all 11 comments

sorted by: hot top controversial new old
[–] 11 points 11 months ago (8 children)

The study was conducted using a labor simulation tool called the Iceberg Index, which was created by MIT and Oak Ridge National Laboratory.

"We built an AI and it told us how good AI is."

  • source
  • hideshow 8 child comments
  • [–] [S] 6 points 11 months ago (7 children)

    Look at it this way, AI is simply exposing the deep absurdity of late capitalism. Much of the economy in the West consists of what Graeber called bullshit jobs which are roles that even people performing them struggle to justify. I'd argue these types of jobs are uniquely vulnerable to replacement by AI.

    It's because these jobs produce nothing of tangible, material value. Building a bridge or diagnosing an illness requires engagement with physical and ethical reality. You are accountable to laws of physics, to human bodies, to measurable outcomes. That sort of a job is going to require a human in the loop. An AI tool can be helpful for the worker where it could help zero down on a diagnosis for example, but the final decision needs to be made by a person who can be held responsible. There is little chance that AI, in the form we have today, can replace such jobs.

    But much of the modern service and knowledge economy operates in a realm of manufactured meaning. Marketing campaigns, branding, corporate compliance, and middle management layers are roles built around persuasion, perception, and bureaucratic performance. They generate what Baudrillard would call simulacra. These are outputs detached from real use-value. AI, as a sophisticated pattern matcher, thrives here precisely because the work was already semantically hollow.

    So while capitalism created these roles to absorb surplus labor and sustain consumption, AI now reveals their contingency. The real contradiction here is between value and bullshit. It is between work that sustains society and work that just sustains the system.

  • source
  • parent
  • hideshow 7 child comments
  • [–] 4 points 11 months ago* (3 children)

    I absolutely believe bullshit jobs are threatened by AI, I'm just skeptical about simulations being produced by pro-business private schools that have every incentive to flatter their corporate sponsors. MIT has received over two hundred million dollars of investments from IBM for AI research and is seeking an additional billion dollars to build out its AI campus. They're beneficiaries of the bubble.

    They have a lot of incentives to lie, here. It sounds like they just built a simulation to tell them what they wanted to hear.

  • source
  • parent
  • hideshow 3 child comments
  • [–] [S] 3 points 11 months ago (2 children)

    It's going to be a self fulfilling prophecy. After all, these studies are produced to convince CEOs to make certain types of business decisions. Their whole point is to convince execs to make the types of policy decisions that they outline.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 2 points 11 months ago (1 child)

    That's why I wanted to highlight the absurdity of "We made an AI to tell us AI is good!" because it shows their nature: a snake eating its own ass.

  • source
  • parent
  • hideshow 1 child comment
  • [–] [S] 3 points 11 months ago

    Oh yeah, the whole thing is absurd of course. In my opinion, AI is just exposing the fact that capitalism is a system of engineered scarcity which forces people to do useless work in order to continue existing.

  • source
  • parent
  • [–] 1 point 11 months ago (2 children)

    Not the person you're replying to, but I think the point is that the study is bullshit, even if the point is apt.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 6 points 11 months ago

    Start with middle management please.

  • source
  • [–] 2 points 11 months ago

    This is the shittiest time line.

  • source
  • 30 comments in c/fuck_ai@lemmy.world view that post
    [–] 50 points 11 months ago (11 children)

    Horse shit. Not a single example of what exactly “AI” will replace, just a bunch of hand-wavy “the simulation says” quotes. Just another “news” article in the endless stream of bullshit propaganda, most likely funded by the AI bullshit machine.

  • source
  • hideshow 11 child comments
  • [+] [S] -12 points 11 months ago (10 children)

    Not a single example of what exactly “AI” will replace

    Lawyers? Doctors? Novelists? Digital art creators? Medical researchers? Astronomers?

    Those are just the ones I could think of, but basically anything that uses words, pattern recognition, or graphics.

  • source
  • parent
  • hideshow 10 child comments
  • [–] 27 points 11 months ago (9 children)

    Anyone who trusts an AI lawyer or doctor will get exactly the kind of service they deserve.

  • source
  • parent
  • hideshow 9 child comments
  • [–] [S] -3 points 11 months ago* (7 children)

    If an AI lawyer or doctor costs 1/10,000 the cost, a lot of people will risk it. Hell, insurance companies would be drooling over the opportunity.

  • source
  • parent
  • hideshow 7 child comments
  • [–] 7 points 11 months ago (1 child)

    insurance companies would be drooling over the opportunity

    It's actually the opposite: https://futurism.com/future-society/insurance-cyber-risk-ai

  • source
  • parent
  • hideshow 1 child comment
  • [–] 2 points 11 months ago

    I think an especially interesting use case for AI is yearly cancer screening. Instead of doing screening only every 5-10 years (and only for specific body parts), you could screen yearly, if the costs of the tests drop dramatically. As we all know, early detection is key to effective treatment.

  • source
  • parent
  • [–] 2 points 11 months ago (2 children)

    None of these things will be replaced. They will be (and are) assistants, which is nice, but the idea that whole professions will be usurped is fantasy.

  • source
  • parent
  • hideshow 2 child comments
  • [–] -2 points 11 months ago

    Better service with a more reliable and repeatable doctor for cheaper? I mean, yeah, everyone deserves that.

    Literally a hard coded bot that just follows the exact same steps but coded by a professional doctor would be better than most american health care. The sheer ability of a bot to be repeatable and consistent immedietely places them in a different echelon for delicate procedures in the medical field. While LLMs have inherit vulnerabilities such as hallucinations, an actual AI trained on all the precedent cases for our legal system would be the greatest lawyer ever created. Effectively free because it can be run on your local computer, and yet with all the legal knowledge of our entire country in a fairly indexable package? Sounds incredible.

    The biggest bit of muance here is that, admittedly, i dont realistically see AI replacing humans entirely at least in the short term. For creative processes that youre not willing to license out to a generative model largely trained on poisoned ai slop, the only way to get a bot to produce anything decent is to give it near human level intelligence at least within its field. Human level intelligence is a much trickier thing to create than basic dumb automation.

  • source
  • parent
  • [–] 72 points 11 months ago (1 child)

    To be clear: this data comes from an agentic AI simulation.

  • source
  • hideshow 1 child comment
  • [–] 20 points 11 months ago

    12% of the workforce is CEOs?

  • source
  • [–] 18 points 11 months ago

    I think it's time for tech CEOs and healthcare CEOs to fear the common man.

    If one CEO doesn't play ball, have them catch an assassin's bullet and let their replacement know he better tighten the fuck up or he will be next.

  • source
  • [–] 12 points 11 months ago (2 children)

    To maximize labor savings, I think the AI could definitely replace the C-Suite of all companies. The humans currently in those roles are already wrong about what is actually needed at the ground level so no one will notice.

  • source
  • hideshow 2 child comments
  • [–] 2 points 11 months ago (1 child)

    The issue is that a lot of times, the "value" these executives bring is not with their managerial, decision making skills. It's their status, industry recognition, and social connections. How can AI replace that?

  • source
  • parent
  • hideshow 1 child comment
  • [–] 10 points 11 months ago (1 child)

    I have questions about the macroeconomic implications of AI replacing jobs. Does this mean that those workers will shift to other non-AI jobs? Or does this mean that unemployment goes up and there just aren't any jobs for people any more?

    The way I am thinking about this, if corporations are able to hoard more wealth and increase profits substantially by getting rid of the need to pay people, how does the economy function if the money that would be paid to those workers is no longer circulating back into the economy?

    If people then will get money from a UBI instead of labor income, who pays the government taxes? Corporations? Consumption taxes on people?

    If corporations manage to get labor costs to near zero, profits go to near infinity, which is the goal of profit maximizers. But then there is no money in the hands of individuals to be able to pay to consume the goods or services these corporations provide? Is this desire to replace human labor with AI not just a living example of the myth of Icarus?

    Any economists out there interested in breaking these issues down into more of a layman language for me? Thanks!

  • source
  • hideshow 1 child comment
  • [–] [S] 2 points 11 months ago

    Does this mean that those workers will shift to other non-AI jobs? Or does this mean that unemployment goes up and there just aren’t any jobs for people any more?

    Well, it's a nuanced matter. In the short term? People will shift jobs. But theres 2 peovlems with that: (1) this creates a glut of workers in the short term because there can only be so many plumbers; (2) it takes a long time to shift careers because there's time needed to retrain.

    For 2, take doctors for example. If AI can do the job, that doesn't mean that MD can just take off and be a plumber. That takes time, and its faster to make new Ai for other doctor positions too. In they retool to a plumber, they need to learn how a water heater works. Add onto that the massive investment into robo-tizing that. Also, it was a huuuuuge investmemt into being a doctor. It takes k-12 grade school, 4 year for your bachelor's, then 4 more for the md. That's 20 years.

    For 1, if there's a glut, that means wages are suppressed. Also, that becomea an attractive field for AI to take over.

  • source
  • parent
  • [–] 9 points 11 months ago

    I feel like you have to make a lot of underlying assumptions to call a hard shot like this, but to be fair I didn't read the study.

  • source
  • [–] 7 points 11 months ago* (2 children)

    I would say we should start a people’s economy that cuts out shareholders and focuses on the needs of the majority and stop honoring a corrupt system that allows billionaire South African apartheid Nazis and Israel to own our government and make us fuedal slaves. Stop honoring a social contract that doesn’t honor the people

  • source
  • hideshow 2 child comments
  • [–] 1 point 11 months ago* (1 child)

    a "people's economy that cuts out shareholders" is equivalent to the country (community) holding all the shares, IMO. such a thing is already doable by transfering the ownership of shares to the communities. no fundamentally new system of economy needed, IMO.

    just introduce taxes to make sure that transfer of ownership actually happens.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 1 point 11 months ago

    Who’s going to do that? The politicians owned by Israel and the oligarchs…never going to happen. I’m thinking more along the lines of community garden collectives, school collectives, college collectives, moving towards trade and barter, ignoring and enforcing an end to consumer debt…stuff like that

  • source
  • parent
  • [–] 5 points 11 months ago (1 child)

    Wow, AI replacing over 10% of the workforce? Talk about a game changer! I hope they're planning for the potential unemployment crisis that could come from this.

  • source
  • hideshow 1 child comment
  • [–] 5 points 11 months ago

    ceos make up 11% of the workforce? That's way higher than I thought!

  • source
  • [–] 2 points 11 months ago

    I propose we do fun stuff like replace all traffic control, cars, airplanes, boats and trains all to ai. Remembering that AI currently only really can choose the best possible word and not a word that has actual intension. If that doesn't end badly too quick, let's do all cooks and pharmacists. Then we can do all wall street investment. If all that works, then we can have several AIs argue with each other about nukes, the winner decides to launch or stay.

  • source
  • [–] 1 point 11 months ago

    Grok did a horrible job of unclogging my toilet.

  • source
  • 23 comments in c/futurology@futurology.today view that post
    [–] 34 points 11 months ago (1 child)

    I guess nobody is asking what happens IF 11.7% of the workforce is replaced. It’s like sawing the branch that you’re sitting on.

    “Sure we helped destroy the economic system, nobody has money to buy our products, but hey we saved 2 million last quarter!”

  • source
  • hideshow 1 child comment
  • [–] 1 point 11 months ago

    companies operating in democratic countries need to realize that eventually they will hit a point where the amount of workers displaced by the technology are going to be enough to negatively impact them via the legal system.

    While AI might be a helpful tool, and /could/ be cost effective in a perfect world. All that means nothing if the general public starts looking at it from a negative POV and starts voting on laws that restrict or ban it.

    If big companies were smart, they would be starting to advocate for something to placebo the general working class, such as a UBI or a supplement for people that were displaced by the tech. I don't expect they will though, and eventually it'll be a lot of money wasted developing into a tech that is likely just going to be outlawed or heavily restricted.

  • source
  • parent
  • [–] 24 points 11 months ago (1 child)

    Is that 11.7% part of middle management or C-suite? That's about the only aspect of any business that could be reliably replaced with AI

  • source
  • hideshow 1 child comment
  • [–] 36 points 11 months ago (4 children)

    I guess if 11.7% of your workforce is unreliable dumbasses then they could be replaced by AI

  • source
  • hideshow 4 child comments
  • [–] 1 point 11 months ago

    I genuinely believe that roughly 10% of most companies are ineffective employees, whether via being a dumbass or just not caring (valid or not).

    fuck, I'm seeing it happen as my own company grows from very small to medium small (a few dozen people). I swear some of my new coworkers are just AI with the way they lie to your face, you point it out, and then they immediately agree with you that yes what they said was incorrect.

  • source
  • parent
  • [–] 11 points 11 months ago

    CEOs and billionaires first.

    I'm hungry.

  • source
  • [–] 7 points 11 months ago (1 child)

    much as $1.2 trillion in wages across finance, health care and professional services.

    Would love to see an actual example of them being better than a person in these fields.

  • source
  • hideshow 1 child comment
  • [–] 5 points 11 months ago

    Oh they're great at lots of it. AI call centers? Completely unnavigable. If you wanted help, instead of customer support you should tweet at the company and hope it goes viral

    AI doctors? They can deny insurance coverages at light speed. Same day rejection is the kind of "innovation" healthcare needs

    AI investment advisors? They'll help you talk yourself into betting on whatever notion you think might be profitable, and tell you that's it's the smartest most clever investment idea they've ever heard

  • source
  • parent
  • [–] 6 points 11 months ago

    I am sure the AI companies will take care of use when they take all of our jobs.

  • source
  • [–] 4 points 11 months ago (1 child)

    Which specific 11% of the workforce are we comfortable being totally unaccountable and unthinking? 

  • source
  • hideshow 1 child comment
  • [–] 4 points 11 months ago* (5 children)

    Honestly the only thing I see current level AI reliably doing without being used as an assistive tool is grunt info work. For example a lot of T1 customer service positions can moreorless be replaced out with the current level of LLM's that we have. Many T1 support roles consist almost entirely of searching the current customers issue, copy/pasting a boilerplate solution list of what may fix it, asking "did that work?" and if not escalating to the next tier. Hallucinations at this level won't have a very big impact outside of annoying the customer and the t2 when it gets escalated because it failed to fix the issue. Said system shouldn't have control over anything, it should strictly be information based. Anything management wise or financial wise or general output of merchandise should not be using these technologies standalone, at most it should be an assistive tool to a human in that position.

  • source
  • hideshow 5 child comments
  • [–] 3 points 11 months ago (1 child)

    Idk if I am a special case or part of the majority, but when I resort to calling a support line, it's because I have exhausted the options available to me via technology, and I need a human to take responsibility. So there has literally never been a time when a purely informational AI support bot has been useful to me.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 2 points 11 months ago*

    I'm the same way here, being said, if that's the case a T1 wouldn't be useful to you anyway, as they just copy/paste the simple solutions you have likely already tried. So really no harm is done in this circumstance.

  • source
  • parent
  • [+] 1 point 11 months ago (2 children)
  • [–] 2 points 11 months ago* (1 child)

    Currently, one AI assisted worker can do the job of ⅘ of a non-AI assisted worker (numbers not made up: https://arxiv.org/abs/2507.09089 )

  • source
  • parent
  • hideshow 1 child comment
  • [–] 2 points 11 months ago*

    for rewording for simplicity because I read it wrong the first time. The linked article said that in their study an AI assisted developer took an average 20% longer to complete a project than the non-AI assisted dev.

    This is actually quite interesting to me, granted their study pool was very small(only 16 devs), but that is an interesting data point.

    Being said, this is also a different field than what I was talking about, since that moved it to development instead of T1 customer service, but the data is nice to see.

  • source
  • parent
  • load more comments (-1 replies)
    [–] 3 points 11 months ago (1 child)

    In comparison level 5 self driving would replace 30% of the workforce, and 5% of workers being unemployed is the expected normal background rate

    11.7% is hardly anything

  • source
  • hideshow 1 child comment
  • 11 comments in c/usa@midwest.social view that post
    [–] 17 points 11 months ago (3 children)

    Let me know when they replace the overpaid C-level parasites with cheaper, faster, more efficient AI CEO/CFO et cetera.

  • source
  • hideshow 3 child comments
  • [+] 25 points 11 months ago (1 child)
  • [–] 4 points 11 months ago

    This isn't even about competency really as the cost of AI is being artificially depressed through all of these businesses running at a loss. It might be insane to stick in a blender with US labor statistics. Hopefully not the kind the WH uses?

    There are a lot of different things falling under the umbrella of AI. You're right to be skeptical of an algorithm's ability to estimate another algorithm's ability in practice. They're testing how well the buzzword tech can move the needle on numbers the fed obsesses over. Really pigeonholing the tech into profitability. Best uses of this tech will be in balancing electricity flows through the country, with inputs including population + industry usage, weather patterns & other renewable energy variables (these are very much a social good but you can already hear capital hissing at the thought). Humans already operate many cybernetic systems that turn on volition, with inputs adjusted to make them exploitative. Streamlining them is going to rip the balls right off for sure. As I often quote Michael Parenti by recently, "You're stupid if you think they're stupid," it goes nicely with "the purpose of a system is what it does."

  • source
  • parent
  • [–] 8 points 11 months ago (2 children)

    I say this as someone generally bullish about AI: bullshit. I use it all the time. It's helpful when you already know what you're doing. Anything you do with AI at scale is going to have a number of fuckups, even if it's mostly reliable — and for most purposes I wouldn't even go that far.

    I see it all the time. I ask Cline to have Claude do a bunch of things and create a markdown file... and it does everything, including generating the markdown, but forgets to put it in a file and then acts confused when you say to put it in a file. If that was some financial report or contract, it could tank a whole business.

  • source
  • hideshow 2 child comments
  • [–] 1 point 11 months ago (1 child)

    Doesn’t that sound just like a real person though? It forgets shit and you have to correct it because you’re not paying it very much and it has pretty little training. You could always pay it 20 bucks a month or you could pay a small salary and insurance.

    I’m certainly not advocating for getting rid of people, but the way you worded it – not trying to start a fight – sounds just like a low level employee

  • source
  • parent
  • hideshow 1 child comment
  • [–] 3 points 11 months ago

    I don't agree with your premise. However, I don't have a good argument against at hand.

    There are intangible benefits to a person, and I feel like the number of roles where AI could perform well enough and mistakes don't cause customer satisfaction issues, regulatory compliance issues, or incur civil liability are vanishingly small. But could they be 10% of jobs? Even 5%? I don't think so.

    I could see an argument that if you have a team of 10 people, AI could let you cut one and expect the other 9 to pick up the slack. But how many teams even have ten people on them? Because I don't think a team of 5 can lose one person and still be capable of the same work. I guess it might depend on the industry — I do have IT blinders on here.

  • source
  • parent
  • [–] 4 points 11 months ago

    Breaking news 11.7% of the work force are doing useless work, work that leadership has no idea how essential it is, and some that a normalized response of text or semantic searching can be replaced with.

  • source
  • [–] 3 points 11 months ago
    [–] 1 point 11 months ago*

    I'm sure there is plenty of chaff to cut in finance and healthcare etcetera, in my heart I want to believe in a coming Burger Decimation, but US labor statistics are absolutely cooked. Having read practically nothing about the model they're using, I would appreciate if anyone can tell how accurate it is.

  • source
  • 4 comments in c/hackernews@lemmy.bestiver.se view that post
    [–] 14 points 11 months ago

    I think a lot of people are really taking AI as more functional that it currently is. I think the bubble is gonna burst in more ways than one

  • source
  • [–] 7 points 11 months ago

    Project Iceberg

    Is the economy the Titanic?

  • source
  • [–] 3 points 11 months ago (1 child)

    So the US is going to setup basic income right?

  • source
  • hideshow 1 child comment
  • load more comments (1 reply)