Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
AI IT

40% of Workers Will Have to Reskill in the Next Three Years Due to AI, Says IBM Study (ibm.com) 129

IBM's business research organization (the IBM Institute for Business Value), released results from a new global study. Its conclusion? "The world of work has changed compared to even six months ago." Executives surveyed estimate that 40% of their workforce will need to reskill as a result of implementing AI and automation over the next three years. That could translate to 1.4 billion of the 3.4 billion people in the global workforce, according to World Bank statistics. Respondents also report that building new skills for existing employees is a top talent issue.

Workers at all levels could feel the effects of generative AI, but entry-level employees are expected to see the biggest shift. Seventy-seven percent of executive respondents say entry-level positions are already seeing the effects of generative AI and that will intensify in the next few years. Only 22% of respondents report the same for executive or senior management roles.

AI can open up more possibilities for employees by enhancing their capabilities. In fact, 87% of executives surveyed believe employees are more likely to be augmented than replaced by generative AI. That varies across functions — 97% of executives think employees in procurement are more likely to be augmented than replaced, compared to 93% for employees in risk and compliance, 93% for finance, 77% for customer service and 73% for marketing...

With AI primed to take on more manual and repetitive tasks, employees surveyed report engaging in impactful work is the top factor they care about beyond compensation and job security — more important than flexible work arrangements, growth opportunities and equity. On top of that, nearly half of employees surveyed believe the work they do is far more important than who they work for or who they work with regularly...

ZDNet explains the report's methodology: To find answers to these questions, IBM pulled data from two prior studies, one survey of 3,000 C-level executives across 28 countries and another of 21,000 workers in 22 nations...

According to IBM IBV research, tech adopters who successfully reskill to adapt "technology-driven job changes report a revenue growth rate premium of 15% on average" and those who focus on AI "see a 36% higher revenue growth rate than their peers." "AI won't replace people — but people who use AI will replace people who don't," said IBM in the report.

The new skill paradigm shifts technical skills that were typically prioritized, such as proficiency in STEM, which was the most critical skill in 2016, to the least priority in 2023. The reason is that now tools like ChatGPT allow workers to do more with less knowledge, as noted by the report. Now there is a bigger emphasis on people skills such as team management, the ability to work effectively in team environments, the ability to communicate effectively, and the willingness to be adaptable to change, which all shifted to top the most critical skills required of the workforce in 2023.

The report ultimately suggests HR leaders redesign work and operating models "to shepherd their organizations into the future."
This discussion has been archived. No new comments can be posted.

40% of Workers Will Have to Reskill in the Next Three Years Due to AI, Says IBM Study

Comments Filter:
  • Nonsense (Score:4, Interesting)

    by Retired Chemist ( 5039029 ) on Sunday August 20, 2023 @09:51AM (#63782182)
    Barders, service workers generally, auto mechanics etc. should see little effect. Even in their jobs could be replaced, it is doubtful if it would be cost effective. People in the computer field may have feel an impact. Instead of doing what they are doing know, they will be cleaning up the disasters caused by so-called AI.
    • People like mechanics may have an improved life at work as a result of AI, because it may help with troubleshooting, and make it easier to find instructions for less-frequent types of repairs.

      In the computer field, I see a similar impact. The jobs that can be replaced by AI, are jobs nobody wants, and _should_ be replaced by AI. The jobs the require thinking, aren't going away.

      • The jobs that can be replaced by AI, are jobs nobody wants, and _should_ be replaced by AI.

        "Computer, I just wrote this code. Please debug it!" Too bad AI can't do that.

        • Re:Nonsense (Score:4, Interesting)

          by Tony Isaac ( 1301187 ) on Sunday August 20, 2023 @10:36AM (#63782288) Homepage

          Actually, AI _can_ debug code, at least for certain classes of bugs.

          Recently I wrote some code that, when it ran, returned an obscure error. I fed the code into ChatGPT and asked it how to fix the problem. It gave me the correct answer.

          • That must've been a fairly short sampling. When I tried this as a test with about 200 lines of code, ChatGPT told me it was too long for it to handle.

            • ChatGPT does have strict input limits. But GitHub Copilot takes in your entire codebase and can help with suggestion and bug analysis of the entire project.

      • People like mechanics may have an improved life at work as a result of AI, because it may help with troubleshooting, and make it easier to find instructions for less-frequent types of repairs.

        I doubt large language models will help with this... and that's the only "AI" they're going to have access to.

        Where is the money in decreasing the amount of work that shops have to do to fix rare problems with old cars? Wouldn't it actually reduce both revenue and profit? So where is the motive to invest in it?

        • I've personally found LLMs to be very helpful in solving mechanical problems that I've faced as a DIYer. It's not hard to imagine that it could help professional mechanics, just as it has helped me as a professional developer.

          Mechanics don't charge by the actual hour, they charge by the *book* hour. So if the book says a certain job should take 2 hours, that's what they bill you for, whether it takes 30 minutes or 4 hours. So if LLMs can help them reduce the number of hours worked, you bet it would add to t

      • Re: (Score:3, Informative)

        by XXongo ( 3986865 )

        People like mechanics may have an improved life at work as a result of AI, because it may help with troubleshooting, and make it easier to find instructions for less-frequent types of repairs.

        40% of auto mechanics are going to have to find other jobs because electric car engines take a lot less maintenance than IC engines. Still have to fix body and chassis problems, but no tune up every 12,000 miles. And many of the problems will be diagnosed remotely, most of which can be done by an AI.

        • People like mechanics may have an improved life at work as a result of AI, because it may help with troubleshooting, and make it easier to find instructions for less-frequent types of repairs.

          40% of auto mechanics are going to have to find other jobs because electric car engines take a lot less maintenance than IC engines. Still have to fix body and chassis problems, but no tune up every 12,000 miles. And many of the problems will be diagnosed remotely, most of which can be done by an AI.

          I'm less sure of that than I used to be. There's pretty severe shortages of workers in many trades; anecdotally, it took me almost 2 months to get my SUV in for a leaky radiator. The fix took 4 hours. Granted it was an import, so there aren't a ton of shops that do them, but even Carmax has been booked out a few weeks for simple things recently. While the equivalent electric might not need as much maintenance they're certainly going to need some.

          There's a decent chance that what this "AI disruption" (or e

        • And many of the problems will be diagnosed remotely, most of which can be done by an AI.

          If you've used ChatGPT to provide code suggestions, you know by now that it still takes a thinking programmer to evaluate the suggestion and determine whether it's correct or not. ChatGPT can get close, but it's that last 20% that's hard.

          • by XXongo ( 3986865 )

            And many of the problems will be diagnosed remotely, most of which can be done by an AI.

            If you've used ChatGPT to provide code suggestions, you know by now that it still takes a thinking programmer to evaluate the suggestion and determine whether it's correct or not. ChatGPT can get close, but it's that last 20% that's hard.

            Diagnosing problems in a car consists of choosing from a list of options that is vastly shorter than the list of "every possible thing that some human could ask an AI to write about".

            There will be a need for humans in those rare cases where it's an unusual and hard to diagnose problem, but actually, the human mechanics have a mixed record of dealing with those, too.

            • The problem isn't so much the length of the list, it's the selection of the correct option from that list. To name a simple example, if your car won't start, it could be the battery, or the alternator, or the timing belt, or the fuel pump, or any number of other possible causes. AI can give you a list of most likely issues, but it can't correctly tell you which precise one. And yes, of course, human mechanics struggle sometimes to make the correct diagnosis too.

              • by XXongo ( 3986865 )

                The problem isn't so much the length of the list, it's the selection of the correct option from that list. To name a simple example, if your car won't start, it could be the battery, or the alternator, or the timing belt, or the fuel pump, or any number of other possible causes.

                Nope; for an electric car, most of the items on that list aren't there. "Battery", yep.

                And, given a number of sensors, yes, can be diagnosed remotely.

                • Clearly, I wasn't talking about electric cars. Yes, it's likely that they will simplify the job of the mechanic. But that influence is completely separate from the potential influence of AI. The point was not to focus on the automotive industry specifically, but on the impact of AI on various trades that aren't seen as "IT". EVs only muddy the issue by bringing in factors that aren't relevant to AI.

                  And regarding sensors...I can't count the number of times I had a "check engine" light, and the mechanic fixed

                  • by XXongo ( 3986865 )

                    Clearly, I wasn't talking about electric cars.

                    You were replying to a thread that started

                    40% of auto mechanics are going to have to find other jobs because electric car engines take a lot less maintenance than IC engines.

                    ?

                    • You'll notice that your quote was replying to *me*, and I said nothing about EVs. YOU brought that into the thread, not me.

        • Oil changes aren't going to be necessary, but electric cars still need work done on the rest of the vehicle. Breaks still need replacement, tires need rotation and the alignment adjusted, etc. The shear number of other things being computerized that seem to breakdown all the damn time will probably compensate to some degree.

          I don't know how much money mechanics make on oil changes, but given how little they cost and how many different places try to compete on price for them (almost as a loss leader so th
    • by dvice ( 6309704 )

      Reskilling means that mechanic learns to use AI for example to help him work faster or to automate some tasks like ordering new parts. It doesn't mean that person has to change career.

      • AI can't automate ordering new parts unless it is also able to keep track of how many parts are being used. And that requires either a very sophisticated vision system, or a lot of data entry. (And if you're doing data entry, you don't need AI to automate the ordering part. That's just bog standard computer programming).
        • by HiThere ( 15173 )

          You are technically correct, and otherwise wrong. The storeroom can change how they accept orders for parts, how they predict which parts should be on hand, etc. They can use AI to do parts of this. You may regret the change, but it can be done if management decides to do it. The analogous changes are possible in many other parts of the business. Consider all the self-service checkout counters at grocery stores.

          I still think the estimate is extremely high, but it's definitely not impossible if AI catch

          • by tlhIngan ( 30335 )

            You are technically correct, and otherwise wrong. The storeroom can change how they accept orders for parts, how they predict which parts should be on hand, etc. They can use AI to do parts of this. You may regret the change, but it can be done if management decides to do it. The analogous changes are possible in many other parts of the business. Consider all the self-service checkout counters at grocery stores.

            The stock room should be using AI to help predict part usage so it always has the parts it needs

          • The storeroom can change how they accept orders for parts

            This has nothing to do with generative AI. Making predictions about your warehouse has always been a good idea, and the idea that executives apparently just found out about a 200 yo concept is baffling. And yes sure AI can help with that, but as a field it's been around for about half a century.

          • by q_e_t ( 5104099 )
            I don't see that AI is a fad, given it has been making gradual in roads into our lives over thirty years.
          • by RedK ( 112790 )

            > The storeroom can change how they accept orders for parts, how they predict which parts should be on hand, etc.

            This isn't AI, this is just parts use trend and stock levels. This has been the norm in the industry for ages. Tracking sales vs time of year vs stock levels and predicting when to order stuff.

            You guys seem to invent problems for AI to solve that have already been solved through simple algorithms.

            • by HiThere ( 15173 )

              Saying is isn't being done with AI doesn't say it can't be done with AI. AI can be used for LOTS of things where you don't need to use it, and it is very likely to change how those things are done. Not necessarily for the better.

        • by XXongo ( 3986865 )

          To the contrary, ordering parts is one thing that an AI can do well, assuming that the AI has computing power at least equal to an Excel spreadsheet.

          (It will, of course, require that the repair bills have to be accurate about what part they're charging the customer for. But that's a feature, not a bug. But everything has bar codes these days, that won't be hard.)

        • A mechanic could dictate his assessment verbally and GPT can format that into nice JSON to send to the ordering API. The same way doctors dictate/take notes during each consultation, and then AI needs to parse that mess into normalised form.
        • by q_e_t ( 5104099 )
          Or RFID or part numbers that can be read. Attempts at automation of parts recognition goes back at least twenty-five years. Issues have included the need to get a clean, 3D image which requires multiple view points as there may be parts that look similar from one angle and that usual false positive and negative issues. Even given a match on shape you might have different materials in which case you still need part numbers or data entry based on taking parts out of a hopper. Another issue is if an mechanic i
        • by ranton ( 36917 )

          There is significant room for improvement where AI could be involved. A mechanic could dictate a recording to the customer describing what repairs he recommends for the car, and AI could use that recording to input those recommendations into a database and generate a form for the client to select which repairs to do. It could answer questions for the client in that form. The AI system could do a better job of anticipating parts which should be on hand, even potentially ordering parts based on the conversati

    • Paper shufflers, memo writers, the low end of the creative fields... Those are going to be hit hard.

      First, by stupid managers trying to save money with the latest trend, too dumb to notice the tech isn't ready yet. But then the tech will be ready, and there will be a massive culling of jobs.

      • I doubt even those people will be hit by AI ever much less than anytime soon.

        This is just IBM's services division trying to drum up new business.

        • They will be hit, but it is going to take more than 3 years. It is funny how many (large, tedious) reports I have worked on over the years managing tremendous amounts of data the value is often the insights gained through the process of working through data to identify information. Sure, that is what the "AI" is supposedly doing, but the reality is that we aren't close to being able to answer many of the "why's" with "AI".

          I can see how AI will make running complex sensitivity analysis significantly easie

          • by HiThere ( 15173 )

            Mmmpfh. If you give it more than 3 years I don't think you can make a good prediction at this point. Too many things are in flux, and there are too many possible interactions. It's definitely going to affect the byte slingers (programmers, writers, etc.) first, because that doesn't require any additional hardware. HOW it will affect them depends on whether they can make the answers more accurate. (Screenwriters don't need much accuracy, but they need some. They need to be able to guess what will move

        • Yes, I've been on the blunt end of IBM/Accenture (and predecessors) consultative services divisions 4 or 5 time in the last 20 years, each time the expense was inordinate & the results were, to be blunt, fucking dire. 3 out of the 5, so 60%, failed completely, finished, kaput. Caused by massively over specified, massively over "Managed" solutions to things that didn't need solving. In each case called in by middle management MBA's climbing the greasy pole.

          If you hear IBM + AI or Accenture + AI in the sa

          • In comparison I feel lucky. I dealt with direct IBM once and Accenture once each in my entire career.

            At 4-5 times I'd have to throw myself off a bridge.

      • On the flip side, proofreaders will probably make a killing fixing the mistakes in AI-generated instructions and code over the next few years until it significantly improves.

        As long as AI keeps making dumb mistakes like recommending tourists visit an Ottawa food bank in a travel blog, the humans have plenty of work to do.

    • "Executives surveyed estimate"

      We all should have stopped reading right there.

  • HR? (Score:5, Insightful)

    by Potor ( 658520 ) <farker1 AT gmail DOT com> on Sunday August 20, 2023 @09:55AM (#63782192) Journal

    The report ultimately suggests HR leaders redesign work and operating models "to shepherd their organizations into the future."

    HR is the first thing that should disappear.

    • Re:HR? (Score:4, Insightful)

      by Opportunist ( 166417 ) on Sunday August 20, 2023 @09:59AM (#63782198)

      The fun bit is that the parts of the company that could easiest be replaced by AI are management and HR. Guess which parts will almost certainly NOT be affected by AI.

    • by XXongo ( 3986865 )

      The report ultimately suggests HR leaders redesign work and operating models "to shepherd their organizations into the future."

      HR is the first thing that should disappear.

      And very likely will be. Companies think of employees as interchangable parts; "we can just automate all that employee stuff, it's no harder than automating warehouse inventory" is absolutely they way they think. Hiring interviews? Chatbot with an algorithm that checks off the right buzzwords. All the touchy-feely stuff about employee relations? Do it with a chatbot. Performance bonuses? Algorithm.

  • The new skill paradigm shifts technical skills that were typically prioritized, such as proficiency in STEM, which was the most critical skill in 2016, to the least priority in 2023. The reason is that now tools like ChatGPT allow workers to do more with less knowledge, as noted by the report.

    By all means, let us replace subject matter experts with large language models and encourage doing more with less knowledge. True innovators will be able to apply this technique in fields that were previously ossified

    • Yeah, it sounds like the MBA types don't get that GPT is coming for their jobs first.

      I'm not sure what it will do to "quick learning generalists" like myself who built a career around becoming a subject matter expert in something when there weren't enough to go around and transitioned to the next thing when the time came. My best guess is that things will become commodity skills faster; they traditionally have taken about 5 years. At less than 2 years my model wouldn't pan out; I need about 3 to make it w

      • by HiThere ( 15173 )

        Management usually doesn't outsource itself. Even when it should.

        OTOH, I expect that this will quickly act to replace large amounts of middle management. But they don't have leverage in this decision.

  • by MpVpRb ( 1423381 ) on Sunday August 20, 2023 @10:05AM (#63782212)

    There's the problem
    They are fully on the hype train
    I have no doubt that things will change
    I have little confidence in executives and their crystal balls

  • It strikes me that IBM is trying to pitch their own AI-augmented consulting services. Sales pitches are *always* overly optimistic.

    • by Luckyo ( 1726890 )

      Obviously. And it's equally obvious that no major human nation has resources anywhere near needed for such reskilling speed if it was even possible.

      And that's before the ability of IBM and its competitors to actually adapt LLMs to tasks other than "creative writing and drawing" as well as "human reading a script" which most certainly does not represent anywhere near 40% of workers. Frankly it's set to replace current generation chatbots and first tier of phone sales and support, not people doing the repetit

      • First my comment on the report. It's written by Consulting, not Research. So it's just a thinly veiled advert.

        > Frankly it's set to replace current generation chatbots and first tier of phone sales and support, not people doing the repetitive tasks like plumbers.

        It won't replace anyone. It's probably the only bit I agree with in the report.

        It is reduce the need for experts in their fields. The newbie plumber using AI will be cheaper and be able to compete against the expert not using AI.

        • by Luckyo ( 1726890 )

          >The newbie plumber using AI will be cheaper and be able to compete against the expert not using AI.

          Reality check: errors newbie plumbers typically do is not in doing incorrect repairs (which is where LLM could actually be helpful). It's mainly in fucking something up in the actual process, like accidentally installing a part upside down.

          And that is why newbie plumbers tend to make very expensive mistakes when they make mistakes. And that is something LLM cannot help with.

  • Watson (Score:4, Informative)

    by phantomfive ( 622387 ) on Sunday August 20, 2023 @10:23AM (#63782252) Journal
    IBM spent a decade trying to sell their own version of an LLM, Watson, but nobody wanted it.

    There is more hype around LLMs right now, but in some ways they are worse than Watson (ie, Watson was almost always right when it said something).
    • by HiThere ( 15173 )

      But Watson was a lot more expensive to run. Some of these things, when "fully trained", can run on a cell phone...with no network connection. (I'm not quite sure what that means. If the stories had said "Raspberry model X" I'd be more certain, but it definitely implies that any decent desktop should be able to run it (as far a computational requirements...it might need a mic or camera).

    • by Equuleus42 ( 723 )

      Agreed. Watson could give a certainty percentage with each answer, which would be really useful with the latest round of LLMs.

  • I can easily imagine this to be true.

    When I got into my new IT position where I had to replace a lot of co-workers because they could not afford more staff, I realized I was up to my ears in work. The only competent co-workers I had was those working overseas I communicated with to get help in my new job, that was hard, because of the company reconstruction they suffered from low-staffing too and had to work overtime, one of them worked till 10 in the afternoon and even saturdays.

    I used my OpenAI account to

    • by gtall ( 79522 )

      You: Alex, I'd like punctuation for $400, please.

      Alex: Periods and semicolons separate distinct ideas.

      You: What prevents my stream of consciousness from becoming blah, blah, and blah?

      Alex: We have a winner!!

    • Can you give an example of what sort of problem it solved for you? I think one of the reasons there is so much disagreement on AI is that there are such a wide range of development tasks, from scripting for games to financial transaction processing to web front ends to engineering simulation codes. Its quite possible AI is useful for some of these but not at all for others.
      • by dvice ( 6309704 )

        1) Creating feedback for students: https://news.stanford.edu/2023... [stanford.edu]
        2) Playing Minecraft (the important part here is how it played, because to me it looks like this could be applied to any game or gamified real world problem, I think this has quite big potential) : https://www.zdnet.com/article/... [zdnet.com]
        3) Khan Academy uses it also as a teacher. You can ask it to help you solve a math problem and it will guide you to solve it yourself.
        4) Translation. ChatGPT is actually quite good at it. If I had to translate so

  • Blockchain! (Score:5, Funny)

    by dysmal ( 3361085 ) on Sunday August 20, 2023 @10:27AM (#63782266)

    "Now you just need Blockchain enhanced AI which will create the synergy needed to expand your presence in the Metaverse for year over year growth while lowering your costs as you move to the cloud!"

    -Salesperson

    • That's right... I almost forgot that IBM was pimping blockchain solutions a few years ago before they started pimping AI solutions again.

      If remember right, the IBM sales hype cycle went from "e-business", to "cloud", to "AI" (Watson, the first attempt), to "blockchain", back to "AI" (GPT, the second attempt). Of course, what they were really selling was the outsourcing of development jobs to whatever developing nation had the cheapest labor at the time.

      • by dysmal ( 3361085 )

        I can't fault IBM for this history of sales fads/attempts. If people are dumb enough to believe the hype, you might as well milk that cash cow!

        I'm sure the /. neck-beards remember their fair share of Y2K fleecing.

    • by Lips ( 26363 )
      Bingo! I pretty much filled my card in 1 sentence.
  • Looks like they are basing it on the statement "According to IBM IBV research, tech adopters who successfully reskill to adapt "technology-driven job changes report a revenue growth rate premium of 15% on average" and those who focus on AI "see a 36% higher revenue growth rate than their peers."

    Broadly available AI is brand new, so its not at al surprising that very early on the growth rate will be high - its easy to grow from near zero. That growth is likely to saturate.

    Using AI may become a useful
    • I'm confused about why people talk about AI like ChatGPT as being a skill at all. Don't you just have to type a question? That's not a skill.
      • by HiThere ( 15173 )

        Well, actually if you want to get a certain answer, it is a skill. But this wasn't an estimate about ChatBots, or at least it didn't appear to be so in the summary, but rather about AI, of which ChatBots are only one aspect, albeit a highly visible one.

  • by williamyf ( 227051 ) on Sunday August 20, 2023 @10:49AM (#63782314)

    Give the perfect example of how workers will have to retrain.

    In common law countries, the figure of the paralegal* is well known. In this age of AI (and specially if one uses specialized AIs, say, an AI tuned for legal usage), law offices will need less paralegals, but paralegals that can write effcient AI prompts, and paralegals that can verify that the AI did not Halucinate.

    So, either 'em paralegals retrain to be more effcient users of AI, or get out of work in the culling when their numbers are reduced.

    Same thing with many more areas of work. Many human works will be replaced not by AI alone... But an AI coupled with a human that knnow how to use it could replace a lot of human only workers.

    * For those of you in Civil law countries, think erin brokovich

  • by VeryFluffyBunny ( 5037285 ) on Sunday August 20, 2023 @10:51AM (#63782318)
    ... based on its specific market & activities and a reporter generalises it to the entire global workforce. Yeah, great reportin' there!
  • I use copilot and ChatGPT to assist my coding, does that mean I've "reskilled"?

    AI can open up more possibilities for employees by enhancing their capabilities. In fact, 87% of executives surveyed believe employees are more likely to be augmented than replaced by generative AI. That varies across functions — 97% of executives think employees in procurement are more likely to be augmented than replaced, compared to 93% for employees in risk and compliance, 93% for finance, 77% for customer service and 7

    • "Learn to Code" is replaced with "Learn to Coal" because Progress!!!(tm).

      Wealthy elites will be more comfortable and everybody else will be doing hard labor.

      It's like a prison camp but they hand out "I Voted" stickers so it's different.

      If you complain the other prisons will reprimand you.

      Good luck, city folks.

    • Using tools like GitHub Copilot and ChatGPT to assist you in coding can be considered a form of augmentation or support for your existing skills, rather than a complete reskilling. Reskilling usually involves acquiring new skills or significantly changing your skillset to adapt to a different job role or industry.
      • Using tools like GitHub Copilot and ChatGPT to assist you in coding can be considered a form of augmentation or support for your existing skills, rather than a complete reskilling. Reskilling usually involves acquiring new skills or significantly changing your skillset to adapt to a different job role or industry.

        Which makes me suspect that 40% is a massive overestimate.

        If we're being optimistic (pessimistic?) then maybe 40% is the hardest affected industries like customer service.

  • More likely, authors of IBM study will have to reskill, as their "studies" are proven wrong.

  • They never really explain what work is going to be requiring GAI skills. Seems like a bullshit article meant to hype up GAI services offered by IBM.

  • Regardless of AI it's pretty common for generational shifts to displace workers as some excuse or another justifies terminating higher paid employees in favor of lesser ones, hence why I don't personally know anyone on the verge of having "the largest generational wealth transfer in history" affect them. For example jobs for people trained as musicians for movie theaters were replaced by people who did other musical work, but they weren't filled by the same people. Those people just had reduced career prosp

    • by CAIMLAS ( 41445 )

      Judging by your UID and your quote of a title of an article flying about in the legacy state-operated print media we've all come to acknowledge as a propaganda column of the government, I'm going to assume you're on the older side of things and don't see boomers transferring wealth to younger generations.

      The passing of the most gluttonous generation the country has (yet) known, with the fewer-in-number remnants of their progeny (or the banks) receiving their inheritance both isn't likely to impact much of a

  • Is that AI coded speech for No More WFH ?
  • It certainly doesn't seem to be getting exponentially better, as many commentators expected. Many say the data it uses is becoming polluted with its own output.

    "Millions of jobs are doomed because of AI" may be a premature conclusion.

    https://gizmodo.com/study-find... [gizmodo.com]
    https://arstechnica.com/inform... [arstechnica.com]

  • AI can't do most jobs. AI can't even code without beginners mistakes.

    The only thing AI will be doing in the 3 three years is exposing lazy copy/paste people.

  • are 40% of workers working in a call centre?!?
    That's the only job I see can be fully substituted by AI.

  • The dubious word "reskill" is definitely not a verb. The IBM lawyers need to regrammar and revocabularize themselves, I think.

  • So they expect to release 60% of their workforce, requiring everyone remaining, who will all be under 48, to be retrained on AI? Sorry RedHat, you're fucked.

  • Oh look, IBM making a completely BS claim about AI. Maybe they will be right this time? *they won't*

    But, it does give me a giggle to think one day in the not too distant future blue collar workers will be the high paid ones, followed by low-skilled service workers, with white collar scraping the bottom of the barrel.

  • by Hoi Polloi ( 522990 ) on Sunday August 20, 2023 @07:25PM (#63783388) Journal

    "Our study however found that 0% of executives and consultants will need to be retrained or will be replaced by AI."

  • Sir, would you like fries with that?
  • Most manufacturing/Production companies won't use AI cause of the costs $$$ and it can't do what's needed ! Merial in Athens, Ga is still using paper for 99% of the forms onsite. I tried for 12 years to get them to computerize all forms and use PC's/Macs/tablets to improve production and reduce errors ! There are 6-10 people in QA/Production whose main job is to review all the paperwork ! Computerizing paper could reduce the people down to 2-3 reviewers and save a bunch of trees ! And here we are in 202
  • And exactly who is going to pay for that?

Behind every great computer sits a skinny little geek.

Working...