Artificial Intelligence / AGI and Jobs

[TherapyforGlencross]

So, here’s a topic that’s been grappling my mind lately. I’m in the my 30’s, give or take, graduated higher level schooling later in life than most. Perhaps, perhaps not. I work in the engineering industry. The amount of AI that I see on a day-to-day basis is astounding. LLM’s are now installed in most programs. Most of my work amounts to reporting, project management, and some out-of-office tasks.

It’s interesting. People rarely utilize AI. They don’t seem worried about the security of their job. I worry for them, as they certainly complete very solid work. The work we specialize in certainly requires critical analysis, but I can’t help but think that a solid foundation could be automized. Reporting, especially at the junior levels, could definitely be taken over by AI. Don’t be scared of one LM. But think of the probability of multiple LLMs forming into a single for increased computing output . That’s where we’ll be headed. Your job won’t be taken over by just one. Simple templating, data entry, initial critique of analysis, ect. What’s the point of hiring co-op students anymore, if the AI can replace their work? Middle managers will certainly be replaced. Math-based careers, gone. I really only see relationship-client managers and AI supervisors (those supervising AI outputs) as being able to withstand the increased usage of AI in consultancy.

Needless to say, it’s caused me anxiety. I’ve tried my best to implement AI in my work, and I think I’ve become more productive because of it. But eventually, let’s say, AI will eventually outdo my productiveness. I don’t really have any perceivable justification for this, but i can see AI causing massive disruptions within 5-10 years max. Does anyone else struggle with this possibility? My work gives me quite a bit of pride. I’d hate to lose that.

[Izzle]

Younger generations are absolutely ####ed. If their entry level jobs aren't being offshored, they're being transferred to AI.

10 years down the line, who's gonna audit the AI output? Not today's juniors since they havent built the skill. Scared how the future would look like.

[DownInFlames]

My 15 year old cousin wants to be a lawyer. Will we even have lawyers in 10 years?


I'm lucky that I have an established business and should be safe from AI and automation replicating what I do for the foreseeable future. I might have to invest in it to compete. That's kinda scary.

[PepsiFree]

In my experience in an industry flooded with AI already, the people who rise to the top are those who use it to their every advantage while maintaining a control over it that keeps the quality of the outputs very high and indistinguishable from something an expert (human) could produce.

The first to lose their jobs or go out of business have been the people resisting it and putting effort into downplaying it or trying to convince people about the “importance” of people over AI. A whole lot of “AI will never truly be able to replace what we do” people sadly replaced by AI real quick.

People who don’t really understand it and don’t care to aren’t far behind.

No sense worrying about it or resisting it. It’s a tool. You can get replaced by a tool I guess, or figure out how to use the tool yourself so the people that pay you don’t feel the need to replace you.

Right now the gap between people who don’t know how to do a job but are great with AI and the people who know how to do a job but don’t know anything about AI is quite small (smaller than the latter wants to admit). The gap between the former and the people who both know how to do a job and are great with AI is still massive. Might as well take advantage.

[MRCboicgy]

This is my field. All I can say is learn how to prompt really, really well for your daily work, and for the big projects that get noticed.

[indes]

I've been trying to push some AI integration into my teams workflow but my boss is very against it. I'm disappointed to be falling behind as everything on my work computer has to be IT approved there's no way for me to start leveraging these new technologies until I can get some buy in. Sucks to not be able to upskill in a meaningful way when I'm sure other people doing the same job as me are moving forward.

[bluejays]

I've seen this coming with robotics 10 years ago. About 4 years ago when ChatGPT was released and pretty quickly showed the world AI capabilities, it became alarming how good it was and good it was right away. The thought previous to that is AI can't replace creativity, but just a couple years later those were the first things to be hit hard. I have a friend in the arts who does wrap designs among other things and he says everyone just uses AI which puts him out of a job pretty fast. Or they take his work and use AI to make it better.


Anyway, I was working for one of the big banks and for ten years now they invested in automation. Now they're investing in AI with the sole purpose of replacing human capital as the consultants see that as the biggest expense on the ledger. 5-10 years is pessimistic. I think we're looking at huge layoffs in the next 2 years. Big corporations are investing heavily on this and it's a matter of time before it becomes accessible to the smaller companies too. Honestly I have no clue where society will be given taxes have to be raised and spending has to take place. There is nobody looking at the effect it's going to have on society. When you think about even a skilled job like a doctor, they're asking you questions and it's one big flow chart depending on what you answer. Even a GP can be replaced soon enough. Maybe they're protected a bit right now by their regulatory boards but they won't be forever. A lot of jobs involving skilled trades will probably be the last to automate. But the hit is coming soon for a lot of people. I really wished governments looked into this more.

[Fuzz]

If a GP gets replaced by a machine that's wrong 40% of the time, I think we are going to see a lot worse health outcomes. For them it's a tool that's probably pretty beneficial. They use it to take notes and summarize for them, which is probably a big time saver and helps them see more patients. It's about using it for what it's good at, not using it for everything.

[bluejays]

Originally Posted by Fuzz

If a GP gets replaced by a machine that's wrong 40% of the time, I think we are going to see a lot worse health outcomes. For them it's a tool that's probably pretty beneficial. They use it to take notes and summarize for them, which is probably a big time saver and helps them see more patients. It's about using it for what it's good at, not using it for everything.

This is what I hear about from a lot of people but I've been part of the discussions at the executive tables and where they want to take it in the world of maximizing profits and minimizing costs, is to replace people. It's not just the corporate world, it's even the public government. There are always discussions on how to save money. I have friends in the medical world who explain how drugs are procured and other ones are not and it's pretty shocking to hear that generic drugs don't at times have the same efficacy as the name brand ones, but are pushed by the government because it saves costs. The world is driven by money (in some industries it's more than others). But don't for a second think that every job is looking to be replaced by this stuff because it helps the bottom line.

[Cappy]

The rise of AI was certainly a consideration for the talk around Universal Basic Income a decade ago.

Then Trump gets elected, the general mood shifts towards the capital class, and we hear nothing about it moving forward.

I don't want to go all Marxist on this board (wake up sheeple style), but the hyperspeed at which companies are adopting AI at the expense of thousands of people certainly has the potential to lead to a dystopian-esque society in the future. The fact that much of our media is getting gobbled up by these elites to feed slop to us and keep us on their side is not a fantasy at this point.

Maybe the AI is all sizzle and no steak at this point, but the slow march has begun. The dotcom bubble was met with "is the internet all hype" but we all know how that turned out.

[Fuzz]

Originally Posted by bluejays

This is what I hear about from a lot of people but I've been part of the discussions at the executive tables and where they want to take it in the world of maximizing profits and minimizing costs, is to replace people. It's not just the corporate world, it's even the public government. There are always discussions on how to save money. I have friends in the medical world who explain how drugs are procured and other ones are not and it's pretty shocking to hear that generic drugs don't at times have the same efficacy as the name brand ones, but are pushed by the government because it saves costs. The world is driven by money (in some industries it's more than others). But don't for a second think that every job is looking to be replaced by this stuff because it helps the bottom line.

Well ya, that's their job. But that's how you also end up with executives making boneheaded moves that everyone else in the company knows will fail. Becuase while executives can set direction, they can't set the capabilities of AI tools.

[TherapyforGlencross]

Originally Posted by Fuzz

Well ya, that's their job. But that's how you also end up with executives making boneheaded moves that everyone else in the company knows will fail. Becuase while executives can set direction, they can't set the capabilities of AI tools.

Soo.. layoffs followed by rehirings?

I don’t see that happening, once we’re laid off there will be increased incentive to see AI prosper.

[Fuzz]

But you can only do that with hope. You can't make a tool better than it is at it's best. And if isn't doing the job, eventually you pay for that.

[bluejays]

Originally Posted by Fuzz

Well ya, that's their job. But that's how you also end up with executives making boneheaded moves that everyone else in the company knows will fail. Becuase while executives can set direction, they can't set the capabilities of AI tools.

I don't disagree. But how many times have you seen things being
"positioned" that it worked or it'll work out in time. Execs make boneheaded moves all the time but the capabilities of AI is coming damn close to being damn good. People are being replaced and that's the goal at the corporate table. If a few eggs are broken along the way so be it. Look at what Shopify's CEO said when it comes to hiring anyone in his company these days. It's happening and it's not a conspiracy. It's a tool but companies are exploiting it for their bottom line more than anything.

[TheIronMaiden]

Any time something starts to remind you of War Hammer 40k it's not going to be pretty.

AI, STCs un bridled technology, can't be good.

https://wh40k.lexicanum.com/wiki/Sta...late_Construct

[Fuzz]

Originally Posted by bluejays

I don't disagree. But how many times have you seen things being
"positioned" that it worked or it'll work out in time. Execs make boneheaded moves all the time but the capabilities of AI is coming damn close to being damn good. People are being replaced and that's the goal at the corporate table. If a few eggs are broken along the way so be it. Look at what Shopify's CEO said when it comes to hiring anyone in his company these days. It's happening and it's not a conspiracy. It's a tool but companies are exploiting it for their bottom line more than anything.

It's generally wrong 40% of the time. I guess it depends on your definition of "good enough". And before I get jumped on, I know there are some things it's much better than 60% on. And if that's your industry or work, then ya, you are in trouble. But it's not general intelligence. Try to get it to draw a map, for instance. It's laughably bad.

[AFireInside]

I do love that people who use AI say things like "learn how to prompt really quickly" as if that's going to save their job. That's not a difficult skill to learn, and if that's what you're being paid for you're going take a massive pay cut, or be replaced.

Overall AI is going to be very bad for humanity, there will be some positives for sure, but I don't see how it will be a good thing in the grand scheme of things.

[edslunch]

Originally Posted by Fuzz

It's generally wrong 40% of the time. I guess it depends on your definition of "good enough". And before I get jumped on, I know there are some things it's much better than 60% on. And if that's your industry or work, then ya, you are in trouble. But it's not general intelligence. Try to get it to draw a map, for instance. It's laughably bad.

The thing is, it doesn't need to be perfect to have a major impact. Imagine a typical scrum development team of 5-6 people including a product owner who sets the requirements and checks the results. They work for a sprint (1-3 weeks) then re-evaluate. Now replace most of them with AI that does the same work in minutes. Is it perfect, no, but the feedback and correction loop has shrunk by orders of magnitude.

No informed person is looking at AI completely replacing work any time soon, but it can already significantly reduce the size of teams working on something. 75% reduction is workforce in some industries is a big freaking deal.

[TheIronMaiden]

Originally Posted by AFireInside

I do love that people who use AI say things like "learn how to prompt really quickly" as if that's going to save their job. That's not a difficult skill to learn, and if that's what you're being paid for you're going take a massive pay cut, or be replaced.

Overall AI is going to be very bad for humanity, there will be some positives for sure, but I don't see how it will be a good thing in the grand scheme of things.

The comparable is the industrial revolution, which, by and large made human life more miserable. Though, it was great for share holder value.

[bluejays]

Originally Posted by Fuzz

It's generally wrong 40% of the time. I guess it depends on your definition of "good enough". And before I get jumped on, I know there are some things it's much better than 60% on. And if that's your industry or work, then ya, you are in trouble. But it's not general intelligence. Try to get it to draw a map, for instance. It's laughably bad.

Sure, and I'm sure you're right. But it's getting better at such a fast pace. There will be hiccups along the way but it's here.

How it starts and how it progresses is interesting. Let's use the example of a doctors office with 10 doctors.

Today - let's use it for listening to the patient/doctor conversation and let it spit out a summary for charting the conversation and next steps. Saves an hour per day per doctor (10 hours saved per day, 50 hours saved per week). With that hour saved, we don't need 10 doctors, instead we just need 9. No need to backfill when a doctor leaves.

2 years from now - let's introduce an automated triage to patients waiting in the waiting room to describe their symptoms, and have the AI give an initial diagnosis/next steps recommendation (to the doctor only). Doctor sees patient with information in hand and potential diagnosis already lined up. Saves 30 minutes more per day.

5 years from now - let's combine both the triaging and in-room conversation to the point where the doctor signs off on the AI diagnosis/next steps. Saves 4 hours a day per doctor. We only now need 4 doctors, and no student to take an initial information.

Anyway, that's just my rough take on examples how these things progress. To some it's great in the sense that you'll get your diagnosis faster and more efficiently, but for society as a whole there are so jobs not needed that the disparity climbs between the haves and have nots. That's my problem with this. A lot of people will eventually be out of jobs.