Artificial Intelligence - Impact on Jobs

[BGARK]

I'd like to open up a discussion about AI and its current and future impact on our jobs and careers. I'm both fascinated and concerned by the rapid advancements in this field, and I'm curious to hear others' perspectives.

Some questions to consider:

1. How do you feel about AI's current capabilities? What's your experience with AI tools in your workplace?

2. What do you think your job or industry will look like in 2025 due to AI advancements?

3. For those in college or considering further education: How are you approaching your studies and career planning in light of AI's potential to automate many white-collar jobs?

4. Parents: How are you discussing future career paths with your children? What advice would you give them?

5. Are there any jobs or industries you believe will remain "AI-proof" in the near future?

In my workplace, we're already using AI daily for various tasks, from customer service to legal document preparation and even complex engineering problems. While the efficiency gains are impressive, it's also making me question the long-term security of many traditional career paths.

I believe it's crucial to have open discussions about this topic, even if some of the potential outcomes seem daunting.

[Sheets Tabuer]

Work in IT, I'm so glad I'm early 50s as I think AI will replace me in a few years, for what I do anyway. Hopefully I'll be able to retire by then.

Really feel sorry for those coming behind us because they'll need to be some kind of universal income or else the only jobs left will be bum wipers.

Can't even say you'll drive a van as they'll all be AI too.

[Raptor7000r]

BGARK said:

I'd like to open up a discussion about AI and its current and future impact on our jobs and careers. I'm both fascinated and concerned by the rapid advancements in this field, and I'm curious to hear others' perspectives.

Some questions to consider:

1. How do you feel about AI's current capabilities? What's your experience with AI tools in your workplace?

2. What do you think your job or industry will look like in 2025 due to AI advancements?

3. For those in college or considering further education: How are you approaching your studies and career planning in light of AI's potential to automate many white-collar jobs?

4. Parents: How are you discussing future career paths with your children? What advice would you give them?

5. Are there any jobs or industries you believe will remain "AI-proof" in the near future?

In my workplace, we're already using AI daily for various tasks, from customer service to legal document preparation and even complex engineering problems. While the efficiency gains are impressive, it's also making me question the long-term security of many traditional career paths.

I believe it's crucial to have open discussions about this topic, even if some of the potential outcomes seem daunting.

I work in IT, no use of it on our specific sector as of yet but many products specifically development are starting to use the buzzwords, no real products seen in our area as of yet.

AI capabilities are rapidly developing but still trying to harness this into real life use cases may take a while longer.

I think for us in the analysis sector we'll still be required as personnel but there will be more automation, how much depends on how capable and technically proficient AI becomes on troubleshooting and the actual human element of implementation.

I think many where humans are still required to press the buttons will remain AI proof but the tools and technologies will harness AI for machine learning, analysis and suggested solutions but will require humans to push the final button.

[gangzoom]

BGARK said:

..........4. Parents: How are you discussing future career paths with your children? What advice would you give them?
...................

In my workplace, we're already using AI daily for various tasks, from customer service to legal document preparation and even complex engineering problems. While the efficiency gains are impressive, it's also making me question the long-term security of many traditional career paths.

The 'industry' I'm in is still on pen/paper for huge amount of work, so digital even without Ai is going to massively introduce efficiency.

For Ai itself, as far as I understand it's not 'intelligent' not even close, but it does do an amazing job of pattern recognition and replicating the past.

Original ideas will therefore stand out even more, if everyone is using Ai to 'solve' a problem, if you want to 'better' everyone else you will need to be even more original/innovative. So those who can come up with truly novel ideas will thrive.

[Olivera]

BGARK said:

In my workplace, we're already using AI daily for various tasks, from customer service to legal document preparation and even complex engineering problems.

I hope you *really* know what you're doing and have a very thorough verification process, as current AI gets many things flat out wrong.

[Rough101]

So far, the only thing we can get AI to do is unpick a mess that other AI made.

[BGARK]

Olivera said:

I hope you *really* know what you're doing and have a very thorough verification process, as current AI gets many things flat out wrong.

Our step one is to provide a draft idea or set of questions that a human can verify. We do this anyway by sometimes asking a colleague to check a "draft" of something.

Also, are we are talking about the latest version GPT 01-Preview, not Chat GPT, as they are vastly different beasts, even Claude 3.5 Sonnet is better than GPT for a lot of text based use cases.

If people are using the free versions, they are all pretty poor at giving any type of coherent response to most questions or challenges.

[MustangGT]

Olivera said:

BGARK said:

In my workplace, we're already using AI daily for various tasks, from customer service to legal document preparation and even complex engineering problems.

I hope you *really* know what you're doing and have a very thorough verification process, as current AI gets many things flat out wrong.

Absolutely this, I have yet to see a single legal-type document produced by AI that was anywhere near correct.

[BGARK]

gangzoom said:

For Ai itself, as far as I understand it's not 'intelligent' not even close,

[fat80b]

It's a fascinating time. My own work usage is:

• We use an AI meeting note taker which is pretty good.

• Whilst I have used the LLMs a bit, I've mostly given up on them for now for a) privacy and b) accuracy concerns, but I do occasionally use them for putting content outlines together.

• I have used them to help write code and the efficiency gain of using them this way is incredible. Want to knock out some python functions to do some simple task, then it is so much faster than writing it yourself or googling for the answer.

But:

The really interesting question for me is what does the future look like for all of the "knowledge workers" of today - I fear that the advent of AI is going to change the meaning of work in fundamental ways that we cannot currently comprehend. And this is because for the first time ever, it is going to be the highest paid jobs that are getting replaced by the technology advancement and that has never happened before.

i.e. Every other technical revolution that has come before (agriculture, industrial, internet etc), has resulted in lower paid jobs being replaced with higher paid jobs, the individuals and the countries they live in get richer, everyone is better off

i.e. the pattern of the past is that subsistence farming gets replaced with industrial farming, coal miners get replaced with software engineers etc.


But in this case, and for the first time ever, it is the highest paid roles that are most at risk, - e.g. lawyers, accountants, software engineers, train drivers, artists etc. Basically, anything that requires a brain / knowledge, and/or talent is at risk of being replaced and the total number of roles in each of those disciplines is going to dramatically decline (to near zero)

All that will be left will be a tiny number of AI input roles, and lower paid manual jobs for the rest of us.

Yes it's a bit of worse case scenario, and yes, people also said similar things as the Internet arrived, but if I was about to go to University today, I have no idea what the "correct" degree is to do any more if you want to choose a discipline that is going to last 40 years!

If the worse case does happen, then it leads to interesting questions about who should "own" the AI, as it could be said that the owner of the AI is going to own all of the money and therefore all of the people - should it be the big tech giants holding governments to ransom, or should it be governments owning the AI for the benefits of the people.

We may well end up with a big reckoning which essentially boils down to extreme capitalism vs extreme socialism when it comes to who benefits from the work the AI does...

Interesting times.

[gangzoom]

BGARK said:

I'm not sure you can really 'test' intelligence, if you're main definition is generating that all important 'idea'.

[ARHarh]

Surely AI as it is right now is no more than a snapshot of the internet, from the past, and a clever bit of programming that can write out the answer in a legible way rather than thinking for its self as such.

The engineering problems it is solving will just be stuff people have put on the web somewhere, about how they solved that problem. There is no way I would trust it, and as far as I can see it is not going to be able to invent anything, or even suggest a new way of doing things.

I see it as just like the flying cars they told me about in the 60's. Strangely 60 years later we still don't have flying cars for everyone.

So glad I am retired and don't need to worry about someone thinking they could replace me with "Google on steroids"

When I left school they told me robots would take every job so I decide to enter a career in electro-mechanical research so I could be the person making the robots. If I was starting out now I would be getting a career in programming so at least I stood a chance of being the one who programs AI

[gangzoom]

fat80b said:

But in this case, and for the first time ever, it is the highest paid roles that are most at risk.....artists

I find this to the most interesting bit of the generative Ai stuff, if an algorithm that has no real concept of death, life, meaning of the world, can generate 'pictures' that looks very similar to our 'dreams' what does 'art' mean? Or is it a good demonstration of what 'dreams' are, ie our brains trying to package our own experiences into manageable chunks and dreams are a simple 'side effect' of moving data around??

Ai development may actually showing us how 'dumb' as a species we are, and how indifferent we might be to bees, where its the collective input of many individuals that allows human civilisation to appear so 'advanced' when in reality we are pretty primitive.

[BGARK]

gangzoom said:

BGARK said:

I'm not sure you can really 'test' intelligence, if you're main definition is generating that all important 'idea'.

We send our children to school, to absorb knowledge, and then do a test.

What other metrics can we use?

Lots of people are testing AI, using human methods, its outperforming us on most knowledge based methods, and last week was a significant shift change with 01 being launched.

I work in a role that needs new ideas to keep us in business. if we can dismiss the poor ideas quickly that really helps.

[nuyorican]

I'm a manual tradesman so I think I'm safe. My worry is being able to drive in the future, or more specifically, park anywhere near the client's house. How do these people living in apartment blocks in London get anything done?

[BGARK]

MustangGT said:

Absolutely this, I have yet to see a single legal-type document produced by AI that was anywhere near correct.

Can we do a test, then you critique the answer?

[98elise]

Sheets Tabuer said:

Work in IT, I'm so glad I'm early 50s as I think AI will replace me in a few years, for what I do anyway. Hopefully I'll be able to retire by then.

Really feel sorry for those coming behind us because they'll need to be some kind of universal income or else the only jobs left will be bum wipers.

Can't even say you'll drive a van as they'll all be AI too.

Agreed.

First thing I got AI to do was write some specific code. It did a good job, including pointers as to what I would need to change to make it work in my specific environment.

That's years of training and experience rendered redundant.

Glad I'm retired.

[nuyorican]

I was watching a YouTube video the other day when about halfway through i realised it was AI. The narrator suddenly started referring to a person as 'Mr 'christian name'. In this case, 'Mr Andy'.

It was weird, like a glitch in the matrix. I had to pause the video and look in the comments where it was confirmed. I know there are loads of AI YouTube videos but until now they've generally been quite obvious. But this was by a real person content creator who seems to have 'gone AI'.

Very strange.

So yeah, content creators.

[Olivera]

BGARK said:

Our step one is to provide a draft idea or set of questions that a human can verify. We do this anyway by sometimes asking a colleague to check a "draft" of something.

Asking a human to verify a lengthy legal document or complex engineering specification that an AI has written is very much non-trivial, with a difficulty that goes up exponentially to the size of the task. Hence it is very easy to miss mistakes.

Also, given an AI is polymath with a huge range of knowledge, how is a human going to verify where it goes beyond their area of expertise? In fact how is a human going to verify anything at all a few generations of AI down the line, if they have been de-skilled to just AI prompters?

[BGARK]

fat80b said:

It's a fascinating time. My own work usage is:

• We use an AI meeting note taker which is pretty good.

• Whilst I have used the LLMs a bit, I've mostly given up on them for now for a) privacy and b) accuracy concerns, but I do occasionally use them for putting content outlines together.

• I have used them to help write code and the efficiency gain of using them this way is incredible. Want to knock out some python functions to do some simple task, then it is so much faster than writing it yourself or googling for the answer.

But:

The really interesting question for me is what does the future look like for all of the "knowledge workers" of today - I fear that the advent of AI is going to change the meaning of work in fundamental ways that we cannot currently comprehend. And this is because for the first time ever, it is going to be the highest paid jobs that are getting replaced by the technology advancement and that has never happened before.

i.e. Every other technical revolution that has come before (agriculture, industrial, internet etc), has resulted in lower paid jobs being replaced with higher paid jobs, the individuals and the countries they live in get richer, everyone is better off

i.e. the pattern of the past is that subsistence farming gets replaced with industrial farming, coal miners get replaced with software engineers etc.


But in this case, and for the first time ever, it is the highest paid roles that are most at risk, - e.g. lawyers, accountants, software engineers, train drivers, artists etc. Basically, anything that requires a brain / knowledge, and/or talent is at risk of being replaced and the total number of roles in each of those disciplines is going to dramatically decline (to near zero)

All that will be left will be a tiny number of AI input roles, and lower paid manual jobs for the rest of us.

Yes it's a bit of worse case scenario, and yes, people also said similar things as the Internet arrived, but if I was about to go to University today, I have no idea what the "correct" degree is to do any more if you want to choose a discipline that is going to last 40 years!

If the worse case does happen, then it leads to interesting questions about who should "own" the AI, as it could be said that the owner of the AI is going to own all of the money and therefore all of the people - should it be the big tech giants holding governments to ransom, or should it be governments owning the AI for the benefits of the people.

We may well end up with a big reckoning which essentially boils down to extreme capitalism vs extreme socialism when it comes to who benefits from the work the AI does...

Interesting times.

I agree with all of that, what concerns me is the amount of people who joke or dismiss this subject, usually the "intellectuals" who believe they cannot be replaced.

What happened only last week with 01-Preview, most people wont even bother to research or try to understand.

Spend $20 and test it for the next month, then come back and comment.