> To prepare for such changes, Gates said it is urgent to discuss policies to mitigate AI's impact on jobs. One idea would be to introduce a tax on AI tokens and robots, a move Gates argued would slow companies' displacement of human workers. "Right now, if you're an employer and you hire someone, you pay payroll taxes on their earnings," he wrote. "But if you buy a robot, you can usually write it off right away as a business expense. The tax system nudges you toward replacing people with machines."
This might be a case of the labor market aligned with government incentives (to increase taxes).
He's been advocating that point since the turn of the century when it was actual robots in mostly industrial jobs.
I still agree with his point, on principal, however. We just currently have a regeme that's diametrically opposed to making large companies and wealthy people pay...<checks notes>...anything unless it was from lottery winnings.
Business will still pay it because robots and AI work around the clock, don’t take sick or vacation days and don’t need health insurance or retirements or workers comp.
I’ve been saying this since generative ai first started showing up on my radar in 2020. This is going to bankrupt entitlement programs if we don’t replace the lost revenue.
My fear is not 100% unemployment. It’s a permanent 10-20%.
It is worth noting that employers write off employee wages the same way they write off the costs of AI.
The "payroll taxes" here is social security and Medicare, which are obviously quite important, but it's sorta obvious why that wouldn't apply to other expenses. The base compensation someone gets has the same tax treatment as any other expense.
That said, benefits, retirement plans, medical insurance, etc all are a part of hiring humans and not machines, so the theme here of cost differentials is still true.
What he is recommending is a bit broader. What he is saying is that revenue from labor (earned income), and capital are taxes very differently, and that it is time to modify the incentives to give the labor market more time to adjust.
It’s surprising that Gates is calling attention and proposing solutions to this growing problem that affects both the American worker and the retiree.
In 2008, Microsoft lobbyists in Washington spearheaded the effort to give international students up to 2 additional years of employment via the new STEM OPT work permit to allow foreigners a total of 3 tries at the H1B visa lottery instead of 1. Those executive actions by 2 different WH administrations have created an uncapped pipeline of white-collar foreign workers into the US since then, surpassing the total number of capped & cap-exempt H1B visas annually.
What needs to be done first, even before taxing corporations & businesses on AI, besides eliminating STEM OPT and making sure CPT, OPT & H1B aren’t abused and are used correctly according to existing laws, is for Congress to place surtaxes on US employers for each worker they hire on CPT, OPT, H1B and offshored positions in other countries. Together with the taxes on AI, those collected surtaxes could go directly to SS and Medicare trust funds and to retrain American workers to survive in the age of AI.
It’s broader than that. By taxing tokens you increase the cost of A.I. relative to the worker potentially replaced by it. The sweet spot would be the employee remains employed and becomes more productive if they use A.I., which is good for everybody.
Technically payroll tax is also a write-off, but yeah there should be an equivalent tax on AI. Normally workers pay half and employers pay the other half. With AI the employer should pay the full amount of course.
I've been saying something similar for a while, tax on automation to fund a UBI or negative income tax or similar.
Personally I don't think the idea that we need to protect jobs for the sake of getting people employed is a worthy cause, rather making sure that everyone can benefit from the fruits of automation and no one goes hungry is what really matters
Ubi is never going to happen. What cpuld.maybe happen is.you harness the anti ai backlash to charge a tax for heavy ai usage on corporations and earmark the money to keep elctricity and water prices down for citizens.
I'm not sure that's right. Hire someone, you pay payroll taxes but you pay less corporate taxes, dividend taxes or capital gains taxes.
Businesses profits including profits from any machine or robot are taxed at ~40% if you include all those taxes. So the token taxes are already in place. As companies make more profits with AI, government revenues increase.
There are some loopholes. eg. Roth 401(k)s and Roth IRAs and such allow lower income retirees to skip paying some of the "robot taxes". I suppose if we want to close those loopholes, new taxes could help (though it would probably just be simpler to reduce those tax breaks).
Sure seems like a good time to take a few days and re-read Grapes of Wrath, if you haven’t recently. How we respond to AI replacement of human labor is our second generational moment (the first being environmentalism and conservation, which is safe to say we thoroughly failed)
Come now: the failure has been decades in the making. We did have that one shining moment when CFCs and the hole in the ozone later we're taken seriously...
Al Gore not being elected was decades ago, almost 3 decades ago at this point in time. So his failure to be elected means the mess is decades in the making.
Also if you notice the last summit to get proper sign up and implementation was Rio in 1992
The US didn't even want to look at the 1995 Kyoto Protocols when he was VP. We gave ourselves a pat on the back for vastly improving the ozone layer and considered everything fixed
it started before hand. Like thinking making people recycle will make much of a difference. Instead of just controlling plastic production into a minimum.
Rather than the news article, you can read Gates' blog post directly - so you skip on the context and selection CBS is feeding you and get his entire thoughts.
>Part of the reason for this is that many commentators underestimate the extent of the impact AI will have. I think there are a few reasons why.
>One is the fact that AI models still make mistakes. It is hard to envision any of them replacing human cognition when, not long ago, they couldn’t solve a simple Sudoku puzzle or figure out how many R’s are in the word strawberry.
>But the reliability problem is being fixed quickly, as researchers create models that can check their own work and improve themselves. Soon they will be substantially better than humans at many tasks.
But I also think that Gates does miss one aspect of why the public's perception of AI is so different from those that work directly with it. There's a twofold bias happening here.
The first is that the public largely does not work with the institutional instances of these models. The overwhelming majority of people's experience is Gemini's search results, which are objectively terrible. Or free Claude/Chat GPT, which are both quite mediocre. Even the "premium" versions of these available to your normal joe public individual user are just okay. So you get a lot of people who's experience with AI is largely the dumbest worst version of it, usually intentionally limited. Furthermore these models never really tell you that they're limited because of the version you have so what is often intentional price tiers gets interpreted as shortcomings.
Secondly, people in tech, Gates included, have a bit of bias in the opposite direction. One thing AI is genuinely very good at right now is writing code. It can more or less replace an entire team of entry level to intermediate coders with one senior guy who's really good at using AI. IMO this creates a bit of bias, in that coders think their job to be historically quite technically difficult (hence the compensation), and now they see a machine replacing what they built their billions doing. Transferrable expertise fallacy is heavily at work here, the presumption that comes from this is that AI will replace jobs all over the place, but the reality is that AI may replace 4/5ths of the tech workforce but never exhibit skills beyond a first year med student in medicine or never meet the skillset of even your most average paralegal.
We'll see, ultimately I do think Gates is more right than wrong here, but even still he's prone to the same biases everyone else in tech has regarding AI's true skillsets.
If its so great at writing code, why is software getting worse and worse? Im assuming companies like Microsoft have access to the best AI models for coding... where are the improvements?
That's a sorta broad question, are the aspects of software that are getting worse in your mind due to errors in code? Or design choices? Or reflective of an ever more complex world that said code needs to operate in?
No, there is plenty of documented concern of the technical debt getting created from AI.
It's not talked about as much because the sexier story is AI nuking all of our jobs but just wait....A new story is emerging. And it's not just the coding "debt", but we are creating an educational gap as well by basically eliminating entry level jobs at scale.
The industry has no choice but to solve this problem soon.
There is no guarantee of how much more advanced models might get-- could be a little, could be a lot. But one thing we can be pretty confident of is that it will be expensive. And execs don't like blowing their budgets up on stuff they don't understand very well. They need subject matter experts.
In addition, my tinfoil hat theory is that government(s) will restrict access to higher performing models, and probably already have done more in that space than we will ever know. So even if this tech could get a lot better, it may not for the general public.
We are going to need to train up more humans. There are like a million reasons why this simply has to happen. Short of a major apocalyptic wipeout, we will need them. If for nothing other than the fact that the legal landscape is slow to change, and if we can't hold bots accountable for their actions then we will need people to blame.
> In addition, my tinfoil hat theory is that government(s) will restrict access to higher performing models, and probably already have done more in that space than we will ever know. So even if this tech could get a lot better, it may not for the general public.
The government doesn't need to do anything. Frontier models are costly and will burn the labs making them. They still have not made a profit (ignoring Anthropic's 'arr' cope) due to training costs, and the business model is "the more customers we have and the more model use, the more expensive it is to serve these models".
Well, for sure, but some of them will potentially choose better ones, and they'll potentially have a chance to compete on those merits, and on a fair playing field we could see some interesting things.
Of course, the fact that it's not a fair playing field can slow progress down. Still, it's an ecosystem. If something gets too out of whack, eventually a rebalancing will occur. Then that will eventually get out of whack... rinse and repeat.
I think, in terms of this issue of juniors never developing senior-level skills and leaving the industry in the lurch eventually - I think it'll be like fish not noticing the waters it lives in being worse than they used to be. 10 years from now, when this bites, no one will really recognize they're getting bitten, because they won't have anything to compare it to. Things will just suck and companies will just accept that as being the same as it always was.
>some of them will potentially choose better ones
Maybe an irrelevant few, but the pressures of the market will push most everyone in that race to the bottom.
Sadly, you have a point. When too many of the wrong environmental variables line up, sometimes things just keep getting worse for a long time. I do expect corrections to come, but they would look very different depending on when and how they happen.
All I know is I think we should definitely keep encouraging junior talent across the board. I continue to believe that generally speaking, the more skills we all have, the better. I suspect many will be surprised at how much the new machines can't do (at least for a while).
> One thing AI is genuinely very good at right now is writing code. It can more or less replace an entire team of entry level to intermediate coders with one senior guy who's really good at using AI.
Let me stop you right there. The only reason this use case is even "working" right now, for the barest notion of what working is, is because the cost is so heavily subsidized. If you are building your entire software organization around the idea that Anthropic is never going to raise prices, you're about to get demolished when the subsidies run out.
It is the worst vendor lock-in I have ever seen in my 20 years of professional software experience. It was bad enough when companies were literally using third-party services to handle user authentication and authorization, but now the thought is that it is possible to outsource your entire workforce consequence-free? LOL
I'm not sure where you read anything regarding costs, that's a pretty separate conversation, this was specifically capabilities. Costs are going to shift over time in a few unpredictable ways, and of course will ultimately be a major factor regarding how companies adopt these capabilities, but costs and capabilities are two very distinct topics within this realm that operate independently. Seems like you're having a different conversation and think they're opposed?
The code creation capabilities are genuinely very good is only true when you finish the sentence.
"The code creation capabilities are genuinely very good for the current price."
If you were made to pay the actual cost of these capabilities, the value judgment would drop considerably. They would not be called very good if every prompt cost you $50.
I understand you're very fixated on having a conversation about costs, I don't disagree that costs are a vital aspect of how and where AI will be integrated over time.
But the above conversation was not about integration or costs, it was about capability. I get a lot of people on reddit sorta have this weird obsession with turning everything in to an argument, but fundamentally speaking framing an unrelated topic as opposed to a statement just so you can be aggressive in your reply isn't really the hallmark of mature or intelligent discussion, is it?
The context of the essay is AI job displacement. You said that AI is genuinely good at writing code. A 20 ton excavator is genuinely good at digging a 4x4x4 hole in your front yard. It will also cost you about $5,000 for the privilege instead of the $400 for some dudes standing at Home Depot.
What is the point about talking about the theoretical downstream impacts of a highly-capable tool if the unsubsidized cost will ultimately look like the above, real-world example? If the unsubsidized cost is so massive, which it may very well end up being, that everyone who got fired because of AI gets hired back somewhere else, who cares if it is actually genuinely good at coding? If the only way the stuff actually works is to set other people's money on fire infinitely, its capabilities are pointless, unless we transcend money entirely.
It seems like you're hell bent on trying to force some sort of debate here, despite me replying several times that your "objection" rests on discussing something outside of the range of my above comment. IDK what else to tell you, you're trying to force some strawman that doesn't exist over and over. It's not gonna magically make it exist, it's just going to make you look like a weirdo who can't accept that the fight they're trying to pick makes no sense lol.
You must have forgotten what you wrote in your original comment, then? You very specifically said that Bill Gates was suffering from a bias because all he sees is tech and how rad these things are at being "genuinely good at writing code." Further, this is in the broader context of the essay wherein the discussion centers around worker displacement due to AI.
If you want to have a conversation about pure capabilities without cost at all, then it gets really simple. LLMs, regardless of who builds or hypes them, are fundamentally not intelligent. They are guessing machines which just take text and return text. They have no memory, they have no consciousness nor comprehension. Their "capabilities" are smashing rocks together at incredible rates to probably match training that has been executed in digital sweatshops in the poorest places on the planet.
IDK what your deal is, you're so obsessed with trying to force an argument, and it's clear that in trying to do that you needed to move outside of the breadth of my comment. IDK why you chose this random comment to obsess over but I'm gonna ignore any further replies - I think at this point it's clear there's something wrong with you.
It’s not just code though. Math, physics, life sciences, clinical work, legal work, materials science, manufacturing, etc. etc.
It’s not just code, models are getting better at every task and it’s happening recursively. There’s all types of machine learning problems being solved besides pure language models as well, including but not limited to solving protein folding, protein binding, drug discovery, materials science, fusion, etc.
People indexing on career safety because Gemini flash is bad have zero idea what’s actually happening right now. None at all. The models are frighteningly good and they’re close to or already at the point where they’re engineering their successors.
Yup. The people choosing to ignore this and believe nothing about it is real are pulling the ostrich strategy and getting angry at anyone else who has their head out of the sand.
Here's the full essay
https://www.gatesnotes.com/a-turbulent-ai-era-and-critical-choices-to-make
The job losses and severe un(der) employment for new graduates was completely predictable. The better question is just what the hell else did anyone expect to happen in places like the US?
The upper management in China made it pretty clear that AI is great, but it is not to be used as a tool to do mass firings and get rid of people. Of course in the US, that's precisely how every single end user company sees AI once you look past all the usual bullshit propaganda.
Billy boy and a number of others are starting to get worried because the elites have already lost control of the process. Couple that with a country whose political and economic climate is already looking a lot like post WW1 Weimar Republic, and anyone who's taken world history in highschool should be able to see how Bad Things™ are going to happen.
The government has been siding with contractors*. The rulings just make it a breach of contract for workers who are employed as contracts have their contracts ripped up. The rulings do not say that AI can’t replace workers.
Compensation is not exactly preventing workers from being fired. I think most US companies offer compensation anyway, so I’m not sure how different this is.
I'm not going to put China on a pedestal here, given their human rights track record. They are a heavy handed government, but in this scenario, their vertical management of every aspect of the economy is well suited. They just go and dictate that no mass firings shall happen to replace people with AI. It sounds noble and all that, and in this instance they're siding with workers. The downside is how they enforce this, if you don't tow the party line, they just "vanish" you.
To be honest, I'm finding it difficult to criticize them, they at least listen to experts on ecological impacts and practical data. They just pick and choose deliberately and issue mandates that you better fucking follow, or else. Even if they do run roughshod on what's considered "good", be it for the environment, people, etc. I just feel like they're taking the right approach here, on this specific thing.
This is not true regarding China. The nuance is that most people employed in China are contract workers. The litigation found that for contracts where substantial changes in the situation the business found itself in was reason to terminate the contract, AI was not a significant enough change that would allow them to do so.
AI is absolutely coming for Chinese workers as well and there are no such legal protections in China yet.
I keep waiting for the people who dismiss this as hype to reckon with the economic reality: when AI starts automating more work, it will not just “assist” workers — it will reduce the number of people needed to do the job. That is the part skeptics keep ignoring. The programmer who says, “I can't finish this now,” may be right in the short term. But in five or 10 years, the competitive pressure will be completely different. Companies will adopt AI where it cuts costs, boosts output, or replaces headcount, and that will reshape the labor market whether people want to admit it or not. My point is simple: pretending this will not affect jobs is shortsighted.
That’s just it: the people most often talking about this live in a very, very insular tech universe and seem to think that “software designer” is a common job. It’s not. A vast majority of the American workforce doesn’t touch computers as part of their work day outside of end user stuff like nursing and teaching.
When AI starts coming for nurses, cops, and teachers, then you can talk to me. It’s not doing that any time soon outside of automating some very specific parts of those jobs. If you can have a whole conversation about job loss that doesn’t include the terms “software” or “data analysis” I might be interested. Otherwise this is just people being really out of touch with the American labor market
I keep waiting for uneducated AI doomers to provide any proof that this automation will be any different from the others. They can't because no proof of this exists
I mean, Gates’ article itself deals with exactly your contention:
>
Another reason people underestimate AI is that analogies to the effects of past innovations are misleading. We have no experience with a technology that can be adopted quickly or that can think and move like a human. When the PC came along, it took twenty years to significantly change how we worked because the software had to be developed, the price had to come down, and people had to learn how to use the tools and incorporate them into their business processes. AI, on the other hand, runs on the devices we already have, and it uses natural language. We don’t have to adapt to it because it can adapt to us. It can watch the same training video that is used to train human workers and learn from existing data.
How can we prove something based on datapoints we don’t yet have? The “doomers” are logically correct that for the moment, AI seems very very different than prior advances in technology.
In prior decades, when automation would come along, new jobs would be created that people who were bumped from their old jobs could fill. This is the first time in human history where the new jobs will likely be able to be filled by the very technology creating them, leaving no room for humans.
> How can we prove something based on datapoints we don’t yet have?
We do have datapoints. It's called every bit of automation in human history
> AI seems very very different than prior advances in technology.
This was said about the internet, heavy machinery and every technology before that. It's innovation. The whole point is to be unique
> new jobs would be created that people who were bumped from their old jobs could fill
Which is exactly what is happening right now and would be happening at an even higher level if Trump didn't hit the "damage economy for no reason" button
>
We do have datapoints. It's called every bit of automation in human history
Not for AI. I have explained to you why those other advances cannot be imputed to AI.
>
This was said about the internet, heavy machinery and every technology before that. It's innovation. The whole point is to be unique
Which doesn’t in any way prove your point. Again, AI is the first technology able to fill the jobs it creates.
>
Which is exactly what is happening right now and would be happening at an even higher level if Trump didn't hit the "damage economy for no reason" button
Read my comment again. Why fill the new jobs with people when we can also fill them with AI. You are wrong and you will continue to be wrong
It's amazing that trillions are being burned on LLMs. Sure some money is going into other Transformer uses, but there's too much faith on Anthropic and OpenAI. Whomst we have also seen aren't spearheaded by serious people.
Edit: the AI bro "giving me a lesson" is so superior he had to insult me, then blocked me immediately. Call out the finances and they get feral, that's how sad these people are.
Anthropic has stated that their value is 30T dollars (hype cope for IPO). An investor visited their HQ and said they are full of "people who believe they are creating a God".
OpenAI has lost a crazy amount of executives in the past few months. Altman once said (not as a joke, unless you think you can tell a joke with the tone of GMan) "we will build the AGI and then ask it how to make a profit". Then to push back on the data center hype he said "we should organize field trips to get data centers, that will change everyone's mind!"
These are not serious people and if you think they are, leave your terminally online bubble.
AI hasn't eliminated jobs overnight. It has shifted work to humans, who now spend hours babysitting AI (rewriting prompts, fact checking for errors, and making sure it doesn't sound robotic). AI is really good at recognizing patterns, reading millions of documents in seconds, and performing repetitive tasks.
These types of technological shifts don’t replace jobs overnight. Outside of the tech world there has been little impact in terms of job reduction and you’re spot on about how people are using it.
The real shift is coming when these AI models are fully integrated into companies main software systems, i.e., SAP, Oracle, Workday, etc. Once we get there you will absolutely see significant reductions throughout companies. You won’t need a demand planner, because the system will just auto adjust demand plans daily, it will auto adjust the production plan to match, it will auto adjust the raw material orders, etc. It will heavily reduce tactical roles and create more centralized strategic roles.
With that being said, this is still 5 years out, and only for companies that are going to be on the more expensive types of software. The time for companies to upgrade to this type of landscape will take varying amounts of time. Some companies in 15 years will pretty much still utilize AI the way they are today IMO because they won’t want to spend for the upgraded software.
I think you can make a lot of parallels here with the industrial revolution or steam engine. There's certainly going to be turmoil, and those who don't/can't adapt will face pain (as a society it's our job to take care of them). But ultimately the idea that this just replaces humans is silly.
A single farming combine replaces something like 150-200 manual laborers. And yet, things like this didn't destroy human's ability to work, it freed us up from spending all day in a field to spending all day doing something far more productive.
AI will ultimately be the same, as we are able to automate away repetitive menial tasks (mental tasks can still be repetitive and menial) society will ultimately progress and create new higher uses for humans.
Now, there will be some workers who are left behind, a worker who's in their 50s and has spent their entire life doing data entry tasks is unlikely to find a career in a few years unless they can adapt to AI quickly.
Factory automation can teach us a lot of tough lessons here too. For instance, we make more cars domestically today than we ever did, but we do so with around half the number of total factory workers and they're paid far worse. Those populations were left behind, and despite America progressing economically at breakneck speed through the 80s, 90s, and 00s the midwest faltered then encountered deep issues with poverty, drug abuse, etc. This need not be the outcome.
Right, and what you’re describing represents a tiny minority of what people actually do for work in America, to say nothing of the rest of the world where the vast, vast majority of people will never touch a computer as part of their work. Do you have any idea how many people farm rice for a living? You want to change the world’s workforce? Automate mother fucking rice. Otherwise you’re just playing around with some jobs in a few dozen big cities
A friend ~~works~~ worked in accounting/finance at the Bill & Melinda Gates Foundation, and they just laid off close to half their finance department (including her) citing AI, so he’s not wrong but also doesn’t seem to care. I will say it was the single most impressive severance package I’ve ever heard of.
Companies are all out here playing chicken with each other. They want everyone else to do what's necessary to maintain a healthy economy, while they only focus on profits.
They also keep it secret when they stop using AI. I know of two major financial firms dropping it this month, one is on the SP500. The reality is that below the executive level that is jacking off over the lies Altman and Co spout, it is a massive shitshow under them. Some people know how to use LLMs to boost their productivity, great! Unfortunately they are a fraction of those using the tools and most either waste money doing useless tasks, or making problems even worse (10xing your dev team means you are writing shit code just as fast with verbose, hard to read code).
This isn’t true. Finance teams are tied to program groups and some program groups are facing reductions mainly as a result of where the board wants to prioritize funding and therefore sunsetting programs, not AI efficiency. No one has been fired for AI, yet. It saves time but is still not accurate and consistent enough
This simply isn’t true. I work in a biglaw firm, we terminated a bunch of admin staff this year including a bunch of people working in accounting because AI made other people so much more efficient.
Interestingly, your lack of ability to synthesize information and follow an argument between multiple comments is leading me to believe that you’re an AI bot
Is this the same guy who asked the deceased pedo Jefferey Epstein to secretly get him VD pills to sneak his wife? Luckily for Billy, that’s a job AI won’t be taking over soon.
I found this somewhere but I think it hits the nail on the head.
At a time of stubbornly high inflation that’s pushed consumer confidence to a seven-month low, Americans are looking up to see tech’s trillion-dollar companies getting bigger and more powerful, all while being led by some of the wealthiest people on the planet.
“If this technology takes off in the way that you see the CEOs of AI Labs talking about, we’re facing a significant impact to our economy, to people’s jobs, to their sense of agency,” data center concerns cut across a number of issues, including water use, higher utility bills, emissions increases, noise pollution and the potential that the facilities are “taking over agricultural land.
Meta endangered children and teens with its social media apps like Instagram and Facebook. As part of the settlement, the company agreed to make substantial changes to its products in addition to paying billions of dollars in penalties.
I stopped listening to Bill Gates once he drugged his own wife. All his perspectives may indeed be valuable, but unfortunately I can’t take this guy seriously anymore.
I’m a UI/UX Designer and am showing my web and app developers this week how I successfully have used AI to export from Figma designs production-ready HTML/CSS/Swift/Kotlin. It’s going to change our entire workflow.
What goons have we got running around where AI is going to beat them at their jobs? I guess it explains so much about our society and culture that its been run by clowns who are fodder for takeover by calculators.
Bill Gates first and foremost is the very definition of a man without a conscience; in any way that you want to defined what it means to have a conscience. For me what having a conscience means... that whatever you choose to do it life, you do it well by following the golden rule. Do unto others, as you would have them do unto you; in others words you treat other humans in the exact same way that you want other humans to treat you. This is how, when all is said and done, and you are lying on your deathbed, you can slip away peacefully.
Bill Gates name being in the Epstein file, is just one of the many shit*y things he has done in his life to break the Golden Rule in order to achieve his wealth and fame, at the cost of others. He basically invented a better typewriter, got it thanks Bill. Now after years of grift, monopolistic practices, bribing others and abusive behavior he is warning us of AI's impact on our lives. The same warning, much greater men than he will ever be already gave us, Isaac Asimov comes to mind.
This is my warning to America. Bill Gates doesn't care about any of us. I read his essay and watched his video. It was little more than heartfelt Ad for AI and how it is going to affect our economy globally. All because of the cadre of Robber Barons at the top of the Global Wall Street Economy who can't control themselves. The warning wouldn't even be necessary if we just said no... to a group of Rich Guys trying to Profit off Replacing the Labor of Man. Instead of trying to figure out other ways to be more productive, inventings things that actually help our species, instead of controlling humans and making them less happy.
That is the bottom-line of what Economics is actually about. Feeding and caring for each other in ways that follows the Golden Rule. This is what Democracy is all about, what an economy run by Capitalism with a Conscience is about. It is an economy with rules and regulations in place, to actually stop people like Bill Gates with his archaic notions of economic control, from becoming Kings while running the world and having an oversized voice.
His wife took a lengthy amount of time after she found out about what happened with Epstein to divorce him. Maybe a lot of that time was spent consulting with lawyers but it showed that divorcing wealth is not easy.
I remember the interview with both Bill and Melinda where she said, in a very snarky tone "But Bill said it didn't happen, so it must not have happened" before all the Epstein stuff went apeshit. In hindsight it made sense, but at the time everyone was thinking some major shit was about to drop but no idea what it could be.
Seems like she made the decision real fast, but you have an excellent point. Bezos divorce was anything but "promptly" and I poorly worded that. It's likely that Melinda promptly made the decision, but in practice it took years.
in its current state I agree with you, but the plan is to use this for general automation (which is something that doesn’t exist and we have never dealt with). This is a single invention meant to automate anything. anything you can think of. It will be paired up with general robotics to increase its breadth of jobs it can automate
we don’t know where this is going, but we have never invented something like this before, this is very different
That may be well be the aspiration, but I have yet to see any convincing evidence that it is achievable with the technology in its current form. And much like Nuclear fusion, you can easily imagine achieving stable fusion at a cost effective price. But achieving it is another matter entirely.
yeah, I could totally see you being right about this
but… a whole bunch of scientists, insiders, government officials, CEOs of companies that make more money than the majority of entire nations, economists, and various other experts think it is the start of general automation. A thingy we know is coming even if you are correct and the smart people are all wrong
I wouldn’t know one way of the other, so I’ll give you enough credit to call this a 50/50 chance
but, I will say that if you are right and the scientists are wrong, and this just turns out to be a program that can write code well, then our economy has huge (albeit different) problems to contend with
I think 50/50 is a strange number to land on. What is your prior assumption?
Mine is that earth shattering technologies are extremely rare and their impact very unpredictable.
Starting from that assumption, my baseline assumption is that this is like a 1/1000 chance of happening. Then when I add my personal use of the tech, my knowledge of it (Stats/Data Science degree) and four years of broken promises about progress... that probably bumps me to 1/10000 chance.
Now, if Mr. Gates and co know something about some new version of the tech that the rest of us have not seen, I would change my mind. However I have a strong piece of evidence that they don't... They are massively in debt, and heading towards an apocalyptically disappointing IPO, and the easiest solution would be to just reveal and demonstrate that tech (which they have not been shy about doing constantly until now to boost value).
So, my suspicion is the whole thing is a damp squib. A somewhat useful technology that will change how we work (much as the panic about computers did - but then took 50 years to really filter into the economy/society).
Edit: And yes, I agree. If the bear case is correct we are in for a hell of a correction in the near future.
Current unemployment rate is 4.1% and we're still waiting to see an acceleration in economic productivity... Nope, the Luddites aren't suddenly going to be right this time.
But there has arguably been an economic recession outside of a handful of stocks and companies (Mag 7) for half a decade. Just as in 2008, younger people are hit hardest in those circumstances.
You do understand that those curves return median earnings?
Here's an older chart, but illustrates the distortion inherent in your dataset. When the tranches are tracked at a granular level, the facts are clear.
Yeah, median being the 50th percentile. And in recent history the the middle class has been shrinking as the upper and upper middle have been growing. You angry about upwards mobility?
Anyways, how many miles do you think the goalpost has been moved from the Luddite AI justification for government intervention?
No it isn't. You don't just get to make up a new unemployment calculation to fit your narrative. Especially with 0 historical context for it. This is the kind of uneducated nonsense trump supporters do
Employed
In the Current Population Survey (CPS), people are classified as employed if, during the survey reference week, they meet any of the following criteria:
worked at least 1 hour as a paid employee (see wage and salary workers)
worked at least 1 hour in their own business, profession, trade, or farm (see self-employed)
were temporarily absent from their job, business, or farm, whether or not they were paid for the time off (see with a job, not at work)
worked without pay for a minimum of 15 hours in a business or farm owned by a member of their family (see unpaid family workers)
This is a red herring. Doesn't matter which you use. The trend is what the argument is about. And the higher unemployment rate here has shown even better improvement over recent decades. Thanks AI?
The made up part is what is derived from the BLS survey. Please make sure you can read at a 4th grade level before embarrassing yourself in the future
There is not a significant portion of the population working 1 hour a week, and no one cares about this organizations definition or calculation of a poverty wage
It's "killing jobs" to pay for it, but it hasn't made a difference in profitablity for customers of the AI labs. Outside of Internet hype, the deployment of this tech hasn't been sunshine and rainbows.
The benefit to software companies must be far from uniform because I'm seeing tremendous value from where I'm sitting. Some features that took us 3 months now take 2 weeks.
So what you're saying is you are pushing shit code because management thought it was a good idea, and now your senior devs are overwhelmed from the flood of tech debt they have to allow into prod. You're being hyperbolic. You're lying and this shit is getting really fucking old.
Yeah, your output is 1500% than normal with LLMs. No, you're just baiting hype.
"back in the day", a lot of folks literally lost their heads when the amount of people with no money and nothing to do reached a critical point while the people in charge merrily ignored the problem.
For what it's worth, the US has a number of structural social problems that are simply unfixable without a complete social, economic, and cultural reset. So the first paragraph happening in the US is not as bad as it sounds.
Yes, I understand that, I'm saying people are rightfully concerned about that, which is what I'm pointing out, as was Mr. Gates in the article I linked to.
Junglebook3 | 17 hours ago
I found this part most novel:
> To prepare for such changes, Gates said it is urgent to discuss policies to mitigate AI's impact on jobs. One idea would be to introduce a tax on AI tokens and robots, a move Gates argued would slow companies' displacement of human workers. "Right now, if you're an employer and you hire someone, you pay payroll taxes on their earnings," he wrote. "But if you buy a robot, you can usually write it off right away as a business expense. The tax system nudges you toward replacing people with machines."
This might be a case of the labor market aligned with government incentives (to increase taxes).
MikeOKurias | 16 hours ago
He's been advocating that point since the turn of the century when it was actual robots in mostly industrial jobs.
I still agree with his point, on principal, however. We just currently have a regeme that's diametrically opposed to making large companies and wealthy people pay...<checks notes>...anything unless it was from lottery winnings.
pseudonominom | 15 hours ago
Just like climate change, school shootings, pedophile politicians and crippling debt:
nothing will be done, the results will be catastrophic, and the GOP will face no consequences.
metalgtr84 | 13 hours ago
Yeah it’s almost comical to suggest what Gates is. Republicans just fundamentally don’t believe the government should help the labor class.
StrongLoan9751 | 13 hours ago
I hate that you're right but you are, 100%.
supamario132 | 12 hours ago
They're facing plenty of consequences. There are 2.2 billion consequences sitting in Trump's bank account from just this term alone
dpzdpz | 11 hours ago
Well good morning Little Miss Sunshine.
PreparationHot980 | an hour ago
Business will still pay it because robots and AI work around the clock, don’t take sick or vacation days and don’t need health insurance or retirements or workers comp.
MrKingsport | 15 hours ago
I’ve been saying this since generative ai first started showing up on my radar in 2020. This is going to bankrupt entitlement programs if we don’t replace the lost revenue.
My fear is not 100% unemployment. It’s a permanent 10-20%.
Initial_Savings3034 | 12 hours ago
Excellent point.
OddlyFactual1512 | 9 hours ago
We'll be lucky if it's that low.
TechnologyEither | 17 hours ago
This is a devastating point
RIP_Soulja_Slim | 16 hours ago
It is worth noting that employers write off employee wages the same way they write off the costs of AI.
The "payroll taxes" here is social security and Medicare, which are obviously quite important, but it's sorta obvious why that wouldn't apply to other expenses. The base compensation someone gets has the same tax treatment as any other expense.
That said, benefits, retirement plans, medical insurance, etc all are a part of hiring humans and not machines, so the theme here of cost differentials is still true.
unsafeideas | 16 hours ago
We not juat tax corporations and billionaires? Will amount to the same thing and actually solve the issues better then taxing tokens.
Archaia | 16 hours ago
What he is recommending is a bit broader. What he is saying is that revenue from labor (earned income), and capital are taxes very differently, and that it is time to modify the incentives to give the labor market more time to adjust.
Advanced-Sky-758 | 9 hours ago
It’s surprising that Gates is calling attention and proposing solutions to this growing problem that affects both the American worker and the retiree.
In 2008, Microsoft lobbyists in Washington spearheaded the effort to give international students up to 2 additional years of employment via the new STEM OPT work permit to allow foreigners a total of 3 tries at the H1B visa lottery instead of 1. Those executive actions by 2 different WH administrations have created an uncapped pipeline of white-collar foreign workers into the US since then, surpassing the total number of capped & cap-exempt H1B visas annually.
What needs to be done first, even before taxing corporations & businesses on AI, besides eliminating STEM OPT and making sure CPT, OPT & H1B aren’t abused and are used correctly according to existing laws, is for Congress to place surtaxes on US employers for each worker they hire on CPT, OPT, H1B and offshored positions in other countries. Together with the taxes on AI, those collected surtaxes could go directly to SS and Medicare trust funds and to retrain American workers to survive in the age of AI.
Additional_Olive3318 | 11 hours ago
It’s broader than that. By taxing tokens you increase the cost of A.I. relative to the worker potentially replaced by it. The sweet spot would be the employee remains employed and becomes more productive if they use A.I., which is good for everybody.
Momoselfie | 14 hours ago
Technically payroll tax is also a write-off, but yeah there should be an equivalent tax on AI. Normally workers pay half and employers pay the other half. With AI the employer should pay the full amount of course.
the_ai_wizard | 13 hours ago
what happens when no other countries have this tax?
website-buyer | 11 hours ago
Exactly that. It’s not enforceable in a way to work the way they expect to work
sviridoot | 8 hours ago
I've been saying something similar for a while, tax on automation to fund a UBI or negative income tax or similar.
Personally I don't think the idea that we need to protect jobs for the sake of getting people employed is a worthy cause, rather making sure that everyone can benefit from the fruits of automation and no one goes hungry is what really matters
Wedgiebro | 32 minutes ago
Ubi is never going to happen. What cpuld.maybe happen is.you harness the anti ai backlash to charge a tax for heavy ai usage on corporations and earmark the money to keep elctricity and water prices down for citizens.
BenE | 12 hours ago
I'm not sure that's right. Hire someone, you pay payroll taxes but you pay less corporate taxes, dividend taxes or capital gains taxes.
Businesses profits including profits from any machine or robot are taxed at ~40% if you include all those taxes. So the token taxes are already in place. As companies make more profits with AI, government revenues increase.
There are some loopholes. eg. Roth 401(k)s and Roth IRAs and such allow lower income retirees to skip paying some of the "robot taxes". I suppose if we want to close those loopholes, new taxes could help (though it would probably just be simpler to reduce those tax breaks).
FearlessPark4588 | 14 hours ago
The cost of goods and services will go down, people will make less, and on net, there will be deflation, if Gates view of the world holds.
hippydipster | 11 hours ago
Which is disastrous given how our overall economic system is set up.
deeziegator | 17 hours ago
Sure seems like a good time to take a few days and re-read Grapes of Wrath, if you haven’t recently. How we respond to AI replacement of human labor is our second generational moment (the first being environmentalism and conservation, which is safe to say we thoroughly failed)
DotComprehensive4902 | 16 hours ago
We failed with environmentalism when Al Gore wasn't elected president
alltehmemes | 16 hours ago
Come now: the failure has been decades in the making. We did have that one shining moment when CFCs and the hole in the ozone later we're taken seriously...
DotComprehensive4902 | 16 hours ago
Al Gore not being elected was decades ago, almost 3 decades ago at this point in time. So his failure to be elected means the mess is decades in the making.
Also if you notice the last summit to get proper sign up and implementation was Rio in 1992
Uncle_Tickle_Monster | 16 hours ago
Pre-internet, Fox News, social media. We’re fucked now.
few | 12 hours ago
And acid rain. We did a good job on that.
But now we're working hard to make up for the few wins back in the 70's, 80' and 90's. 😟
Dan_Berg | 12 hours ago
The US didn't even want to look at the 1995 Kyoto Protocols when he was VP. We gave ourselves a pat on the back for vastly improving the ozone layer and considered everything fixed
Biggie62 | 11 hours ago
it started before hand. Like thinking making people recycle will make much of a difference. Instead of just controlling plastic production into a minimum.
Zank_Frappa | 14 hours ago
Does he still live in that massive 10,000 square foot home? Hard to take the guy's message seriously when he lives like that.
dfsw | 13 hours ago
Seriously?
Zank_Frappa | 13 hours ago
Yes, seriously. Should the ruling class be allowed to live like kings while preaching austerity for the plebes?
hhmmn | 13 hours ago
Buddy - not sure I'm ready for the digital tractor
deeziegator | an hour ago
me neither
CardiologistFit6090 | 12 hours ago
environmentalism and conservation is not a movement, not even an after thought.
RIP_Soulja_Slim | 16 hours ago
Rather than the news article, you can read Gates' blog post directly - so you skip on the context and selection CBS is feeding you and get his entire thoughts.
https://www.gatesnotes.com/a-turbulent-ai-era-and-critical-choices-to-make
RIP_Soulja_Slim | 16 hours ago
This part mirrors a lot of my thoughts around AI:
>Part of the reason for this is that many commentators underestimate the extent of the impact AI will have. I think there are a few reasons why.
>One is the fact that AI models still make mistakes. It is hard to envision any of them replacing human cognition when, not long ago, they couldn’t solve a simple Sudoku puzzle or figure out how many R’s are in the word strawberry.
>But the reliability problem is being fixed quickly, as researchers create models that can check their own work and improve themselves. Soon they will be substantially better than humans at many tasks.
But I also think that Gates does miss one aspect of why the public's perception of AI is so different from those that work directly with it. There's a twofold bias happening here.
The first is that the public largely does not work with the institutional instances of these models. The overwhelming majority of people's experience is Gemini's search results, which are objectively terrible. Or free Claude/Chat GPT, which are both quite mediocre. Even the "premium" versions of these available to your normal joe public individual user are just okay. So you get a lot of people who's experience with AI is largely the dumbest worst version of it, usually intentionally limited. Furthermore these models never really tell you that they're limited because of the version you have so what is often intentional price tiers gets interpreted as shortcomings.
Secondly, people in tech, Gates included, have a bit of bias in the opposite direction. One thing AI is genuinely very good at right now is writing code. It can more or less replace an entire team of entry level to intermediate coders with one senior guy who's really good at using AI. IMO this creates a bit of bias, in that coders think their job to be historically quite technically difficult (hence the compensation), and now they see a machine replacing what they built their billions doing. Transferrable expertise fallacy is heavily at work here, the presumption that comes from this is that AI will replace jobs all over the place, but the reality is that AI may replace 4/5ths of the tech workforce but never exhibit skills beyond a first year med student in medicine or never meet the skillset of even your most average paralegal.
We'll see, ultimately I do think Gates is more right than wrong here, but even still he's prone to the same biases everyone else in tech has regarding AI's true skillsets.
DarthJDP | 13 hours ago
If its so great at writing code, why is software getting worse and worse? Im assuming companies like Microsoft have access to the best AI models for coding... where are the improvements?
RIP_Soulja_Slim | 13 hours ago
That's a sorta broad question, are the aspects of software that are getting worse in your mind due to errors in code? Or design choices? Or reflective of an ever more complex world that said code needs to operate in?
Diazpora | 12 hours ago
No, there is plenty of documented concern of the technical debt getting created from AI.
It's not talked about as much because the sexier story is AI nuking all of our jobs but just wait....A new story is emerging. And it's not just the coding "debt", but we are creating an educational gap as well by basically eliminating entry level jobs at scale.
https://youtu.be/zp9GTOVAyoI?si=rsocadgtLT-vq8b1
slfnflctd | 12 hours ago
> we are creating an educational gap
The industry has no choice but to solve this problem soon.
There is no guarantee of how much more advanced models might get-- could be a little, could be a lot. But one thing we can be pretty confident of is that it will be expensive. And execs don't like blowing their budgets up on stuff they don't understand very well. They need subject matter experts.
In addition, my tinfoil hat theory is that government(s) will restrict access to higher performing models, and probably already have done more in that space than we will ever know. So even if this tech could get a lot better, it may not for the general public.
We are going to need to train up more humans. There are like a million reasons why this simply has to happen. Short of a major apocalyptic wipeout, we will need them. If for nothing other than the fact that the legal landscape is slow to change, and if we can't hold bots accountable for their actions then we will need people to blame.
Olangotang | 11 hours ago
> In addition, my tinfoil hat theory is that government(s) will restrict access to higher performing models, and probably already have done more in that space than we will ever know. So even if this tech could get a lot better, it may not for the general public.
The government doesn't need to do anything. Frontier models are costly and will burn the labs making them. They still have not made a profit (ignoring Anthropic's 'arr' cope) due to training costs, and the business model is "the more customers we have and the more model use, the more expensive it is to serve these models".
hippydipster | 11 hours ago
> The industry has no choice but to solve this problem soon.
Nonsense, they have lots of choices, and you can be sure they'll choose the worst ones.
slfnflctd | 11 hours ago
Well, for sure, but some of them will potentially choose better ones, and they'll potentially have a chance to compete on those merits, and on a fair playing field we could see some interesting things.
Of course, the fact that it's not a fair playing field can slow progress down. Still, it's an ecosystem. If something gets too out of whack, eventually a rebalancing will occur. Then that will eventually get out of whack... rinse and repeat.
hippydipster | 11 hours ago
I think, in terms of this issue of juniors never developing senior-level skills and leaving the industry in the lurch eventually - I think it'll be like fish not noticing the waters it lives in being worse than they used to be. 10 years from now, when this bites, no one will really recognize they're getting bitten, because they won't have anything to compare it to. Things will just suck and companies will just accept that as being the same as it always was.
>some of them will potentially choose better ones
Maybe an irrelevant few, but the pressures of the market will push most everyone in that race to the bottom.
slfnflctd | 10 hours ago
All my homies hate the race to the bottom!
Sadly, you have a point. When too many of the wrong environmental variables line up, sometimes things just keep getting worse for a long time. I do expect corrections to come, but they would look very different depending on when and how they happen.
All I know is I think we should definitely keep encouraging junior talent across the board. I continue to believe that generally speaking, the more skills we all have, the better. I suspect many will be surprised at how much the new machines can't do (at least for a while).
fbiguy22 | 13 hours ago
Writing good code has little to do with the overall end quality of a service. That’s a project management issue.
Hacking_the_Gibson | 12 hours ago
> One thing AI is genuinely very good at right now is writing code. It can more or less replace an entire team of entry level to intermediate coders with one senior guy who's really good at using AI.
Let me stop you right there. The only reason this use case is even "working" right now, for the barest notion of what working is, is because the cost is so heavily subsidized. If you are building your entire software organization around the idea that Anthropic is never going to raise prices, you're about to get demolished when the subsidies run out.
It is the worst vendor lock-in I have ever seen in my 20 years of professional software experience. It was bad enough when companies were literally using third-party services to handle user authentication and authorization, but now the thought is that it is possible to outsource your entire workforce consequence-free? LOL
RIP_Soulja_Slim | 11 hours ago
I'm not sure where you read anything regarding costs, that's a pretty separate conversation, this was specifically capabilities. Costs are going to shift over time in a few unpredictable ways, and of course will ultimately be a major factor regarding how companies adopt these capabilities, but costs and capabilities are two very distinct topics within this realm that operate independently. Seems like you're having a different conversation and think they're opposed?
Hacking_the_Gibson | 7 hours ago
The code creation capabilities are genuinely very good is only true when you finish the sentence.
"The code creation capabilities are genuinely very good for the current price."
If you were made to pay the actual cost of these capabilities, the value judgment would drop considerably. They would not be called very good if every prompt cost you $50.
RIP_Soulja_Slim | 7 hours ago
I understand you're very fixated on having a conversation about costs, I don't disagree that costs are a vital aspect of how and where AI will be integrated over time.
But the above conversation was not about integration or costs, it was about capability. I get a lot of people on reddit sorta have this weird obsession with turning everything in to an argument, but fundamentally speaking framing an unrelated topic as opposed to a statement just so you can be aggressive in your reply isn't really the hallmark of mature or intelligent discussion, is it?
Hacking_the_Gibson | 4 hours ago
The context of the essay is AI job displacement. You said that AI is genuinely good at writing code. A 20 ton excavator is genuinely good at digging a 4x4x4 hole in your front yard. It will also cost you about $5,000 for the privilege instead of the $400 for some dudes standing at Home Depot.
What is the point about talking about the theoretical downstream impacts of a highly-capable tool if the unsubsidized cost will ultimately look like the above, real-world example? If the unsubsidized cost is so massive, which it may very well end up being, that everyone who got fired because of AI gets hired back somewhere else, who cares if it is actually genuinely good at coding? If the only way the stuff actually works is to set other people's money on fire infinitely, its capabilities are pointless, unless we transcend money entirely.
RIP_Soulja_Slim | 4 hours ago
It seems like you're hell bent on trying to force some sort of debate here, despite me replying several times that your "objection" rests on discussing something outside of the range of my above comment. IDK what else to tell you, you're trying to force some strawman that doesn't exist over and over. It's not gonna magically make it exist, it's just going to make you look like a weirdo who can't accept that the fight they're trying to pick makes no sense lol.
Take care.
Hacking_the_Gibson | 3 hours ago
You must have forgotten what you wrote in your original comment, then? You very specifically said that Bill Gates was suffering from a bias because all he sees is tech and how rad these things are at being "genuinely good at writing code." Further, this is in the broader context of the essay wherein the discussion centers around worker displacement due to AI.
If you want to have a conversation about pure capabilities without cost at all, then it gets really simple. LLMs, regardless of who builds or hypes them, are fundamentally not intelligent. They are guessing machines which just take text and return text. They have no memory, they have no consciousness nor comprehension. Their "capabilities" are smashing rocks together at incredible rates to probably match training that has been executed in digital sweatshops in the poorest places on the planet.
RIP_Soulja_Slim | 3 hours ago
I'm speaking directly to what I wrote.
IDK what your deal is, you're so obsessed with trying to force an argument, and it's clear that in trying to do that you needed to move outside of the breadth of my comment. IDK why you chose this random comment to obsess over but I'm gonna ignore any further replies - I think at this point it's clear there's something wrong with you.
Dickasaurus_Rex_ | 3 hours ago
It’s not just code though. Math, physics, life sciences, clinical work, legal work, materials science, manufacturing, etc. etc.
It’s not just code, models are getting better at every task and it’s happening recursively. There’s all types of machine learning problems being solved besides pure language models as well, including but not limited to solving protein folding, protein binding, drug discovery, materials science, fusion, etc.
People indexing on career safety because Gemini flash is bad have zero idea what’s actually happening right now. None at all. The models are frighteningly good and they’re close to or already at the point where they’re engineering their successors.
i8abug | an hour ago
Yes, exactly. And coders are just first to see through the window
Dickasaurus_Rex_ | an hour ago
Yup. The people choosing to ignore this and believe nothing about it is real are pulling the ostrich strategy and getting angry at anyone else who has their head out of the sand.
Tall-Control8992 | 16 hours ago
Here's the full essay https://www.gatesnotes.com/a-turbulent-ai-era-and-critical-choices-to-make
The job losses and severe un(der) employment for new graduates was completely predictable. The better question is just what the hell else did anyone expect to happen in places like the US?
The upper management in China made it pretty clear that AI is great, but it is not to be used as a tool to do mass firings and get rid of people. Of course in the US, that's precisely how every single end user company sees AI once you look past all the usual bullshit propaganda.
Billy boy and a number of others are starting to get worried because the elites have already lost control of the process. Couple that with a country whose political and economic climate is already looking a lot like post WW1 Weimar Republic, and anyone who's taken world history in highschool should be able to see how Bad Things™ are going to happen.
methodinmadness7 | 15 hours ago
Can you elaborate a bit on what China did? I hadn’t heard of that.
Comfortable_Jury369 | 15 hours ago
There are a few lawsuits where workers were laid off and ostensibly replaced with AI, and the government has been siding with the workers.
A_Novelty-Account | 11 hours ago
The government has been siding with contractors*. The rulings just make it a breach of contract for workers who are employed as contracts have their contracts ripped up. The rulings do not say that AI can’t replace workers.
chollida1 | 13 hours ago
> There are a few lawsuits where workers were laid off and ostensibly replaced with AI, and the government has been siding with the workers.
Can you share these lawsuits? I hadn't heard this and its encouraging news.
Comfortable_Jury369 | 13 hours ago
https://www.theguardian.com/world/2026/may/13/china-court-awards-compensation-sacked-worker-replaced-by-ai
Not optimistic about this being extended to the US, or universally assisting people who can't prove their firing was related to AI.
methodinmadness7 | 13 hours ago
Compensation is not exactly preventing workers from being fired. I think most US companies offer compensation anyway, so I’m not sure how different this is.
chollida1 | 13 hours ago
Oh, rats, that's China so not really relevant in North America.
Appreciate you sending this my way regardless!!
Advanced-Patient-161 | 12 hours ago
I'm not going to put China on a pedestal here, given their human rights track record. They are a heavy handed government, but in this scenario, their vertical management of every aspect of the economy is well suited. They just go and dictate that no mass firings shall happen to replace people with AI. It sounds noble and all that, and in this instance they're siding with workers. The downside is how they enforce this, if you don't tow the party line, they just "vanish" you.
To be honest, I'm finding it difficult to criticize them, they at least listen to experts on ecological impacts and practical data. They just pick and choose deliberately and issue mandates that you better fucking follow, or else. Even if they do run roughshod on what's considered "good", be it for the environment, people, etc. I just feel like they're taking the right approach here, on this specific thing.
A_Novelty-Account | 11 hours ago
This is not true regarding China. The nuance is that most people employed in China are contract workers. The litigation found that for contracts where substantial changes in the situation the business found itself in was reason to terminate the contract, AI was not a significant enough change that would allow them to do so.
AI is absolutely coming for Chinese workers as well and there are no such legal protections in China yet.
Wedgiebro | 24 minutes ago
This is still better then any us state has done on the issue
Initial_Savings3034 | 12 hours ago
CCP leadership supports industrial policy that bolsters employment.
They have a justifiable concern about lots of unemployed young people marching on Beijing.
Ok_Possible_2260 | 16 hours ago
I keep waiting for the people who dismiss this as hype to reckon with the economic reality: when AI starts automating more work, it will not just “assist” workers — it will reduce the number of people needed to do the job. That is the part skeptics keep ignoring. The programmer who says, “I can't finish this now,” may be right in the short term. But in five or 10 years, the competitive pressure will be completely different. Companies will adopt AI where it cuts costs, boosts output, or replaces headcount, and that will reshape the labor market whether people want to admit it or not. My point is simple: pretending this will not affect jobs is shortsighted.
1-Dollar-Doge-Coins | 12 hours ago
No one is saying it won't affect jobs. Just that the expected impact is likely overstated.
jollyllama | 27 minutes ago
That’s just it: the people most often talking about this live in a very, very insular tech universe and seem to think that “software designer” is a common job. It’s not. A vast majority of the American workforce doesn’t touch computers as part of their work day outside of end user stuff like nursing and teaching.
When AI starts coming for nurses, cops, and teachers, then you can talk to me. It’s not doing that any time soon outside of automating some very specific parts of those jobs. If you can have a whole conversation about job loss that doesn’t include the terms “software” or “data analysis” I might be interested. Otherwise this is just people being really out of touch with the American labor market
Joltizer | 14 hours ago
I keep waiting for uneducated AI doomers to provide any proof that this automation will be any different from the others. They can't because no proof of this exists
A_Novelty-Account | 11 hours ago
I mean, Gates’ article itself deals with exactly your contention:
> Another reason people underestimate AI is that analogies to the effects of past innovations are misleading. We have no experience with a technology that can be adopted quickly or that can think and move like a human. When the PC came along, it took twenty years to significantly change how we worked because the software had to be developed, the price had to come down, and people had to learn how to use the tools and incorporate them into their business processes. AI, on the other hand, runs on the devices we already have, and it uses natural language. We don’t have to adapt to it because it can adapt to us. It can watch the same training video that is used to train human workers and learn from existing data.
How can we prove something based on datapoints we don’t yet have? The “doomers” are logically correct that for the moment, AI seems very very different than prior advances in technology.
In prior decades, when automation would come along, new jobs would be created that people who were bumped from their old jobs could fill. This is the first time in human history where the new jobs will likely be able to be filled by the very technology creating them, leaving no room for humans.
Joltizer | 11 hours ago
> How can we prove something based on datapoints we don’t yet have?
We do have datapoints. It's called every bit of automation in human history
> AI seems very very different than prior advances in technology.
This was said about the internet, heavy machinery and every technology before that. It's innovation. The whole point is to be unique
> new jobs would be created that people who were bumped from their old jobs could fill
Which is exactly what is happening right now and would be happening at an even higher level if Trump didn't hit the "damage economy for no reason" button
A_Novelty-Account | 10 hours ago
> We do have datapoints. It's called every bit of automation in human history
Not for AI. I have explained to you why those other advances cannot be imputed to AI.
> This was said about the internet, heavy machinery and every technology before that. It's innovation. The whole point is to be unique
Which doesn’t in any way prove your point. Again, AI is the first technology able to fill the jobs it creates.
> Which is exactly what is happening right now and would be happening at an even higher level if Trump didn't hit the "damage economy for no reason" button
Read my comment again. Why fill the new jobs with people when we can also fill them with AI. You are wrong and you will continue to be wrong
Joltizer | 10 hours ago
> Not for AI
I have already lectured you as to why you're wrong here
> Again, AI is the first technology able to fill the jobs it creates
Incorrect. Jobs that are not filled by people aren't even jobs. Please at least attempt to use basic critical thinking skills
> You are wrong and you will continue to be wrong
Please do not project and learn from what I am teaching you
StickFit1459 | 9 hours ago
Just so you know, you’re 100% wrong here, and frankly you don’t sound very bright. Stop preening like you’ve won something.
Joltizer | 9 hours ago
I am 100% correct, unfortunately for you. I really don't care what uneducated people like you think
Wedgiebro | 23 minutes ago
You sound like an idiot btw
Olangotang | 11 hours ago
It's amazing that trillions are being burned on LLMs. Sure some money is going into other Transformer uses, but there's too much faith on Anthropic and OpenAI. Whomst we have also seen aren't spearheaded by serious people.
Edit: the AI bro "giving me a lesson" is so superior he had to insult me, then blocked me immediately. Call out the finances and they get feral, that's how sad these people are.
Joltizer | 11 hours ago
Just because you don't like AI does not mean that Anthropic and OpenAI are not staffed with serious people
Olangotang | 10 hours ago
I like AI :)
Anthropic has stated that their value is 30T dollars (hype cope for IPO). An investor visited their HQ and said they are full of "people who believe they are creating a God".
OpenAI has lost a crazy amount of executives in the past few months. Altman once said (not as a joke, unless you think you can tell a joke with the tone of GMan) "we will build the AGI and then ask it how to make a profit". Then to push back on the data center hype he said "we should organize field trips to get data centers, that will change everyone's mind!"
These are not serious people and if you think they are, leave your terminally online bubble.
Joltizer | 10 hours ago
> . An investor visited their HQ and said they are full of "people who believe they are creating a God".
This is not evidence. This is cope from someone who gets their view of these companies from social media
> "we will build the AGI and then ask it how to make a profit"
"CEO overhypes product. Weather at 7"
> leave your terminally online bubble.
Please don't project so much
Olangotang | 10 hours ago
The only winning move is not to play with AI bros.
YoureEconIlliterate | 10 hours ago
Please don't block your superiors when we are teaching you why you're wrong
You're welcome to ignore AI, and I expect you to do so. It's why you've struggled so much in your CS career and financially
cotton-candy-dreams | 11 hours ago
🤣👏🏼
hippydipster | 11 hours ago
"Proof" is not available to us, in general, and almost every time someone demands it, it is only selectively.
Joltizer | 11 hours ago
TL;DR no proof beyond doom scrolling
InclinationCompass | 6 hours ago
Automobiles replaced many jobs and industries, but it also created many.
TGAILA | 16 hours ago
AI hasn't eliminated jobs overnight. It has shifted work to humans, who now spend hours babysitting AI (rewriting prompts, fact checking for errors, and making sure it doesn't sound robotic). AI is really good at recognizing patterns, reading millions of documents in seconds, and performing repetitive tasks.
ofesfipf889534 | 15 hours ago
These types of technological shifts don’t replace jobs overnight. Outside of the tech world there has been little impact in terms of job reduction and you’re spot on about how people are using it.
The real shift is coming when these AI models are fully integrated into companies main software systems, i.e., SAP, Oracle, Workday, etc. Once we get there you will absolutely see significant reductions throughout companies. You won’t need a demand planner, because the system will just auto adjust demand plans daily, it will auto adjust the production plan to match, it will auto adjust the raw material orders, etc. It will heavily reduce tactical roles and create more centralized strategic roles.
With that being said, this is still 5 years out, and only for companies that are going to be on the more expensive types of software. The time for companies to upgrade to this type of landscape will take varying amounts of time. Some companies in 15 years will pretty much still utilize AI the way they are today IMO because they won’t want to spend for the upgraded software.
RIP_Soulja_Slim | 15 hours ago
I think you can make a lot of parallels here with the industrial revolution or steam engine. There's certainly going to be turmoil, and those who don't/can't adapt will face pain (as a society it's our job to take care of them). But ultimately the idea that this just replaces humans is silly.
A single farming combine replaces something like 150-200 manual laborers. And yet, things like this didn't destroy human's ability to work, it freed us up from spending all day in a field to spending all day doing something far more productive.
AI will ultimately be the same, as we are able to automate away repetitive menial tasks (mental tasks can still be repetitive and menial) society will ultimately progress and create new higher uses for humans.
Now, there will be some workers who are left behind, a worker who's in their 50s and has spent their entire life doing data entry tasks is unlikely to find a career in a few years unless they can adapt to AI quickly.
Factory automation can teach us a lot of tough lessons here too. For instance, we make more cars domestically today than we ever did, but we do so with around half the number of total factory workers and they're paid far worse. Those populations were left behind, and despite America progressing economically at breakneck speed through the 80s, 90s, and 00s the midwest faltered then encountered deep issues with poverty, drug abuse, etc. This need not be the outcome.
hippydipster | 11 hours ago
What are humans going to learn to do that AI won't?
A_Novelty-Account | 11 hours ago
Why would you expect overnight replacement? It will happen gradually but with increasing pace.
jollyllama | 22 minutes ago
Right, and what you’re describing represents a tiny minority of what people actually do for work in America, to say nothing of the rest of the world where the vast, vast majority of people will never touch a computer as part of their work. Do you have any idea how many people farm rice for a living? You want to change the world’s workforce? Automate mother fucking rice. Otherwise you’re just playing around with some jobs in a few dozen big cities
_afox_ | 16 hours ago
A friend ~~works~~ worked in accounting/finance at the Bill & Melinda Gates Foundation, and they just laid off close to half their finance department (including her) citing AI, so he’s not wrong but also doesn’t seem to care. I will say it was the single most impressive severance package I’ve ever heard of.
BrodinGodofSwole | 13 hours ago
Companies are all out here playing chicken with each other. They want everyone else to do what's necessary to maintain a healthy economy, while they only focus on profits.
Olangotang | 11 hours ago
They also keep it secret when they stop using AI. I know of two major financial firms dropping it this month, one is on the SP500. The reality is that below the executive level that is jacking off over the lies Altman and Co spout, it is a massive shitshow under them. Some people know how to use LLMs to boost their productivity, great! Unfortunately they are a fraction of those using the tools and most either waste money doing useless tasks, or making problems even worse (10xing your dev team means you are writing shit code just as fast with verbose, hard to read code).
SkoHens | 13 hours ago
This isn’t true. Finance teams are tied to program groups and some program groups are facing reductions mainly as a result of where the board wants to prioritize funding and therefore sunsetting programs, not AI efficiency. No one has been fired for AI, yet. It saves time but is still not accurate and consistent enough
A_Novelty-Account | 11 hours ago
> No one has been fired for AI, yet.
This simply isn’t true. I work in a biglaw firm, we terminated a bunch of admin staff this year including a bunch of people working in accounting because AI made other people so much more efficient.
SkoHens | 11 hours ago
At the foundation. Idc about your firm
A_Novelty-Account | 5 hours ago
Your statement was general to finance teams, not just the foundation which the other commenter ostensibly has more information than you about.
SkoHens | 5 hours ago
Nope, I work there. I was explaining how finance teams are structured at the foundation and that AI has not replaced finance staff
jollyllama | 14 minutes ago
Interestingly, your lack of ability to synthesize information and follow an argument between multiple comments is leading me to believe that you’re an AI bot
hippydipster | 11 hours ago
> No one has been fired for AI, yet.
That's not a claim you could possibly know.
SkoHens | 11 hours ago
Bro I work there
hippydipster | 11 hours ago
Ah, your comment read like a general statement about the world.
NoseSide | 13 hours ago
Is this the same guy who asked the deceased pedo Jefferey Epstein to secretly get him VD pills to sneak his wife? Luckily for Billy, that’s a job AI won’t be taking over soon.
One-Emu-1103 | 12 hours ago
I found this somewhere but I think it hits the nail on the head.
At a time of stubbornly high inflation that’s pushed consumer confidence to a seven-month low, Americans are looking up to see tech’s trillion-dollar companies getting bigger and more powerful, all while being led by some of the wealthiest people on the planet.
“If this technology takes off in the way that you see the CEOs of AI Labs talking about, we’re facing a significant impact to our economy, to people’s jobs, to their sense of agency,” data center concerns cut across a number of issues, including water use, higher utility bills, emissions increases, noise pollution and the potential that the facilities are “taking over agricultural land.
Meta endangered children and teens with its social media apps like Instagram and Facebook. As part of the settlement, the company agreed to make substantial changes to its products in addition to paying billions of dollars in penalties.
miagi_do | 14 hours ago
I stopped listening to Bill Gates once he drugged his own wife. All his perspectives may indeed be valuable, but unfortunately I can’t take this guy seriously anymore.
ness1210 | 13 hours ago
Bill Gates should be questioned for his role in the Epstein files. He should be in prison not preaching to the rest of us.
More text to meet this dumb requirement.
Gh0st_Pirate_LeChuck | 7 hours ago
I’m a UI/UX Designer and am showing my web and app developers this week how I successfully have used AI to export from Figma designs production-ready HTML/CSS/Swift/Kotlin. It’s going to change our entire workflow.
postconsumerwat | 3 hours ago
What goons have we got running around where AI is going to beat them at their jobs? I guess it explains so much about our society and culture that its been run by clowns who are fodder for takeover by calculators.
americanspirit64 | 16 hours ago
Bill Gates first and foremost is the very definition of a man without a conscience; in any way that you want to defined what it means to have a conscience. For me what having a conscience means... that whatever you choose to do it life, you do it well by following the golden rule. Do unto others, as you would have them do unto you; in others words you treat other humans in the exact same way that you want other humans to treat you. This is how, when all is said and done, and you are lying on your deathbed, you can slip away peacefully.
Bill Gates name being in the Epstein file, is just one of the many shit*y things he has done in his life to break the Golden Rule in order to achieve his wealth and fame, at the cost of others. He basically invented a better typewriter, got it thanks Bill. Now after years of grift, monopolistic practices, bribing others and abusive behavior he is warning us of AI's impact on our lives. The same warning, much greater men than he will ever be already gave us, Isaac Asimov comes to mind.
This is my warning to America. Bill Gates doesn't care about any of us. I read his essay and watched his video. It was little more than heartfelt Ad for AI and how it is going to affect our economy globally. All because of the cadre of Robber Barons at the top of the Global Wall Street Economy who can't control themselves. The warning wouldn't even be necessary if we just said no... to a group of Rich Guys trying to Profit off Replacing the Labor of Man. Instead of trying to figure out other ways to be more productive, inventings things that actually help our species, instead of controlling humans and making them less happy.
That is the bottom-line of what Economics is actually about. Feeding and caring for each other in ways that follows the Golden Rule. This is what Democracy is all about, what an economy run by Capitalism with a Conscience is about. It is an economy with rules and regulations in place, to actually stop people like Bill Gates with his archaic notions of economic control, from becoming Kings while running the world and having an oversized voice.
ReefLedger | 13 hours ago
I feel dumber after wasting my time reading this.
Advanced-Patient-161 | 11 hours ago
Allegedly, he tried to sneak antibiotics into his wife because he caught syphilis from fucking underaged girls.
His wife seems to have found out about the Epstein rumors early on, and then promptly divorced him.
Maybe he's just being honest now?
Ketaskooter | 9 hours ago
His wife took a lengthy amount of time after she found out about what happened with Epstein to divorce him. Maybe a lot of that time was spent consulting with lawyers but it showed that divorcing wealth is not easy.
Advanced-Patient-161 | 8 hours ago
I remember the interview with both Bill and Melinda where she said, in a very snarky tone "But Bill said it didn't happen, so it must not have happened" before all the Epstein stuff went apeshit. In hindsight it made sense, but at the time everyone was thinking some major shit was about to drop but no idea what it could be.
Seems like she made the decision real fast, but you have an excellent point. Bezos divorce was anything but "promptly" and I poorly worded that. It's likely that Melinda promptly made the decision, but in practice it took years.
Joltizer | 14 hours ago
We are really still doing the lump of labor fallacy in 2026. In the econ sub no less
We have automated jobs throughout history and we are richer and better off for it. This is no different
Otherwise_Cupcake_65 | 13 hours ago
in its current state I agree with you, but the plan is to use this for general automation (which is something that doesn’t exist and we have never dealt with). This is a single invention meant to automate anything. anything you can think of. It will be paired up with general robotics to increase its breadth of jobs it can automate
we don’t know where this is going, but we have never invented something like this before, this is very different
dillanthumous | 10 hours ago
That may be well be the aspiration, but I have yet to see any convincing evidence that it is achievable with the technology in its current form. And much like Nuclear fusion, you can easily imagine achieving stable fusion at a cost effective price. But achieving it is another matter entirely.
Otherwise_Cupcake_65 | 5 hours ago
yeah, I could totally see you being right about this
but… a whole bunch of scientists, insiders, government officials, CEOs of companies that make more money than the majority of entire nations, economists, and various other experts think it is the start of general automation. A thingy we know is coming even if you are correct and the smart people are all wrong
I wouldn’t know one way of the other, so I’ll give you enough credit to call this a 50/50 chance
but, I will say that if you are right and the scientists are wrong, and this just turns out to be a program that can write code well, then our economy has huge (albeit different) problems to contend with
dillanthumous | 4 hours ago
I think 50/50 is a strange number to land on. What is your prior assumption?
Mine is that earth shattering technologies are extremely rare and their impact very unpredictable.
Starting from that assumption, my baseline assumption is that this is like a 1/1000 chance of happening. Then when I add my personal use of the tech, my knowledge of it (Stats/Data Science degree) and four years of broken promises about progress... that probably bumps me to 1/10000 chance.
Now, if Mr. Gates and co know something about some new version of the tech that the rest of us have not seen, I would change my mind. However I have a strong piece of evidence that they don't... They are massively in debt, and heading towards an apocalyptically disappointing IPO, and the easiest solution would be to just reveal and demonstrate that tech (which they have not been shy about doing constantly until now to boost value).
So, my suspicion is the whole thing is a damp squib. A somewhat useful technology that will change how we work (much as the panic about computers did - but then took 50 years to really filter into the economy/society).
Edit: And yes, I agree. If the bear case is correct we are in for a hell of a correction in the near future.
Joltizer | 13 hours ago
No, there is no plan for general automation. Some fringe extremists claim it's the goal despite it not being possible
This is not any different from any other automation
hardsoft | 17 hours ago
Guy has gone off the deep the and trying to stay relevant. AI isn't going to drive systemic unemployment higher anymore than Microsoft Excel did.
Initial_Savings3034 | 17 hours ago
It is likely to drive hiring choices away from people and towards utilization of "agents".
hardsoft | 17 hours ago
Current unemployment rate is 4.1% and we're still waiting to see an acceleration in economic productivity... Nope, the Luddites aren't suddenly going to be right this time.
Initial_Savings3034 | 17 hours ago
https://www.lisep.org/tru
The percentage of people, particularly the young that are trapped in zero hour jobs with poverty wages is closer to 25%.
In the Bureau of Labor statistics metric, one hour of paid employment in a week is enough to say a worker is "employed".
https://www.bls.gov/cps/definitions.htm#employed
dillanthumous | 10 hours ago
But there has arguably been an economic recession outside of a handful of stocks and companies (Mag 7) for half a decade. Just as in 2008, younger people are hit hardest in those circumstances.
hardsoft | 17 hours ago
https://fred.stlouisfed.org/series/MEPAINUSA672N
But yeah I get it you're a conspiracy theorist and no data counts except for what you cherry pick.
We should never have allowed the printing press because scribes lost jobs.
And LISEP is down significantly since 2010 so the trend there doesn't reveal any alarmism either...
Initial_Savings3034 | 13 hours ago
You do understand that those curves return median earnings?
Here's an older chart, but illustrates the distortion inherent in your dataset. When the tranches are tracked at a granular level, the facts are clear.
https://www.census.gov/library/visualizations/2015/demo/real-household-income-at-selected-percentiles--1967-to-2014.html
Any specific objection to the LISEP methodology?
hardsoft | 13 hours ago
Yeah, median being the 50th percentile. And in recent history the the middle class has been shrinking as the upper and upper middle have been growing. You angry about upwards mobility?
Anyways, how many miles do you think the goalpost has been moved from the Luddite AI justification for government intervention?
Initial_Savings3034 | 12 hours ago
When an entire population's earnings is presented as one curve, the extremes distort the graphic representation.
In this case, it makes it appear that wages for the median decile (and those below) have grown.
In fact, wages in the highest 3 deciles are the only ones trending upward.
The original article quoting Gates addresses precisely this.
hardsoft | 12 hours ago
The extremes don't distort the median, by definition.
You're probably thinking of mean, or average.
Initial_Savings3034 | 12 hours ago
https://www.census.gov/content/dam/Census/library/visualizations/2015/demo/real-hh-inc.pdf
Joltizer | 14 hours ago
No it isn't. You don't just get to make up a new unemployment calculation to fit your narrative. Especially with 0 historical context for it. This is the kind of uneducated nonsense trump supporters do
Initial_Savings3034 | 13 hours ago
https://www.bls.gov/cps/definitions.htm#employed
Employed In the Current Population Survey (CPS), people are classified as employed if, during the survey reference week, they meet any of the following criteria:
worked at least 1 hour as a paid employee (see wage and salary workers) worked at least 1 hour in their own business, profession, trade, or farm (see self-employed) were temporarily absent from their job, business, or farm, whether or not they were paid for the time off (see with a job, not at work) worked without pay for a minimum of 15 hours in a business or farm owned by a member of their family (see unpaid family workers)
Not a new metric.
Joltizer | 13 hours ago
Yes, good. That is the correct definition. Use that instead of the made up leftist trash you linked earlier
Initial_Savings3034 | 13 hours ago
"Made up leftist trash" which was derived from the BLS survey, stating the obvious.
One hour a week isn't employment sufficient to pay rent.
hardsoft | 13 hours ago
This is a red herring. Doesn't matter which you use. The trend is what the argument is about. And the higher unemployment rate here has shown even better improvement over recent decades. Thanks AI?
1-Dollar-Doge-Coins | 12 hours ago
> One hour a week isn't employment sufficient to pay rent.
What % of workers are only working 1 hour per week? Just curious how big of a deal this is.
Initial_Savings3034 | 12 hours ago
Unknown.
The point is that the BLS considers this the threshold of employment.
Joltizer | 13 hours ago
The made up part is what is derived from the BLS survey. Please make sure you can read at a 4th grade level before embarrassing yourself in the future
There is not a significant portion of the population working 1 hour a week, and no one cares about this organizations definition or calculation of a poverty wage
Initial_Savings3034 | 13 hours ago
https://cdn.prod.website-files.com/63ba0d84fe573c7513595d6e/63c1bb4dc740e1acb5d3b6dd_TRU%20White%20Paper.pdf
Perhaps you didn't read the source material closely.
Junglebook3 | 17 hours ago
There's 1/3 of the job openings for junior software engineers today than they were in the years prior to Gen AI. It's already killing jobs.
Olangotang | 11 hours ago
It's "killing jobs" to pay for it, but it hasn't made a difference in profitablity for customers of the AI labs. Outside of Internet hype, the deployment of this tech hasn't been sunshine and rainbows.
Junglebook3 | 10 hours ago
The benefit to software companies must be far from uniform because I'm seeing tremendous value from where I'm sitting. Some features that took us 3 months now take 2 weeks.
Olangotang | 10 hours ago
So what you're saying is you are pushing shit code because management thought it was a good idea, and now your senior devs are overwhelmed from the flood of tech debt they have to allow into prod. You're being hyperbolic. You're lying and this shit is getting really fucking old.
Yeah, your output is 1500% than normal with LLMs. No, you're just baiting hype.
Junglebook3 | 10 hours ago
OK bud.
hardsoft | 17 hours ago
Real scribes lost their jobs due to the invention of the printing press.
No long term structural unemployment issues justifying government banning or restricting the technology.
Tall-Control8992 | 16 hours ago
"back in the day", a lot of folks literally lost their heads when the amount of people with no money and nothing to do reached a critical point while the people in charge merrily ignored the problem.
For what it's worth, the US has a number of structural social problems that are simply unfixable without a complete social, economic, and cultural reset. So the first paragraph happening in the US is not as bad as it sounds.
hardsoft | 16 hours ago
Same fear mongering we heard at the beginning of the digital revolution.
And the arguments about the rate of change aren't reflected in economic numbers. They just aren't.
Initial_Savings3034 | 13 hours ago
Applied for an entry level job, recently?
It's not just data centric jobs - those people displace less qualified applicants for manual positions.
https://arstechnica.com/ai/2026/08/ai-is-hitting-entry-level-jobs-hardest-stanford-study-finds/
hardsoft | 13 hours ago
Ok I'm convinced we need the government to outlaw AI to save the data center caller jobs /s
Junglebook3 | 12 hours ago
Entry level software jobs are 1/3 what they were 4 years ago. Those jobs are the ticket to the American dream for millions. Gone.
hardsoft | 12 hours ago
Then there would be downwards pressure on wages
https://fred.stlouisfed.org/series/LES1252881600Q
Junglebook3 | 12 hours ago
Yes, I understand that, I'm saying people are rightfully concerned about that, which is what I'm pointing out, as was Mr. Gates in the article I linked to.
Joltizer | 14 hours ago
The only correct comment with any economic literacy sits deep in the negatives