Please stick to one thread. Iām not reading three different ones.
Yes, and thatās enough functionally. Actual understanding is not necessary if you can fake it to the point that people actually think AI is cognitive. Hell, did you read that article about Richard Dawkins now thinking Claude is conscious?
No itās not LOL. Like you can pretend to be a very coginitvely smart and important person, but reality is different.
People say a lot of things. Now we get to make memes about them.
You do realizeā¦
Yes. I realize a lot of stuff. Maybe if you challenge yourself a bit more you could have figured that out yourself. I also realize books donāt have moving images like a video and that a laptop is not a blackboard. Thank you.
You were praising AI for causing people to be unemployed, essentially.
So in your line of thinking, if I say itās a good thing that the oil crisis accelerated the transition to renewable energy, am I praising Trump and the Irani regime?
If I tell you itās a good thing Europe is improving their defence, am I praising Putin or Trump?
But good thing we can get rid of many doctors and teachers eventually, like you said about us engineers. Weāll need them for actually meaningful work in the trades and the hospitality industry.
Please stick to one thread. Iām not reading three different ones.
Make fewer comments in the same thread then.
No itās not LOL.
Source on that please.
Like you can pretend to be a very coginitvely smart and important person, but reality is different
Never pretended to be either, I just said that if Iām not challenged with what pretty much amounts to puzzles of a sort, I get bored. I also said several times that this is because I have very severe ADHD not because Iām very ācoginitivelyā (sic) smart or anything.
Yes. I realize a lot of stuff. Maybe if you challenge yourself a bit more you could have figured that out yourself. I also realize books donāt have moving images like a video and that a laptop is not a blackboard. Thank you.
Well your previous comment sure sounded like you didnāt know the difference, comparing AI to a blackboard. AI is closer to a teacher than to a textbook in interactivity. It regurgitates previously learned information, sureā¦ But so do teachers. And while it canāt really reason, it gets the job done better than a lot of teachers in small towns in particular, where thereās really no competition for the jobs (rather, everyoneās competing for the teachers).
So in your line of thinking, if I say itās a good thing that the oil crisis accelerated the transition to renewable energy, am I praising Trump and the Irani regime? If I tell you itās a good thing Europe is improving their defence, am I praising Putin or Trump?
The oil crisis has an upside because using oil is actually bad for the environment. Whatās the upside of millions of people losing their jobs? That they can enter other industries and drive wages down there? Itās great for the capitalists, not for the rest of us.
I responded to this comment. You interjected yourself into that. Also, have some dignity and donāt do a āno uā. Youāre an adult for christs sake.
You kinda did by calling other peopleās work janitorial and factory work as if that was a bad thing.
Well your previous comment sure sounded like you didnāt know the difference, comparing AI to a blackboard.
That is not my lack of understanding, but yours. Both are tools for a teacher to use. How is this tool any different just because it appears to ātalk backā? A powerpoint is more advanced than a book and a video more advanced than a still image. Teachers who dealt with the inventions at their time also thought it would change the job and the world completely. It didnāt.
it gets the job done better than a lot of teachers
Yeah youāre really going to have to back that one up. Especially since the data show otherwise
The oil crisis has an upside because using oil is actually bad for the environment. Whatās the upside of millions of people losing their jobs?
Read my initial comment that you responded toā¦ Iām not going to regurgitate shit because you canāt comprehend or remember.
And Iām not going to respond to your other comment, since I already told you to use just one thread. Paste it here or I wonāt read it and it will just be a waste of your time.
And I responded to that reply in particular (i.e society adapting). Youāre now trying to merge threads that were originally about different facets of the same conversation. I have a tendency to branch a lot in my trains of thought, to the point where ideally I should be making several comments to reply to you, but I thought two might be a more manageable amount.
Sure here you are.
So neither one of us is going to prove anything. But Iām wondering what quality do you think a teacher has that an actually functional future teaching AI canāt solve to at least 80% of the same capability for 10-20% of the cost, especially if you consider that the teacher can only attend to one student at a time, or address the whole class at once, but not attend to everyoneās individual questions and such at the same time. If thereās an average of about 25 students per class, 45 minutes a lesson, thatās less than 2 minutes per student per lesson. Except most of that time usually goes towards teaching the whole class, so really most students get zero individual attention/mentorship. Which is an easy problem to solve by hiring more teachersā¦ But to get even a single new teacher in most countries with decent social systems, you first have to finance a masterās degree and then pay (in case of many European nations) union wages, which are actually pretty high compared to a lot of other jobs. And even then that person might decide not to stay in the role because teaching is very taxing work mentally, especially if you want to do it well. I know people who have worked in a school for a few years and then quit. Hell, I know a teacher who quit to work in straight up corporate tech support (and I do mean corporate: it was a B2B company, all the customers were companies), because that wasā¦ somehow less grueling. But make kids above a certain age be in classes of 50 or 70 instead of 20-30 and let AI do most of the work. Youāve suddenly eliminated close to half the teaching positions and itās easier to fill all the positions with people who are truly so passionate about it that being degraded by unruly kids every day doesnāt burn them out.
That is not my lack of understanding, but yours. Both are tools for a teacher to use. How is this tool any different just because it appears to ātalk backā?
How is a teacher any different just because it appears to ātalk backā?
In the end both the teacherās brain and the LLM powering a future teacherbot, are neural networks. And if we are to believe that we live in a deterministic universe, free will as we know it may not even be real - in which case we have more in common with AI than we think, we just have vastly more stimuli and past experiences affecting our output and are of course significantly more complex.
Yeah youāre really going to have to back that one up. Especially since the data show otherwise
Two studies from a nation already defunding its education (with individual states sabotaging their curricula too, since itās not nationally standardized), highlighting trends that were taking place before even ChatGPT came out and people actually started talking about AI. Iām talking about future agents that will wrap the LLM into something that can actually be proactive in a classroom setting, not just reactive.
And Iām not going to respond to your other comment, since I already told you to use just one thread. Paste it here or I wonāt read it and it will just be a waste of your time.
ā¦ You replied to two of my replies. I replied to your replies. The point of the thread system here is to keep things organized so we know which comment was replying to which comment. Otherwise weād just be on an oldschool bbs. But fine, Iāll paste it here.
Yes. In response to your derogatory remarks. How does that make me have a god complex?
Which remarks exactly seemed derogatory to you?
Of course being a doctor is more interesting
Then why didnāt you choose a more meaningful job instead of the interesting one? You could be doing more with patients in person. Itās the nurses and orderlies who do most of the actual work in a hospital, from what Iāve seen the few times Iāve been in one.
Could it be that you found those jobs boring? Not challenging you in the right way? Why else did you not take one of those roles, given that theyāre, by some standards, even more meaningful than that of a doctor? An orderly in particular requires much less education too, you couldāve jumped straight to helping people after high school if that was actually the primary metric by which you chose your job. I suspect you want to think that, but in reality you took the job thatās well paid and has the types of challenges that make you feel good when you solve them.
That sounds a lot like something that someone who doesnāt actually have a point would say. So, please humour me.
You: āGood, let 'em ruin themselves, we need the people in healthcare and education anyway.ā
Millions of people are losing their jobs worldwide, many more have studied for 3 years to get a degree they wonāt ever be able to use. Youāre saying thatās a good thing. Why? Because in 5-10 years time, some companies might potentially go under for laying people off today and running out of the talent pool to hire from later? Those people need jobs now, and the healthcare system isnāt going to need that many, especially since much of the western world canāt afford its current healthcare systems right now.
Somehow a bunch of people losing their jobs is good because what, 2%, of them can grab vacant jobs in your industry? Hospitals employ a lot of people and there are vacancies like any other industry, but theyāre not going to magically conjure up all the money to hire a bunch of new people just because theyād be beneficial to have around. And Iām not even going into education, because the vast majority of people working in education are teachers and teachers need a degree, usually a masterās in pedagogy. And like I said, that doesnāt even guarantee youāll have a job for long now that governments are looking into using more AI to āmake education more efficientā (cut costs).
And yet Iām the one āpraising AIā. Iām the one saying that capitalists will fight to replace every worker they can with AI, youāre the one saying itās somehow a good thing. The positives donāt even outweigh the negatives in the economic system we live in. Unless youāre part of the ownership class. I donāt know about you, but Iām not a billionaire so Iām not really benefiting here.
You start a sentence with that and expect me to read the rest of the gibberish
Pardon me, you just seemed ignorant of the whole issue of government efficiency being a thing taxpayers are usually looking for. Especially since, you know, taxes are actively being raised in some countries to be able to even afford the current spending. Or perhaps you handwaved it away because you understood itās going to be hard to argue your point unless in a perfect world without ever-increasing financial pressures.
E.g here in Estonia, in the last 5 years weāve had the income tax raised, VAT raised twice, and the price of medical visits (ER visits and first visit per case of a specialty doctor, but not GP visits) went from 5 euros to 20 euros to try to get people to go to the doctor. Oh and in the same timeframe we got a nice new vehicle tax (not a bad thing in of itself) that comes with a new registration tax that also applies retroactively to old cars if ownership is transferred (can be a couple hundred to a few thousand euros to transfer ownership of a car worth 500 for an example). AND public transit costs money for tickets again. And weāre STILL running a deficit, something Estonia didnāt really do in the past.
As a result, different government departments are always trying to save money whereever possible. That includes things like having more students per teacher so that fewer teachers need to be paid. And this is Estonia, our government debt is just under 25% of GDP. If you take for an example Belgium, a nation with some of the highest taxes in the EU, that also brings in tons of income from diplomats and MEPs spending money there), their debt is over 100% of GDP. Finland is going to be over 100% too. Thereās no set debt to GDP ratio thatās bad, but the higher the debt, the higher the interest payments. And to pay the interest, there need to be more taxes.
Doctors being replaced with AI doesnāt mean theyāre going to be all replaced at once, nor are they going to be explicitly āreplacedā by AI like software engineers. Rather, individual doctors are going to be expected to do more because āyou now have AI helping youā. Fewer young doctors will be hired because data will show that having X% fewer doctors of specialty Y per capita would result in only Z% fewer positive patient outcomes. Then in a few years, as AI tools get better, even fewer doctors will be needed. Etc. 2 years ago already we had an article here saying the national health insurance system found ways to save 21 million a year using ādigital technologies and workforce reformā. Theyāre looking to either save another 100 million a year, or raise it via more taxation. Easiest way to do the former is to lay people off, which is what āworkforce reformā really means.
If it seems to you that I ever implied that AI is going to replace doctors and teachers WITHOUT negatively affecting quality of care and educationā¦ Sorry, no, thatās not what I meant. I meant the savings are going to be so significant that governments will do it despite the reduction in quality. Itāll be deemed as āgood enoughā.
Call me a pessimist or whatever, but one thing governments and corporations have in common is that they like cutting costs at the expense of the common folk. Corporations will sell AI solutions to governments, who will have to take them because of the aging populace and therefore shrinking tax base.
> And I responded to that reply in particular (i.e society adapting)
I donāt care why it happened. Point is you tried to make this my faultā¦
> So neither one of us is going to prove anything.
This is not a āboth sidesā thing. You made a claim that it is enough to pretend to know something to teach someone else something, which is an outrageous claim. Yet you tried to put the burden of proof on me and when that didnāt stick try to share the blame. Not going to work and Iām not very fond of your bad faith arguments.
> But Iām wondering what quality do you think a teacher has that an actually functional future teaching AI canāt solve
Weāre going around in circles. You already asked this and I already told you (hint: itās the human element).
> How is a teacher any different just because it appears to ātalk backā?
Uhm. Iām pretty sure a teacher knows what itās saying and doesnāt appear to talk back but actually does.
> In the end both the teacherās brain and the LLM
That doesnāt answer my question at allā¦
> Two studies from a nation already defunding its educationā¦
Trying to pick apart the data that backs up what I said is not the same as backing up your own claim. Try a little harder.
> The point of the thread system hereā¦
Yeah in comments between different users, not the same one.
> Which remarks exactly seemed derogatory to you?
Read my previous comment. And you didnāt answer how that made me into having a god complexā¦
> Then why didnāt you choose a more meaningful job instead of the interesting one?
I am doing the most meaningful oneā¦ there is not much difference within health care. There is however a large difference with corporate jobs, like I already mentioned in my first comment.
> Somehow a bunch of people losing their jobs is good because what, 2%, of them can grab vacant jobs in your industry?
Not good for those people in the short term obviously, but better for the world in the long run. And health care is just one example. There are many other sectors that could use human capital instead of something an AI could replace.
Yes economic hardship is coming in the near future, especially when the AI bubble bursts.Ā Ā But yes we will find a way to adapt in the long run, as we always have.
So how would I fund all the health care and other jobs? Large corporations are already hardly contributing anything, and if theyāre not even providing jobs, the argument that theyāre good for the economy is completely gone. That opens up a path to simply tax the hell out of these corps. And I know that it is not going to be as easy, but I also do not need to make an entire economic balance sheet for one comment on Lemmy. That is ridiculous.
> And yet Iām the one āpraising AIā.
Indeed you are. I told you AIās limitations, that they canāt be a person, canāt reason, can only fake having knowledge, can only fake interaction and understanding, only read a lot of books (or saw a lot of scans or EKGs or lab results, I know other AIās other than LLMs exist doofus, I work with them on a daily basis) but that it doesnāt equal human experience that also incorporates a lot of stuff that you wonāt learn in any book, especially in healthcare.
You on the other hand seem to think an AI is already doing a better job than teachers without anything to back it up. So yeah, praising AI beyond its capabilities.
> you just seemed ignorant of the whole issue of government efficiency being a thing taxpayers are usually looking for
Just because I donāt over explain everything like you do doesnāt mean Iām ignorant. This is just more of your own god complex showing, your own arrogance and ignorance like I already told you a bunch of times. And you wonder why Iām annoyed with you? This is not how you have a constructive discussion. So either keep your arrogance in check or you can discuss with yourself.
You also seem to have forgotten a teeny tiny problem here: an influx of patients because of aging baby boomers. The baby boom is about to turn into a health care boom. The first baby boomers are already nearing 80 years old. They will not stop presenting themselves to hospitals or care facilities because you or society doesnāt want to spend the money on it. This definitely means we will need more people in the short to medium term. Especially the people that you consider ājanitorsā or āfactory workersā.
> Doctors being replaced with AI doesnāt meanā¦
You also donāt need to mansplain this to me. This is a daily topic of discussion at the hospital where I work.
I donāt care why it happened. Point is you tried to make this my faultā¦
Why is that even something we need to assign blame for? I replied to your reply, because it was on a different subject. It was supposed to be a different discussion than the healthcare one, until you decided to merge them. You want one discussion, I wanted two different ones. Big deal, letās move on.
This is not a āboth sidesā thing. You made a claim that it is enough to pretend to know something to teach someone else something, which is an outrageous claim. Yet you tried to put the burden of proof on me and when that didnāt stick try to share the blame. Not going to work and Iām not very fond of your bad faith arguments.
This is something that weāre only going to be finding out in the future. But like I said, I donāt expect AI to be as good as actually good teachers. I do expect it to be cheaper though. And good teachers arenāt ubiquitous. Thereās currently no data either way, but youāre saying itās impossible, Iām saying itās probably not and governmentās gonna want to save money.
Weāre going around in circles. You already asked this and I already told you (hint: itās the human element).
Seen plenty of teachers in my time in school who had very little human element in their lessons. Plus like I said, they donāt actually have time for their students.
Uhm. Iām pretty sure a teacher knows what itās saying and doesnāt appear to talk back but actually does.
And where does a teacherās knowledge come from? If the answer is books and studies, I have excellent news, LLMs have been trained on more of those than a human could ever learn.
Trying to pick apart the data that backs up what I said is not the same as backing up your own claim. Try a little harder.
What was the point of your data though? Itās irrelevant, it has nothing to do with what Iām saying.
What Iām saying is that LLMs are cheaper than humans. Iām also saying that the quality of teaching will soon be āgood enoughā. Who decides whatās good enough? Bureaucrats in the government. Itās subjective. Thereās no exact standard.
Yeah in comments between different users, not the same one.
ORRRR you can use different comment chains to reply about different subjects. Thereās no rule of one comment per user lol
Read my previous comment.
You just said you had a problem with me saying I dislike menial work. I never insulted you. I never insulted the people you work with, who Iām pretty sure work much harder than you or me. I said Iām incapable of doing that type of work because itād bore me to death.
Not good for those people in the short term obviously, but better for the world in the long run. And health care is just one example. There are many other sectors that could use human capital instead of something an AI could replace.
Yes, but unfortunately only physical jobs are now safe from AI. 5 years ago we didnāt imagine software engineers being replaced by AI anytime soon. Now we donāt imagine subpar software engineers having jobs in a few years.
And health care is just one example. There are many other sectors that could use human capital instead of something an AI could replace.
And theyāll also be under fire from AI.
Indeed you are. I told you AIās limitations, that they canāt be a person, canāt reason, can only fake having knowledge, can only fake interaction and understanding
But what matters are the results and the cost, not the actual understanding. If an AI has a similar success rate in diagnosis and treatment compared to a doctor, but is significantly cheaper and can be scaled up and down to see more or fewer patients depending on need.
AI doesnāt understand art either, yet artists are being replaced. AI doesnāt understand code either, yet software engineers arenāt being hired anymoreā¦ And we used to think weād be the last to be replaced, since weāre the ones actually creating it. That was our hubris. Youāre now where I was a few years ago with yours.
only read a lot of books (or saw a lot of scans or EKGs or lab results
And itās being trained off EVERY scan and EKG in the future. You only see those of your patients and maybe a few other study cases. AI will learn from millions. Which is the reason itāll be able to mostly perform diagnostics without even understanding what itās doing.
I know other AIās other than LLMs exist doofus, I work with them on a daily basis
Yet you implied all they have is book knowledge. You should then know there are far more accurate types of AI than LLMs, and weāre only getting better at creating them.
but that it doesnāt equal human experience that also incorporates a lot of stuff that you wonāt learn in any book, especially in healthcare.
You just admitted that you realize they donāt just learn from books. Once the deals are there (and the big AI companies are already working with governments so it wonāt be long), theyāll literally be trained on your experience as you document everything in your EMR software. So why pretend itās only ever going to be stuff learned āin any bookā?
Just because I donāt over explain everything like you do doesnāt mean Iām ignorant.
Well youāre simply handwaving important issues away, so I was hoping you were ignorant, but turns out youāre just arguing in bad faith.
You also donāt need to mansplain this to me. This is a daily topic of discussion at the hospital where I work.
A daily discussion and yet you think itāll be impossible for it to happen? Alrighty then.
The god complex I mentioned is your idea that youāre irreplaceable. Youāre not. Iām not. None of us are. You and I are easier to replace than a construction worker, plumber or electrician. Weāre not coming out of this without abolishing private property. Taxation wonāt save anything, because corporations canāt have significant revenue taxes for fairly obvious reasons and profit taxes are too easy to skirt. VAT works the best, but good luck selling the general population on increasing that, itās too visible in how it affects prices.
Thereās currently no data either way, but youāre saying itās impossible
What I said it is that the AI doesnāt understand things, doesnāt get what someone is saying. AI in its current form doesnāt understand a person. That is fact. I literally never said that it would not be possible in the future.
This is something that weāre only going to be finding out in the future.
The claim you made was that AI is better than a portion of teachers right now and that is what I denied. How is it not a big deal when it turns out youāre simply wrong? How about we move on after you agree you were wrong?
Seen plenty of teachers in my time in school who had very little human element in their lessons.
That is your biased interpretation. They were still humans and not computers.
And where does a teacherās knowledge come from?
Canāt you figure that out for yourself? I donāt really get the feeling you really like to challenge yourself mentally, as this is an easy question to answer with a little brain work.
But fine: in the beginning their formal training as a teacher, as taught by other, more experienced teacher. And after a few years on the job their own experience as well. Reducing it to just books is a straw man argument that even you know is utter bullshit.
What was the point of your data though? Itās irrelevant, it has nothing to do with what Iām saying.
It has everything to do with what you were saying, since you claimed that the AI was already doing better than the bottom part of teachers. So why are we not seeing an improvement in education results from the time they were implemented? Strange how things are irrelevant when they disprove your claim. Youāre arguing in bad faith again.
What Iām saying is that LLMs are cheaper than humans.
Now that is irrelevant because we were discussing that you though AI was doing better than teachers. Now you want to bring money into this? And I thought I couldnāt say you were praising AI? Then why do you keep doing it?
Yes, but unfortunately only physical jobs are now safe from AI
Nonsense. We have broadly discussed teaching and there are many other jobs that require physical communication with another human.
AI will learn from millions.
But it will still not understand. Which is what is necessary to make the translation from data to patients.
AI can read a gazillion scans and put out a result with a confidence % or whatever. But in the end the decision needs to be made what is best for this individual patient. It only knows books, guidelines and scans. That is not the hard part of medicine. Itās weighing all the options and information and deciding for each patient what needs to be done, taking into account a lot of factors that necessitate human interaction. This is where big data fails and the human element comes in. This is also what you fail to understand.
You also need a human to explain things to a patient. We are experiencing more and more patients every day who put their health complaints through ChatGPT and donāt understand an iota of what itās saying and draw their own conclusions that cannot be drawn. You cannot bombard a patient with data and information like ChatGPT. Youāre way too stuck in your own
but turns out youāre just arguing in bad faith
LOL nice āno uā coming from mister cognitive challenge
you think itāll be impossible for it to happen? Alrighty then.
Yes that would be pretty foolish of me to think. Good thing I never said that. More straw manning.
The god complex I mentioned is your idea that youāre irreplaceable.
Never said that, nor hold this idea. Look above why a computer can never replace a person, since as I already mentioned a 1000 times, it lacks the human aspect. This is not something that is specific to me, so I donāt know why youāre making this about me specifically.
profit taxes are too easy to skirt
And that is impossible to change?
Edit: if all youāve got is straw man arguments weāre done here.
When did I ever say AI would be better than humans in general? I said itās better than some humans and significantly cheaper than humans. You say Iām using strawmans, yet youāre not arguing the points Iām actually making, only the ones you want me to be making so you could win. Bye bye, Iām out.
Please stick to one thread. Iām not reading three different ones.
No itās not LOL. Like you can pretend to be a very coginitvely smart and important person, but reality is different. People say a lot of things. Now we get to make memes about them.
Yes. I realize a lot of stuff. Maybe if you challenge yourself a bit more you could have figured that out yourself. I also realize books donāt have moving images like a video and that a laptop is not a blackboard. Thank you.
So in your line of thinking, if I say itās a good thing that the oil crisis accelerated the transition to renewable energy, am I praising Trump and the Irani regime? If I tell you itās a good thing Europe is improving their defence, am I praising Putin or Trump?
Aww so butthurt LOL. Grow the fuck up.
Make fewer comments in the same thread then.
Source on that please.
Never pretended to be either, I just said that if Iām not challenged with what pretty much amounts to puzzles of a sort, I get bored. I also said several times that this is because I have very severe ADHD not because Iām very ācoginitivelyā (sic) smart or anything.
Well your previous comment sure sounded like you didnāt know the difference, comparing AI to a blackboard. AI is closer to a teacher than to a textbook in interactivity. It regurgitates previously learned information, sureā¦ But so do teachers. And while it canāt really reason, it gets the job done better than a lot of teachers in small towns in particular, where thereās really no competition for the jobs (rather, everyoneās competing for the teachers).
The oil crisis has an upside because using oil is actually bad for the environment. Whatās the upside of millions of people losing their jobs? That they can enter other industries and drive wages down there? Itās great for the capitalists, not for the rest of us.
I responded to this comment. You interjected yourself into that. Also, have some dignity and donāt do a āno uā. Youāre an adult for christs sake.
Sure here you are.
You kinda did by calling other peopleās work janitorial and factory work as if that was a bad thing.
That is not my lack of understanding, but yours. Both are tools for a teacher to use. How is this tool any different just because it appears to ātalk backā? A powerpoint is more advanced than a book and a video more advanced than a still image. Teachers who dealt with the inventions at their time also thought it would change the job and the world completely. It didnāt.
Yeah youāre really going to have to back that one up. Especially since the data show otherwise
Read my initial comment that you responded toā¦ Iām not going to regurgitate shit because you canāt comprehend or remember.
And Iām not going to respond to your other comment, since I already told you to use just one thread. Paste it here or I wonāt read it and it will just be a waste of your time.
And I responded to that reply in particular (i.e society adapting). Youāre now trying to merge threads that were originally about different facets of the same conversation. I have a tendency to branch a lot in my trains of thought, to the point where ideally I should be making several comments to reply to you, but I thought two might be a more manageable amount.
So neither one of us is going to prove anything. But Iām wondering what quality do you think a teacher has that an actually functional future teaching AI canāt solve to at least 80% of the same capability for 10-20% of the cost, especially if you consider that the teacher can only attend to one student at a time, or address the whole class at once, but not attend to everyoneās individual questions and such at the same time. If thereās an average of about 25 students per class, 45 minutes a lesson, thatās less than 2 minutes per student per lesson. Except most of that time usually goes towards teaching the whole class, so really most students get zero individual attention/mentorship. Which is an easy problem to solve by hiring more teachersā¦ But to get even a single new teacher in most countries with decent social systems, you first have to finance a masterās degree and then pay (in case of many European nations) union wages, which are actually pretty high compared to a lot of other jobs. And even then that person might decide not to stay in the role because teaching is very taxing work mentally, especially if you want to do it well. I know people who have worked in a school for a few years and then quit. Hell, I know a teacher who quit to work in straight up corporate tech support (and I do mean corporate: it was a B2B company, all the customers were companies), because that wasā¦ somehow less grueling. But make kids above a certain age be in classes of 50 or 70 instead of 20-30 and let AI do most of the work. Youāve suddenly eliminated close to half the teaching positions and itās easier to fill all the positions with people who are truly so passionate about it that being degraded by unruly kids every day doesnāt burn them out.
How is a teacher any different just because it appears to ātalk backā?
In the end both the teacherās brain and the LLM powering a future teacherbot, are neural networks. And if we are to believe that we live in a deterministic universe, free will as we know it may not even be real - in which case we have more in common with AI than we think, we just have vastly more stimuli and past experiences affecting our output and are of course significantly more complex.
Two studies from a nation already defunding its education (with individual states sabotaging their curricula too, since itās not nationally standardized), highlighting trends that were taking place before even ChatGPT came out and people actually started talking about AI. Iām talking about future agents that will wrap the LLM into something that can actually be proactive in a classroom setting, not just reactive.
ā¦ You replied to two of my replies. I replied to your replies. The point of the thread system here is to keep things organized so we know which comment was replying to which comment. Otherwise weād just be on an oldschool bbs. But fine, Iāll paste it here.
Which remarks exactly seemed derogatory to you?
Then why didnāt you choose a more meaningful job instead of the interesting one? You could be doing more with patients in person. Itās the nurses and orderlies who do most of the actual work in a hospital, from what Iāve seen the few times Iāve been in one.
Could it be that you found those jobs boring? Not challenging you in the right way? Why else did you not take one of those roles, given that theyāre, by some standards, even more meaningful than that of a doctor? An orderly in particular requires much less education too, you couldāve jumped straight to helping people after high school if that was actually the primary metric by which you chose your job. I suspect you want to think that, but in reality you took the job thatās well paid and has the types of challenges that make you feel good when you solve them.
You: āGood, let 'em ruin themselves, we need the people in healthcare and education anyway.ā
Millions of people are losing their jobs worldwide, many more have studied for 3 years to get a degree they wonāt ever be able to use. Youāre saying thatās a good thing. Why? Because in 5-10 years time, some companies might potentially go under for laying people off today and running out of the talent pool to hire from later? Those people need jobs now, and the healthcare system isnāt going to need that many, especially since much of the western world canāt afford its current healthcare systems right now.
Somehow a bunch of people losing their jobs is good because what, 2%, of them can grab vacant jobs in your industry? Hospitals employ a lot of people and there are vacancies like any other industry, but theyāre not going to magically conjure up all the money to hire a bunch of new people just because theyād be beneficial to have around. And Iām not even going into education, because the vast majority of people working in education are teachers and teachers need a degree, usually a masterās in pedagogy. And like I said, that doesnāt even guarantee youāll have a job for long now that governments are looking into using more AI to āmake education more efficientā (cut costs).
And yet Iām the one āpraising AIā. Iām the one saying that capitalists will fight to replace every worker they can with AI, youāre the one saying itās somehow a good thing. The positives donāt even outweigh the negatives in the economic system we live in. Unless youāre part of the ownership class. I donāt know about you, but Iām not a billionaire so Iām not really benefiting here.
Pardon me, you just seemed ignorant of the whole issue of government efficiency being a thing taxpayers are usually looking for. Especially since, you know, taxes are actively being raised in some countries to be able to even afford the current spending. Or perhaps you handwaved it away because you understood itās going to be hard to argue your point unless in a perfect world without ever-increasing financial pressures.
E.g here in Estonia, in the last 5 years weāve had the income tax raised, VAT raised twice, and the price of medical visits (ER visits and first visit per case of a specialty doctor, but not GP visits) went from 5 euros to 20 euros to try to get people to go to the doctor. Oh and in the same timeframe we got a nice new vehicle tax (not a bad thing in of itself) that comes with a new registration tax that also applies retroactively to old cars if ownership is transferred (can be a couple hundred to a few thousand euros to transfer ownership of a car worth 500 for an example). AND public transit costs money for tickets again. And weāre STILL running a deficit, something Estonia didnāt really do in the past.
As a result, different government departments are always trying to save money whereever possible. That includes things like having more students per teacher so that fewer teachers need to be paid. And this is Estonia, our government debt is just under 25% of GDP. If you take for an example Belgium, a nation with some of the highest taxes in the EU, that also brings in tons of income from diplomats and MEPs spending money there), their debt is over 100% of GDP. Finland is going to be over 100% too. Thereās no set debt to GDP ratio thatās bad, but the higher the debt, the higher the interest payments. And to pay the interest, there need to be more taxes.
Doctors being replaced with AI doesnāt mean theyāre going to be all replaced at once, nor are they going to be explicitly āreplacedā by AI like software engineers. Rather, individual doctors are going to be expected to do more because āyou now have AI helping youā. Fewer young doctors will be hired because data will show that having X% fewer doctors of specialty Y per capita would result in only Z% fewer positive patient outcomes. Then in a few years, as AI tools get better, even fewer doctors will be needed. Etc. 2 years ago already we had an article here saying the national health insurance system found ways to save 21 million a year using ādigital technologies and workforce reformā. Theyāre looking to either save another 100 million a year, or raise it via more taxation. Easiest way to do the former is to lay people off, which is what āworkforce reformā really means.
If it seems to you that I ever implied that AI is going to replace doctors and teachers WITHOUT negatively affecting quality of care and educationā¦ Sorry, no, thatās not what I meant. I meant the savings are going to be so significant that governments will do it despite the reduction in quality. Itāll be deemed as āgood enoughā.
Call me a pessimist or whatever, but one thing governments and corporations have in common is that they like cutting costs at the expense of the common folk. Corporations will sell AI solutions to governments, who will have to take them because of the aging populace and therefore shrinking tax base.
> And I responded to that reply in particular (i.e society adapting)
I donāt care why it happened. Point is you tried to make this my faultā¦
> So neither one of us is going to prove anything.
This is not a āboth sidesā thing. You made a claim that it is enough to pretend to know something to teach someone else something, which is an outrageous claim. Yet you tried to put the burden of proof on me and when that didnāt stick try to share the blame. Not going to work and Iām not very fond of your bad faith arguments.
> But Iām wondering what quality do you think a teacher has that an actually functional future teaching AI canāt solve
Weāre going around in circles. You already asked this and I already told you (hint: itās the human element).
> How is a teacher any different just because it appears to ātalk backā?
Uhm. Iām pretty sure a teacher knows what itās saying and doesnāt appear to talk back but actually does.
> In the end both the teacherās brain and the LLM
That doesnāt answer my question at allā¦
> Two studies from a nation already defunding its educationā¦
Trying to pick apart the data that backs up what I said is not the same as backing up your own claim. Try a little harder.
> The point of the thread system hereā¦
Yeah in comments between different users, not the same one.
> Which remarks exactly seemed derogatory to you?
Read my previous comment. And you didnāt answer how that made me into having a god complexā¦
> Then why didnāt you choose a more meaningful job instead of the interesting one?
I am doing the most meaningful oneā¦ there is not much difference within health care. There is however a large difference with corporate jobs, like I already mentioned in my first comment.
> Somehow a bunch of people losing their jobs is good because what, 2%, of them can grab vacant jobs in your industry?
Not good for those people in the short term obviously, but better for the world in the long run. And health care is just one example. There are many other sectors that could use human capital instead of something an AI could replace.
Yes economic hardship is coming in the near future, especially when the AI bubble bursts.Ā Ā But yes we will find a way to adapt in the long run, as we always have.
So how would I fund all the health care and other jobs? Large corporations are already hardly contributing anything, and if theyāre not even providing jobs, the argument that theyāre good for the economy is completely gone. That opens up a path to simply tax the hell out of these corps. And I know that it is not going to be as easy, but I also do not need to make an entire economic balance sheet for one comment on Lemmy. That is ridiculous.
> And yet Iām the one āpraising AIā.
Indeed you are. I told you AIās limitations, that they canāt be a person, canāt reason, can only fake having knowledge, can only fake interaction and understanding, only read a lot of books (or saw a lot of scans or EKGs or lab results, I know other AIās other than LLMs exist doofus, I work with them on a daily basis) but that it doesnāt equal human experience that also incorporates a lot of stuff that you wonāt learn in any book, especially in healthcare.
You on the other hand seem to think an AI is already doing a better job than teachers without anything to back it up. So yeah, praising AI beyond its capabilities.
> you just seemed ignorant of the whole issue of government efficiency being a thing taxpayers are usually looking for
Just because I donāt over explain everything like you do doesnāt mean Iām ignorant. This is just more of your own god complex showing, your own arrogance and ignorance like I already told you a bunch of times. And you wonder why Iām annoyed with you? This is not how you have a constructive discussion. So either keep your arrogance in check or you can discuss with yourself.
You also seem to have forgotten a teeny tiny problem here: an influx of patients because of aging baby boomers. The baby boom is about to turn into a health care boom. The first baby boomers are already nearing 80 years old. They will not stop presenting themselves to hospitals or care facilities because you or society doesnāt want to spend the money on it. This definitely means we will need more people in the short to medium term. Especially the people that you consider ājanitorsā or āfactory workersā.
> Doctors being replaced with AI doesnāt meanā¦
You also donāt need to mansplain this to me. This is a daily topic of discussion at the hospital where I work.
Why is that even something we need to assign blame for? I replied to your reply, because it was on a different subject. It was supposed to be a different discussion than the healthcare one, until you decided to merge them. You want one discussion, I wanted two different ones. Big deal, letās move on.
This is something that weāre only going to be finding out in the future. But like I said, I donāt expect AI to be as good as actually good teachers. I do expect it to be cheaper though. And good teachers arenāt ubiquitous. Thereās currently no data either way, but youāre saying itās impossible, Iām saying itās probably not and governmentās gonna want to save money.
Seen plenty of teachers in my time in school who had very little human element in their lessons. Plus like I said, they donāt actually have time for their students.
And where does a teacherās knowledge come from? If the answer is books and studies, I have excellent news, LLMs have been trained on more of those than a human could ever learn.
What was the point of your data though? Itās irrelevant, it has nothing to do with what Iām saying.
What Iām saying is that LLMs are cheaper than humans. Iām also saying that the quality of teaching will soon be āgood enoughā. Who decides whatās good enough? Bureaucrats in the government. Itās subjective. Thereās no exact standard.
ORRRR you can use different comment chains to reply about different subjects. Thereās no rule of one comment per user lol
You just said you had a problem with me saying I dislike menial work. I never insulted you. I never insulted the people you work with, who Iām pretty sure work much harder than you or me. I said Iām incapable of doing that type of work because itād bore me to death.
Yes, but unfortunately only physical jobs are now safe from AI. 5 years ago we didnāt imagine software engineers being replaced by AI anytime soon. Now we donāt imagine subpar software engineers having jobs in a few years.
And theyāll also be under fire from AI.
But what matters are the results and the cost, not the actual understanding. If an AI has a similar success rate in diagnosis and treatment compared to a doctor, but is significantly cheaper and can be scaled up and down to see more or fewer patients depending on need.
AI doesnāt understand art either, yet artists are being replaced. AI doesnāt understand code either, yet software engineers arenāt being hired anymoreā¦ And we used to think weād be the last to be replaced, since weāre the ones actually creating it. That was our hubris. Youāre now where I was a few years ago with yours.
And itās being trained off EVERY scan and EKG in the future. You only see those of your patients and maybe a few other study cases. AI will learn from millions. Which is the reason itāll be able to mostly perform diagnostics without even understanding what itās doing.
Yet you implied all they have is book knowledge. You should then know there are far more accurate types of AI than LLMs, and weāre only getting better at creating them.
You just admitted that you realize they donāt just learn from books. Once the deals are there (and the big AI companies are already working with governments so it wonāt be long), theyāll literally be trained on your experience as you document everything in your EMR software. So why pretend itās only ever going to be stuff learned āin any bookā?
Well youāre simply handwaving important issues away, so I was hoping you were ignorant, but turns out youāre just arguing in bad faith.
A daily discussion and yet you think itāll be impossible for it to happen? Alrighty then.
The god complex I mentioned is your idea that youāre irreplaceable. Youāre not. Iām not. None of us are. You and I are easier to replace than a construction worker, plumber or electrician. Weāre not coming out of this without abolishing private property. Taxation wonāt save anything, because corporations canāt have significant revenue taxes for fairly obvious reasons and profit taxes are too easy to skirt. VAT works the best, but good luck selling the general population on increasing that, itās too visible in how it affects prices.
What I said it is that the AI doesnāt understand things, doesnāt get what someone is saying. AI in its current form doesnāt understand a person. That is fact. I literally never said that it would not be possible in the future.
The claim you made was that AI is better than a portion of teachers right now and that is what I denied. How is it not a big deal when it turns out youāre simply wrong? How about we move on after you agree you were wrong?
That is your biased interpretation. They were still humans and not computers.
Canāt you figure that out for yourself? I donāt really get the feeling you really like to challenge yourself mentally, as this is an easy question to answer with a little brain work.
But fine: in the beginning their formal training as a teacher, as taught by other, more experienced teacher. And after a few years on the job their own experience as well. Reducing it to just books is a straw man argument that even you know is utter bullshit.
It has everything to do with what you were saying, since you claimed that the AI was already doing better than the bottom part of teachers. So why are we not seeing an improvement in education results from the time they were implemented? Strange how things are irrelevant when they disprove your claim. Youāre arguing in bad faith again.
Now that is irrelevant because we were discussing that you though AI was doing better than teachers. Now you want to bring money into this? And I thought I couldnāt say you were praising AI? Then why do you keep doing it?
Nonsense. We have broadly discussed teaching and there are many other jobs that require physical communication with another human.
But it will still not understand. Which is what is necessary to make the translation from data to patients.
AI can read a gazillion scans and put out a result with a confidence % or whatever. But in the end the decision needs to be made what is best for this individual patient. It only knows books, guidelines and scans. That is not the hard part of medicine. Itās weighing all the options and information and deciding for each patient what needs to be done, taking into account a lot of factors that necessitate human interaction. This is where big data fails and the human element comes in. This is also what you fail to understand.
You also need a human to explain things to a patient. We are experiencing more and more patients every day who put their health complaints through ChatGPT and donāt understand an iota of what itās saying and draw their own conclusions that cannot be drawn. You cannot bombard a patient with data and information like ChatGPT. Youāre way too stuck in your own
LOL nice āno uā coming from mister cognitive challenge
Yes that would be pretty foolish of me to think. Good thing I never said that. More straw manning.
Never said that, nor hold this idea. Look above why a computer can never replace a person, since as I already mentioned a 1000 times, it lacks the human aspect. This is not something that is specific to me, so I donāt know why youāre making this about me specifically.
And that is impossible to change?
Edit: if all youāve got is straw man arguments weāre done here.
When did I ever say AI would be better than humans in general? I said itās better than some humans and significantly cheaper than humans. You say Iām using strawmans, yet youāre not arguing the points Iām actually making, only the ones you want me to be making so you could win. Bye bye, Iām out.
When did I say you said humans in general?