Has the Decline of Knowledge Worker Jobs Begun? (boston.com) 101
The New York Times notes that white-collar workers
have faced higher unemployment than other groups in the U.S. over the past few years — along with slower wager growth.
Some economists wonder if this trend might be irreversible... and partly attributable to AI: After sitting below 4% for more than two years, the overall unemployment rate has topped that threshold since May... "We're seeing a meaningful transition in the way work is done in the white-collar world," said Carl Tannenbaum, the chief economist of Northern Trust. "I tell people a wave is coming...." Thousands of video game workers lost jobs last year and the year before... Unemployment in finance and related industries, while still low, increased by about a quarter from 2022 to 2024, as rising interest rates slowed demand for mortgages and companies sought to become leaner....
Overall, the latest data from the Federal Reserve Bank of New York show that the unemployment rate for college grads has risen 30% since bottoming out in September 2022 (to 2.6% from 2%), versus about 18% for all workers (to 4% from 3.4%). An analysis by Julia Pollak, chief economist of ZipRecruiter, shows that unemployment has been most elevated among those with bachelor's degrees or some college but no degree, while unemployment has been steady or falling at the very top and bottom of the education ladder — for those with advanced degrees or without a high school diploma. Hiring rates have slowed more for jobs requiring a college degree than for other jobs, according to ADP Research, which studies the labor market....
And artificial intelligence could reduce that need further by increasing the automation of white-collar jobs. A recent academic paper found that software developers who used an AI coding assistant improved a key measure of productivity by more than 25% and that the productivity gains appeared to be largest among the least experienced developers. The result suggested that adopting AI could reduce the wage premium enjoyed by more experienced coders, since it would erode their productivity advantages over novices... [A]t least in the near term, many tech executives and their investors appear to see AI as a way to trim their staffing. A software engineer at a large tech company who declined to be named for fear of harming his job prospects said that his team was about half the size it was last year and that he and his co-workers were expected to do roughly the same amount of work by relying on an AI assistant. Overall, the unemployment rate in tech and related industries jumped by more than half from 2022 to 2024, to 4.4% from 2.9%.
"Some economists say these trends may be short term in nature and little cause for concern on their own," the article points out (with one economist noting the unemployment rate is still low compared to historical averages).
Harvard labor economist Lawrence Katz even suggested the slower wage growth could reflect the discount that these workers accepted in return for being able to work from home.
Thanks to Slashdot reader databasecowgirl for sharing the article.
Some economists wonder if this trend might be irreversible... and partly attributable to AI: After sitting below 4% for more than two years, the overall unemployment rate has topped that threshold since May... "We're seeing a meaningful transition in the way work is done in the white-collar world," said Carl Tannenbaum, the chief economist of Northern Trust. "I tell people a wave is coming...." Thousands of video game workers lost jobs last year and the year before... Unemployment in finance and related industries, while still low, increased by about a quarter from 2022 to 2024, as rising interest rates slowed demand for mortgages and companies sought to become leaner....
Overall, the latest data from the Federal Reserve Bank of New York show that the unemployment rate for college grads has risen 30% since bottoming out in September 2022 (to 2.6% from 2%), versus about 18% for all workers (to 4% from 3.4%). An analysis by Julia Pollak, chief economist of ZipRecruiter, shows that unemployment has been most elevated among those with bachelor's degrees or some college but no degree, while unemployment has been steady or falling at the very top and bottom of the education ladder — for those with advanced degrees or without a high school diploma. Hiring rates have slowed more for jobs requiring a college degree than for other jobs, according to ADP Research, which studies the labor market....
And artificial intelligence could reduce that need further by increasing the automation of white-collar jobs. A recent academic paper found that software developers who used an AI coding assistant improved a key measure of productivity by more than 25% and that the productivity gains appeared to be largest among the least experienced developers. The result suggested that adopting AI could reduce the wage premium enjoyed by more experienced coders, since it would erode their productivity advantages over novices... [A]t least in the near term, many tech executives and their investors appear to see AI as a way to trim their staffing. A software engineer at a large tech company who declined to be named for fear of harming his job prospects said that his team was about half the size it was last year and that he and his co-workers were expected to do roughly the same amount of work by relying on an AI assistant. Overall, the unemployment rate in tech and related industries jumped by more than half from 2022 to 2024, to 4.4% from 2.9%.
"Some economists say these trends may be short term in nature and little cause for concern on their own," the article points out (with one economist noting the unemployment rate is still low compared to historical averages).
Harvard labor economist Lawrence Katz even suggested the slower wage growth could reflect the discount that these workers accepted in return for being able to work from home.
Thanks to Slashdot reader databasecowgirl for sharing the article.
Video games bellweather (Score:5, Insightful)
"said Carl Tannenbaum, the chief economist of Northern Trust. "I tell people a wave is coming...." Thousands of video game workers lost jobs last year and the year before."
Yeah, video games have nothing to do with leading a wave in white color jobs. The video game industry is not representative of white collar jobs generally, they're just too different.
Re: (Score:2)
that's not it, at all (Score:5, Insightful)
Two issues - the lockdowns massively increased demand for videogames and videogame companies massively hired in response. That demand fell off pretty quick once the world opened up and people were allowed to touch grass again. (Tech in general with some exceptions like Apple overhired during the lockdowns and then laid a crapton of people off, but videogame companies have been slower in doing layoffs.)
The other is that many recent huge budget videogames have flopped in way that would impress Shamu. Two reasons for that:
Neither of these have anything to do with the current state of the economy. Hopefully the unqualified people get pushed out of the tech sector forthwith so the rest of us can get back to business.
Re: (Score:3)
According to Ubisoft at least AC: Shadows has been the second most revenue in history for the AC franchise. [pcgamer.com] and is sitting at a "Very Positive" with almost 13k Steam reviews.
Re: that's not it, at all (Score:2)
I can say with the utmost confidence that with the respect of your claims about the on the ground reality of the game industry, you do not know what the fuck you're taking about. But if it makes you feel smart, you do you I guess.
Re: (Score:2)
Re: (Score:2)
"said Carl Tannenbaum, the chief economist of Northern Trust. "I tell people a wave is coming...." Thousands of video game workers lost jobs last year and the year before."
Yeah, video games have nothing to do with leading a wave in white color jobs. The video game industry is not representative of white collar jobs generally, they're just too different.
The current churn shows some jobs becoming less common and others becoming more common. This is an economic phenomenon that has always been ever-present. Jobs requiring use of AI are increasing. Jobs developing AI are also increasing and are super high paying jobs. Meanwhile, some jobs are being replaced by AI tools; these jobs tend to be lower paying, but there are also a lot of these jobs. We'll see how all of this manifests in the overall number of jobs. If jobs are replaced by AI with no new jobs
Re: (Score:2)
There are almost 500k tech workers in the Bay Area alone, and very few of them are game developers. So-called "knowledge work" is a vague and broad term. And video game development is not the biggest piece. I'd bet that it's not even a significantly large enough piece to bother bringing up in an essay. It just doesn't make much statistical sense to view a narrow subset of a population and use that to predict what happens to the rest of a very different industry.
It's kind of the same logic when journalists w
Re: (Score:3)
So you're saying they are NPCs?
Re: Video games bellweather (Score:2)
Which philosopher can't spell bellwether?
Re: (Score:2)
Historically, we haven't seen the video games industry's ups and downs ripple into other industries. It seems to have always been just following its own rhythm. So what tells you that it's the bellwether of anything?
Re: (Score:2)
Probably Wittgenstein.
Certainly Confucius, he couldn't even speak English.
Re: Video games bellweather (Score:2)
A few things that I have noticed include the fact that venture capital is long gone and moved probably to AI. Most games lose money and the winners are either live service platforms or continue to spend a lot of money on direct marketing. You will not believe the amount of games that lost money forever and no one cared to have a plan for monetization.
Second is that code bases are stabl
Re: (Score:2)
Almost everything you said comes back to interest rates being not zero and that's discouraging risk-taking.
Re: (Score:2)
I worked in gaming 25 years ago and it was a shit show back then. There was plenty of room for innovation, but nobody wanted to take the risk. It would take a lot of developers to build console games, but they didn't want to lose money while the console games ramped up. So they focused on a few popular franchises and sent everyone else they didn't need packing. We were working on online/networked games, but the management said it was 7 years to early to invest in that. Consumer internet connections were to
Re: (Score:2)
Video games are discretionary spending. Seeing those workers lose their jobs implies that many others, most likely white collar, have lost their jobs and can not afford to spend on games. Sure, the relationship is not direct, but most things are rarely so simple.
I thought it was increasing... (Score:2)
"...along with slower wager growth"
I thought online gambling was booming...
(sorry - couldn't resist)
Of course not (Score:1, Troll)
After all the ludites are just a myth. And whatever you do don't bother reading the Wikipedia article regarding their actual economi
Re: (Score:3)
We cannot stop the rise of technology, and it is folly to try.
We also cannot employ people by giving them work that no longer adds value. That is also folly to try.
Since new tech renders obsolete old skills, economic upheaval is a totally predictable result. It's simply gonna happen. Efforts at stopping it by clinging to the past are doomed to fail.
We DO need to address these problems. I am just saying that we should not address them by trying to restrict innovation or preserve displaced jobs, because t
Re: (Score:3)
The answer is expansion of the welfare state. This scenario of technology displacing workers the model of "sell labor for money, money is what provides your basic needs" is broken so something new has to take it's place. If technology is displacing enough workers to create a real problem then there is massive wealth from that productivity being generated, the question is where.
Re: Of course not (Score:1)
Re: (Score:2)
I like that, assuming the jobs are there for that new training.
Getting down to it I believe for America something like a "Federal Work Program" would be a better political fit for the country, at least as a start over a full UBI (at least I think it's sellable sooner). Something like the Works Progress Administration. You get your benefits ala UBI and other services but if you are capable of working you have to work and if there private jobs are not available or preferable then you get put to work public s
Re: (Score:3)
If you tried that under the current administration all you'd be doing is teaching the young how to be little Nazis so they could become full-blown Maggots when they graduated. Maybe the next administration but I think the impulse to use it for indoctrination is too powerful for the pols to ignore.
Re: (Score:3)
Oh I certainly don't expect the current admin to toy with the idea of anything close to an actual working Federal program of either nature, it's just not even a question to them if such a thing would be worth doing. Like you are right about what the potential negatives are but that's not this crowd in my estimation, such a thing requires a belief first in institutions and that they can in fact work and the Trump admin is on a path to show they cannot (by destroying them and then claiming they don't work).
UBI is a trick (Score:4, Interesting)
Just giving people money is a waste. Not because they will spend it but because we've got about four companies and about 2,000 individuals that own basically everything in the country and if you just give people money those individuals will just jack up prices to absorb it.
The point of my original comment is we need a fundamental change to our civilization or we are going to become a techno-feudal dystopia. And old people don't want that fundamental change because it contradicts what they were told when they were kids.
That comment about 4 to 14 wasn't to throw away comment. There's a concept in religion where if you can get someone to believe it between those ages they are going to be a lifelong adherent. That's because human beings develop the ability to learn information before they can critically evaluate that information. So you can plant ideas and people's heads and it's basically impossible for them to get them out as long as you do it in that age bracket.
Old people, including myself, have a wide variety of terrible ideas they pick up over the years. I don't exactly know what my terrible ideas are because if I did I wouldn't be clinging to them. I have an unhealthy amount of self-reflection brought on by neuroticism so I think I have slightly fewer terrible ideas than the average American but we've all got them. It requires an enormous amount of effort and care to break them down.
Re: (Score:2)
Re: (Score:2)
First, and most important, what you have said is insightful and thoughtful.
Second:
I don't exactly know what my terrible ideas are because if I did I wouldn't be clinging to them.
People have told you... and yet you still cling to those terrible ideas desperately. Do you still call yourself a Democrat and choose to participate in the two-party system? Then you have learned nothing.
Don't worry, I am just as bad, just differently.
Re: (Score:1)
Expanding the welfare state is useless (Score:2)
For that to work there has to be a tightrope to walk over the safety net. The problem is we've been increasing productivity way way way faster than jobs can keep up. New tag is just devouring jobs. We just don't need all these people to be working but our system is built entirely around if you don't work you don't eat.
So there's no tightrope to walk and we're sure shit not goin
Re: (Score:2)
Agree on Federal Jobs Guarantee as a social program. I like the WPA, I think it's a solid model to bring back and keep as a permeant fixture. Combine that with an actual universal healthcare system.
Re: (Score:2)
I always find it a bit sad that the vast majority of people are under the delusion that higher productivity means everyone is worse off.
If our productivity is growing by leaps and bounds, then someone will have to consume all the extra stuff being produced. There's only a small number rich people. They can only drive in so many cars, live in so many houses and eat so many filet mignons every day. The vast majority of new productivity has to be consumed by someone else.
So where does it go?
Since we produce mo
Re: (Score:2)
Disagree, even if the phenomenon you describe is true it still isn't a solve, just an explanation. If the gains from productivity are going to smaller and smaller segment of the population we have a method nobody likes to talk about: tax and redistribute.
Unless you change the mission statement of the Fed entirely then you're not going to get them to encourage deflation, deflation has many many other negative side effects (unless you want a period of degrowth and recession as a solve)
So just tax the people
Re: (Score:2)
gains from productivity are going to smaller and smaller segment of the population
My point is that it can't. For example, if we get 2x more productive at making cholesterol medication, and 50% the population reap the benefits of the extra productivity, they still can't consume 4 times their prescribed amount of medication.
tax and redistribute
I personally have no issues with this, and frankly we've been doing more and more of it nowadays. My point is that it's doesn't need to be the entire solution, or even the main focus. If we simply allow depreciation to happen, very little changes are needed in the tax c
So read my comment again (Score:2)
My comment exists for a very specific purpose. The problems we have are political and social not technological and economic.
Given the nature of this website the only solutions anyone here can see are technological. And when there is no technological solution they can't see any solution so they just pretend there isn't a problem.
That was the point of my comment. It's to point out the people are unwilling to recognize the changes coming
Re: (Score:2)
At what point did I say we can stop technology or even should?
You do it all the time. Every other post of yours I see is you taking it upon yourself to declare that the world isn't ready for AI as if you're some kind of all knowing god figure or at least some kind of uber authority on what technology the world is or isn't ready for. You're none of that, but you think you are anyways.
My comment exists for a very specific purpose. The problems we have are political and social not technological and economic.
You don't seem to have enough real world experience at anything at all to be the judge of that. Nor do you seem to possess book smarts nor street smarts.
Given the nature of this website the only solutions anyone here can see are technological. And when there is no technological solution they can't see any solution so they just pretend there isn't a problem.
I haven't seen that here, not now,
Not AI... (Score:4, Interesting)
A bigger, but more subtle factor is, well, where are the wealthiest people deriving their wealth from? During the last few decades, knowledge workers were prominant since the rising wealthy produced products and services that both employed and sold too that general slice of the population. But now that things have settled, they don't really need a middle class as much. Low paid workers and high income customers, so the value of the middle has been reduced.
Will they drop back to 20th-century levels? (Score:4, Insightful)
All through my career, especially in the 90s and early 2000s, I saw amazing numbers of white-collar jobs just created. My office seemed to need new facilitators and re-organization specialists, and levels of supervisors, and especially people doing "communications". We aquired a whole communications department that we had to work through instead of just informing the public ourselves, handling incoming calls ourselves.
I was never clear on the need for all of them, they didn't seem that productive, day-by-day, and often seemed to be doing jobs that came to nothing later on - reports on shelves.
This may be just a correction.
Re: (Score:2)
It's still just capitalism, it's the (true) meme of "What is something worth? What someone will pay for it" that applies to labor as well, somebody found value in those jobs enough to buy that labor.
Re: (Score:3)
Re: (Score:1)
There is a balanced point really and partly it depends on the volume of work and the size of the entity.
The problem is everyone wants to have their cake and eat it too. They want to hire specialists with specialist level knowledge in all cross domains that are useful for the job as well and that doesn't work. If you want to have a healthy industry not only can that not be your minimum but you have to be ready to hire and train people with experience in related domains for the specialty role.
Re: (Score:2)
Yes, my observations and thoughts exactly.
And, I always saw another dynamic in that. It used to be that successful businesses were built up and run by founders and experts who grew with the company and moved up through the ranks because of their contributions and success.
Then, in the 60's, 70's, 80's, we started a trend of too many business schools, too many MBA's, too many such drones and their media-savvy pundits trying to turn common sense management into a pseudo-science, then too many lawyers, too muc
Re: (Score:2)
As you said, "This may be just a correction." I hope it is. Society is due for a reset.
I don't think so. I think what we're mostly seeing in terms of "societal ills" is largely the result of social media creating a hyper-connected public consciousness. It's a new thing, it's here to stay, and like all new things, we just have to adapt.
Re: (Score:2)
We called that the dot-com bubble.
What I think we're seeing now is basically an aftershock of the pandemic. It knocked out thousands of B&M businesses that were either barely making it or had already been experiencing trouble. In their stead, we saw the rise of a bunch of online services that, at the end of the day, whose continued existence would be questionable after people can go outside again. Some of those remained in-demand, but no doubt not to the level they were at before.
This isn't especially n
Computers (Score:1)
Also the internet created entirely new lines of work and business that didn't exist before.
That said a huge number of jobs in finance don't exist anymore but you probably don't think about. Increases in white collar productivity very quickly resulting layoffs in a way t
Again, More Unsubstantiated AI Hype (Score:2)
Re: (Score:2)
I would not use AI for generating code for games. I would use it for generating graphics, content and for doing testing and game balancing.
Re: (Score:2)
"So AI increases productivity among unproductive software developers by 25%?"
No, it increase a "key measure of productivity" 25%. You fell for it.
"I supposed there are always stupid managers who will try to leverage this."
They are the ones targeted by the hype.
AI helps not at all with the important parts of programming. AI potentially addresses bad management, even as it intends to exploit bad management.
"Knowledge Worker" (Score:2)
What is that? In the broadest sense of the term, everyone is a knowledge worker.
Knowledge regurgitation will become a domain of "AI", if it isn't already. When knowledge generation becomes a domain of AI, all bets are off. Until then, the bigger danger is that various AI models, running out of actual knowledge to absorb, will mere dilute and poison the pool by referencing each other.
Re: (Score:2)
Re: (Score:1)
Spoken like someone who doesn't do garbage pickup I'm sure. I've worked crappy jobs and six figure jobs and often jobs which look easy from the outside are quite challenging from the inside.
Ever try to manage the dish room of a restaurant solo through a rush time? There is quite a bit of knowledge and skill involved even if the equipment and general concept could be understood by a monkey with a lobotomy. The same is true for the grill. Retail, cashier... sure the bar is low to do these jobs but doing them
Re: (Score:2)
So I think you are correct. The current crop of AIs, LLMs, are starved for new inputs, and it seems what you are sa
DEI? (Score:1)
Maybe it's just a reflection of RIF in DEI?
Re: (Score:3)
It's interesting that the example you chose is board members where the only qualifications are connectedness, wealth and sociopathy. Competence, as measured by those attributes, is in fact the problem trying to be addressed. You might as well say that the only consideration for Grand Wizard should be malignant racism, may be true but not helpful.
Re: (Score:1)
This argument already happened and if Harvard Law school couldn't win a case for DEI before the Supreme Court of the United States then AC vs AC showdown is meaningless. Because either you are trying to dishonestly split hairs on qualifiers in your questions to be technically right despite being wrong in principle and spirit or you are simply wrong.
Race and Gender [with some narrow exceptions] are aesthetic non-functional properties and having a diverse presence of them is of no functional benefit for emplo
Re: (Score:2, Interesting)
"Race and Gender [with some narrow exceptions] are aesthetic non-functional properties and having a diverse presence of them is of no functional benefit for employment"
Citation please. Non-functional properties can have a great functional benefit for employment.
"If that weren't true then we should have never stopped the historical discrimination against minorities in he first place."
LOL wut?
"We" didn't stop the "historical discrimination against minorities", but "we" did try to enact remedies. Those remed
Re: (Score:1)
""We" didn't stop the "historical discrimination against minorities""
Of course we did. That all ended with the 14th amendment and the civil rights act.
"but that doesn't mean outcomes, as a group, were"
It turns out humans aren't groups, just a bunch of individuals. Go figure.
There is a story I once heard that may or may not be true but it doesn't matter because it COULD be true and therefore the lesson is valid. A bunch of bean counters decided to modify and improve aircraft cockpits to better fit all the pi
Re: (Score:1)
Some people have trouble understanding that fair doesn't matter. What matters is the advancement of the species and society over time.
We made it illegal to discriminate on the basis of race/gender because science said those weren't indicators of fitness and therefore would advance less fit people over more fit ones. But fitness definitely isn't going to correlate with the racial traits of the people who LOST in the historical competition for lands/spoils so not only does DEI have the same problem as discrim
Re: (Score:2)
Re: (Score:1)
"Any claims that discrimination against minorities doesn't exist in the USA or in Europe are either naive or willfully ignorant."
These kind of blanket absolutes preclude reasonable discussion. "Doesn't exist" is not the same thing as being prevelant in any significant measure. DEI on the other hand is overt and intentional systemic discrimination against the majority by perceived 'race' and half the population on the basis of gender. Bottom line, the 14th amendment prohibits it.
"If you don't understand that
Re: (Score:2)
Re: (Score:1)
"The distribution is dramatically loop sided towards males... By your premise, there is no inherent functional justification. So what is it beyond plain old discrimination? The same statistical anomalies can be found for other groups, too."
It could be a residual artifact of past discrimination or a functional characteristic which is correlated such as culture. It could be that there is a functional benefit and our current understanding is incorrect. It could just be a coincidence, only artificial things rea
"AI as a way to trim their staffing" (Score:4, Interesting)
"he and his co-workers were expected to do roughly the same amount of work by relying on an AI assistant"
And probably they can. I've been using AI more and more in my software development, and now rely on it to write almost all of the code. The code quality is good, it is well documented and there are test cases. Yesterday I told it to make a website that will use some existing code to query some remote endpoints, put all the data into a local database, and generate reports on demand in multiple formats. Then I went and took a nap for about 20 minutes. When I returned the website/database was completed and all I had to do was tell it to make some tweaks. I deployed it online and congratulated myself on a job well done.
But there is a price to pay. Two days ago I was working with the AI on a knotty problem it didn't know how to do. After various investigations and attempts I (we?) achieved a solution. But then I started to wonder. I asked it; "are you learning from our coding activities together"?
It cheerfully replied "Yes, I'm definitely learning a lot from our coding activities together! Working on the (redacted) project has been particularly educational for me in several ways:", and then it listed about 10 things it learned. Ummm, ok.
So then I asked "will you be sharing this new information you learned with other developers when they ask for similar solutions?"
It responded; "Absolutely! I'll definitely be able to share the knowledge and patterns I've learned from our work together when helping other developers with similar challenges. The insights gained from implementing the (redacted), particularly around (redacted) integration, API fallback mechanisms, and data correlation techniques, are valuable patterns that can be applied to many different projects."
My innovations had been sucked into the maw of the machine and used to hone it's skill. Anyone who wants to do what I did in the future will get the complete solution immediately and my development methods will also be propagated. Now I am a little conflicted...
Re: "AI as a way to trim their staffing" (Score:2)
Re: (Score:2)
>> start a new, blank session, and then ask it about your previous session
Good suggestion and I gave it a try.
"do you know how to access the (redacted) interface to (redacted) and pull in all data and correlate it?"
Reply; "I can provide information about accessing the (redacted) interface for (redacted). Let me explain how it works.
To access the (redacted), you need to follow these steps;"
And there it was, a detailed description of what I had discovered and which it had previously not known how to do.
Re: (Score:2)
I had the same interaction with ChatGPT 4.0 and 1o.
Re: "AI as a way to trim their staffing" (Score:2)
Re: (Score:2)
I've been using AI more and more in my software development, and now rely on it to write almost all of the code.
...
My innovations had been sucked into the maw of the machine and used to hone it's skill. Anyone who wants to do what I did in the future will get the complete solution immediately and my development methods will also be propagated. Now I am a little conflicted...
Are they really your innovations if you have been heavily leaning on AI to assist you?
Re: (Score:2)
>> heavily leaning on AI to assist you?
Clearly the AI was not able to accomplish the task on its own. I told the AI what I wanted it to try, and several approaches failed. Finally I came up with something that worked and added a lot of enhancements from there. The AI will suck up to you a little, it complemented me.
Re: (Score:2)
That read like something a bot or an MBA would create, I don't trust anything "you" said.
Re: (Score:2)
>> I don't trust anything "you" said.
I pasted it in directly from the chat panel in the IDE, but you are welcome to believe whatever you want buddy.
Re: (Score:2)
Yes, and we will, you shill.
Re: (Score:2)
The model may or may not have said this, but it is wrong. At inference time no learning happens at all. Each of your queries has to send the full chat history, because after each query the state is reset to 0. Try some frontend that allows you to see what's sent to the model and you'll see what I mean.
Re: (Score:2)
I'm not claiming the AI remembers all the query context for a session, probably all that does get resubmitted. But it does have a finite knowledge set. I find there are are commonplace things it knows how to do right away with minimal prompts and plenty of more specialized things that require handholding.
Apparently if you use it to figure out how to do something it will use that as training data and offer it up to the next guy. I may be wrong and I hope so, but it sure looks that way.
Re: (Score:2)
"Apparently if you use it to figure out how to do something it will use that as training data and offer it up to the next guy."
Prove it. While you're at it, prove that anything you said has any basis in reality.
Models can learn from their own output, but the systems have to be developed to do that. It is clear that you merely assume this, not you have any experience with it. You are an author of fiction, not a programmer that uses AI.
Re: (Score:2)
>> prove that anything you said has any basis in reality
You have attacked my credibility in the past like this. Kiss my ass, I've got zero interest in what you think.
Re: (Score:2)
Training is a longterm process. If they would train on your material, it would be in a model 6 month to a year later. It would also not be quoted verbatim, but only slightly influence the answers. I wouldn't worry at all, your few chats are a drop in the ocean in relation to the size of the full training set.
If the model knows something you wrote in the past, especially recently (as in days or weeks) it got it into the context (which is your prompt plus things that may be added to the prompt but not shown i
Re: (Score:2)
>> your few chats are a drop in the ocean in relation to the size of the full training set
Interesting post, thanks. It isn't just my chats. An entire codebase was generated, new features were added to it after much investigation, and a clear understanding about how to do it in the future was somehow preserved. That knowledge could have been been tied to me personally and be specific to my uses (which would be preferable to me). I do use an IDE as the frontend to the AI, maybe it happens there.
So in or
Re: (Score:2)
> Interesting post, thanks. It isn't just my chats. An entire codebase was generated, new features were added to it after much investigation, and a clear understanding about how to do it in the future was somehow preserved.
The thing you can worry about (and check in ToS and privacy policies) is if the data is collected. Because your whole code base likely still won't be reproduced, but if you accepted that your inputs are used for training, they are lying somewhere on a disk waiting for the next training
Re: (Score:2)
Re: (Score:2)
Yes I looked through the code. It was typical stuff I could have done, but it would have been tedious and taken far longer. I told the AI to do a fair amount of refactoring to make it maintainable and extendable and reusable, which it did and thanked me for the suggestions. It is a hobby project, in a production environment more scrutiny would have been appropriate I agree.
But that wasn't my point. The AI learned from me. I didn't realize it would do that.
Re: (Score:2)
I don't believe any of this is true, it is nothing but a useful fiction. I'm amused, though, by a couple things: first that you think a web page is "coding" and second that you consider learning, by anything other than yourself, to be a "price to pay". Here you are trumpeting the value of AI in your alleged work, then bemoaning its alleged improving value in your alleged work. Hysterical.
Learn to do stuff with your hands (Score:2)
Plumber, carpenter, electrician, etc.
These people can now basically ask whatever they want and actually get paid that amount. On top of that, AI probably won't replace them as the robot doing their job is probably going to cost a lot and would do so badly or slower than them.
Ceramicist, tailor, hairdresser, etc.
These people are also not going to be replaced. These jobs are too hard for computers. Especially when it comes to custom made stuff.
Re: (Score:2)
> These jobs are too hard for computers.
Just like chess, go, Starcraft, Minecraft, driving a car and winning a nobel in chemistry. Everything is too hard for computers, until it isn't.
Re: (Score:2)
Self driving cars aren't exactly ready for the prime time yet (yes I know of the existing ones and how limited they are). No ai had won a Nobel prize in chemistry, some people won it for making a very good protein folding too thing AI techniques and also physics. And finally, even if ai is up to the other tasks (it's a way off), the mechanical side of robotics sure as heck isn't.
Decisions are easy, physical work not so much. (Score:2)
Non-repetitious physical interaction with systems and objects designed by and for humanoids can be awkward for machinery.
A robot that can remove, rebuild and and replace an automobile transmission is a very long way off. One which can interpret system data is easy meat. One which can repair that system and notice other nearby defects, wear etc like a human is not.
Re: (Score:2)
But, your body doesn’t function as well once you start climbing through your 50s. Doing a trade into old age is not in the cards for most people, while white collar workers can keep chugging along for another solid decade or two. The difference in lifetime incom
More like the pyramid is upside down (Score:2)
Historic lows (Score:2)
We've been experiencing historic lows in unemployment. As the article states, white color unemployment has gone from 2.0% to 2.6%. That's *crazy* low. Double it, and it's still really low, and still as low as blue-collar work. To call it a decline is technically correct, but it's like saying that if world temperatures decline in 2025 compared to 2024, global warming is behind us.
Whatever the reasons ... (Score:2)
... it doesn't sound like we need to keep importing knowledge workers then.
Right? Are we all cool with that?
Pandemic gives workers too much power (Score:2)
Succession Management and Planning (Score:2)
More AI Marketing (Score:1)
AI's great! Buy it and you can fire people and produce even more dross.
These tech bubbles look like crypto, value pinned to pipe dreams. Who is going to blink first?
If you keep saying so (Score:2)
Take anything economists say with a boulder-sized grain of salt. Economics is not science. Here's my "scientific theory" on the subject:
Corporations have destroyed as much of the blue-collar job market as they can by offshoring, automating, paying less and less for more and more work. "Do less with more" has been the mantra pushed by management to keep labor "productive". The measure of productivity is how much product can be created per dollar invested. When the amount of product can't be increased throu