple in practice but I am 100% already learning and paying for new AI tools. I'll learn how I can utilize them and acquire new skills to make up for the ones that are becoming commoditized. Its the incessant wave of innovation thats barreling on me and for now I have the energy to ride it. My name isn't at the top of the AI hit list but damn if its not scribbling my name in the margins for later.
Hi Brian, loved the piece. Per your question, I’ve had quite an eclectic anecdotal experience with not the loss of my job, but a spike in demand for something previously not in my field: personal programming and deepfake training. For context, I founded a boutique editorial company about 10 years ago. Among the many editorial throughlines I worked within, I did executive ghostwriting for a string of C-Suite professionals. I wrote an article recently about how one exec created a deepfake with his face and aspects of my linguistic voice (without my knowledge or consent), and the mutual risk exposures that come from not understanding the ownership differences for different architectural systems: Gen-AI, Agentic, LLM, and API.
Here’s the piece: https://open.substack.com/pub/anovareport/p/is-chat-gpt-generative-ai-or-llm?utm_source=app-post-stats-page&r=50mu3p&utm_medium=ios
As an AI tool creator I always aspired to create things that open new doors that didn’t exist before so people who didn’t have the option to achieve X now can. But is this too naive? Do companies inevitably use the same capabilities to replace existing methods because it’s more profitable? What can toolmakers and innovators do to improve things?
Yes, you are as naive as those scientists that were working on a new final weapon to "end all wars".