In the boring real world, AI will not replace humans, it’ll just make the existing humans more efficient. Efficiency is usually a good thing, although there needs to be a fight against letting all the profits go to the most powerful.
But in the actual boring real world AI doesn’t make people more efficient. Studies show it makes people less efficient and ruins their ability to think critically and complete tasks they started assigning to the AI.
At the same time the data centres powering the AI suck huge amounts of energy and water to achieve the inefficiency and degradation of the work force’s capability.
Arguably AI is creating jobs because inefficiencies in worker output now requires more workers to do the same amount of work. When I say worker output I am specifically referring to the output of office based workers, and even then this seems to be almost exclusively those in software engineering and management. There is little to no change to the output efficiency of trades at any level. Again arguably more work is available trades to help support the ever increasing data centres.
I belive they exist, it’s just the phrase “studies show that…” has become synonymus with “I’ve once saw a blogpost that…” or “that one Insta reel mentioned that…”
There are some ways in which AI (even generative AI) can make people more efficient. I’m not talking about ChatGPT writing your essays for you, or Copilot writing your code.
Like, for coding, it can be useful to get an idea of how to solve a problem. You may later throw out 90% of what the AI generated, but it can get you unstuck, or get you looking at the problem differently.
For writing tasks, it can sometimes give you a first draft that you can then put into your own words.
I would imagine that if you’re making a movie on a really tight budget and have no drawing skills, you could use GenAI to create storyboards that explain how you want something to look better than you could with just words, or better than you could with your own poor drawings. None of those AI generated storyboards will end up in the final product, but they’re potentially useful as a way for you to communicate how you want something to look.
The key difference is who is in charge. If you’re a coder and AI is one of the many tools you have available to you, you can probably find some uses for it. But, it’s going to be terrible if your job is to proof-read and approve code written and submitted by AIs. In that case you’re going to have to look at a whole bunch of code that is specifically designed to look plausibly realistic, in order to find the logic flaw camouflaged by all the other believable things.
No but seriously, does it? There have been a few studies now that show that it doesn’t really make work efficiënt. Hell, it seems to have a negative influence on our brains.
Of course we can blame the users and call it a garbage in garbage out scenario, but even with decent data it hallucinates data that doesn’t exists.
We don’t really know yet just from an empirical standpoint. I theory it seems useful in the narrow areas that it’s good at (translation/summarization and any tasks analogous to those).
We have a running joke at work that any problem it can’t solve is a skill issue with prompt engineering. At least… I hope it’s a joke.
There have been a few studies now that show that it doesn’t really make work efficiënt.
Anecdotally, I’ve been an infinite times more efficient in some tasks thanks to AI. (perhaps I’m the moron)
Hell, it seems to have a negative influence on our brains.
Sure, that could very well be. I would welcome a study that compares its negative effect to social media. My guess is that this is worse.
Also browsing through a manual is probably better for our brains that just making queries to a search engine. Just like writing on a paper is better than typing. Meditation is better than watching an episode of a series, or 100 times better than going into an endless youtube frenzy.
I’m all for making great brain health easier to catch, but I’m probably too lazy to do all of those things regurarily.
It makes non-programmers more efficient. If you reach the limits of what text editors or Excel can do natively with REGEX and have to write scripts for data formatting, AI is a godsend. Takes 10 Minutes to generate a small script, glance over it to make sure it’s ok, test it and start batch formating files rather than spend hours doing it by hand.
In the boring real world, AI will not replace humans, it’ll just make the existing humans more efficient. Efficiency is usually a good thing, although there needs to be a fight against letting all the profits go to the most powerful.
It will make all labour “unskilled” which, at least at first, will allow big business to pay everyone less.
deleted by creator
And the gymnecks…
But in the actual boring real world AI doesn’t make people more efficient. Studies show it makes people less efficient and ruins their ability to think critically and complete tasks they started assigning to the AI.
At the same time the data centres powering the AI suck huge amounts of energy and water to achieve the inefficiency and degradation of the work force’s capability.
Arguably AI is creating jobs because inefficiencies in worker output now requires more workers to do the same amount of work. When I say worker output I am specifically referring to the output of office based workers, and even then this seems to be almost exclusively those in software engineering and management. There is little to no change to the output efficiency of trades at any level. Again arguably more work is available trades to help support the ever increasing data centres.
Got references to a few of them? I assume you do, because you must’ve read them.
Here’s a recent one from MIT to get you started. Their sources are pretty good, too, as expected. https://arxiv.org/abs/2506.08872
nah sorry I didn’t note down the links to them and CBF digging them up, it’s cool if you don’t believe they exist, you do you =)
I belive they exist, it’s just the phrase “studies show that…” has become synonymus with “I’ve once saw a blogpost that…” or “that one Insta reel mentioned that…”
There are some ways in which AI (even generative AI) can make people more efficient. I’m not talking about ChatGPT writing your essays for you, or Copilot writing your code.
Like, for coding, it can be useful to get an idea of how to solve a problem. You may later throw out 90% of what the AI generated, but it can get you unstuck, or get you looking at the problem differently.
For writing tasks, it can sometimes give you a first draft that you can then put into your own words.
I would imagine that if you’re making a movie on a really tight budget and have no drawing skills, you could use GenAI to create storyboards that explain how you want something to look better than you could with just words, or better than you could with your own poor drawings. None of those AI generated storyboards will end up in the final product, but they’re potentially useful as a way for you to communicate how you want something to look.
The key difference is who is in charge. If you’re a coder and AI is one of the many tools you have available to you, you can probably find some uses for it. But, it’s going to be terrible if your job is to proof-read and approve code written and submitted by AIs. In that case you’re going to have to look at a whole bunch of code that is specifically designed to look plausibly realistic, in order to find the logic flaw camouflaged by all the other believable things.
Okay, sure. I’ll clarify.
It’ll make existing humans who are not complete morons to start with more efficient.
No but seriously, does it? There have been a few studies now that show that it doesn’t really make work efficiënt. Hell, it seems to have a negative influence on our brains.
Of course we can blame the users and call it a garbage in garbage out scenario, but even with decent data it hallucinates data that doesn’t exists.
We don’t really know yet just from an empirical standpoint. I theory it seems useful in the narrow areas that it’s good at (translation/summarization and any tasks analogous to those).
We have a running joke at work that any problem it can’t solve is a skill issue with prompt engineering. At least… I hope it’s a joke.
Anecdotally, I’ve been an infinite times more efficient in some tasks thanks to AI. (perhaps I’m the moron)
Sure, that could very well be. I would welcome a study that compares its negative effect to social media. My guess is that this is worse.
Also browsing through a manual is probably better for our brains that just making queries to a search engine. Just like writing on a paper is better than typing. Meditation is better than watching an episode of a series, or 100 times better than going into an endless youtube frenzy.
I’m all for making great brain health easier to catch, but I’m probably too lazy to do all of those things regurarily.
This must be a form of AI that hasn’t been invented yet. I’ve yet to meet a useful implementation.
As a software developer, AI has made me ten times more efficient. Before AI I used to produce 2-5 new bugs per day of work…
I suppose it could be useful if you have the sort of boss that thinks productivity is based on the number of lines of code you write.
I will never understand programmers claiming that AI makes them more efficient. As if Emmett and snippets hasn’t already existed for a decade or so
They were joking, they “went from 2-5 bugs per day to ten times that!”
It makes non-programmers more efficient. If you reach the limits of what text editors or Excel can do natively with REGEX and have to write scripts for data formatting, AI is a godsend. Takes 10 Minutes to generate a small script, glance over it to make sure it’s ok, test it and start batch formating files rather than spend hours doing it by hand.
Because regular people are out there batch formatting files all day …
Perhaps you’re just too awesome.
Yeah, apologies. I was busy imagining a you 'puter.