Will AI replace most tech jobs?

[Alchemisst]

AI is already pretty impressive and seems to be getting better all the time. I have seen some pretty amazing things it can do e.g be used to program apps and websites and other automation with just prompts. I have seen some people in these fields that were previously said that AI will not replace them now changing their minds. I think AI overall will be a benefit to automate all the mundane jobs/ tasks like driving etc.

How much of a threat is AI to tech (and all jobs in general) and is there anything you are doing to adapt/ overcome these challenges?

[moneytaichi]

I work at a matured start-up. AI just replaced the entire content marketing team a couple of months ago. I'm using AI to research and publish the marketing content with a jaw-dropping speed. The first draft done by AI is better than my 2nd or 3rd draft in the old time. I think AI is going to replace LOTS of jobs in many fields really quickly. Even if it's not fully matured in certain fields right now, AI can learn and improve. That's both scary (from job POV) and exciting (from productivity POV). I also wonder if young people would ever learn how to write without AI's aid.

[twinstudy]

I don't think AI will replace jobs other than menial or mid level jobs.

I asked Chat GPT a pretty basic prompt in my area of work - I asked it to assess whether a particular case in my field of law was likely to be appealed and if so on what grounds of appeal, and it had no idea whatsoever. So if it can't even figure out the likelihood of an appeal or the grounds of appeal, it's not going to be able to draft appeal submissions, and that means trial lawyers' jobs are safe.

Granted, tech isn't the same field, but I assume that well-paid software engineers are doing more than just adapting known code in a mechanical way. There's got to be some insight or creativity in there - otherwise what are you being paid for?

Chat GPT is not capable of generating fresh analysis - it can only synthesise from known principles, and it does so poorly, with numerous subtle (sometimes obvious) inaccuracies. It's very good at providing a basic overview of a known topic with 80-90% accuracy - you just won't know what the 10-20% inaccurate bits are.

[Fru-Gal]

I don’t like Chat GPT because of Sam Altman/the company. Specialized LLMs like for law, medicine, etc. are what will revolutionize work, not Chat GPT. Of course OpenAI is working hard to sell into those specialty fields (such as defense) as well. But there will be many options to choose from. Training your own private/personal LLM is a good idea for staying ahead of the curve.

[Metalcat]

AI won't replace my profession, but it can and will dramatically alter it, the same way it will with all healthcare professions.

It's already taking over note writing, which has a massive impact on the distribution of clinician time and resources, but the note-writing apps are also data-miners. ALL of that patient data is being processed for these companies to develop even more advanced tools.

In therapy and other disciplines, these apps are listening to appointments and analyzing what the diagnostic process looks like IRL, what the interventions are, and what the outcomes are. This massive amount of client data is definitely going to be used to develop therapy-capable chatbots. It's inevitable.

This will absolutely decimate the population of basic skill, generic anxiety&depression therapists.

[Ron Scott]

There are obviously things AI excels at and things it can’t do. I think 2025 will be a year in which tech firms focus strongly on offering solutions that leverage the real strengths of AI.

It is already outperforming doctors in various diagnoses and interpreting medical imaging/radiology. As above, it can write and enhance marketing copy, and so on. It is starting to get good at self-driving, customer service, accounting, and drug development.

The large cloud providers (AWS, Google, MS) are heavily invested in AI and will be rolling out compelling solutions for their customers—which just happen to be the largest companies in the world.

For perspective:

1. AI doesn’t have to replace everyone to have the effect. If it can help a team of 20 accountants do the work of 30 the die is cast and it will continue to improve.

2. Without AI the introduction of COVID vaccines would have taken much longer…and scientists say these vaccines saved tens of millions of lives worldwide. That ain’t no ChatGPT hallucination…

[scottish]

AI is often wrong.   The sectors it's going to change the most are the ones based on soft skills, where there is no right answer or being strictly correct is less important.    It's also not creative, so it won't be coming up with new solutions to problems.

In it's current form, it's not going to have a big impact on workers with technical jobs.      Medical imaging is an interesting case.   Hopefully the AI results are being overseen by trained medical staff.

[Baguettestache]

Could you say a little more about how it has replaced a marketing team? I understand it can write articles yes but do they really convert leads to clients better than a marketing team?
ChatGPT seems to generate texts that are politically correct, maybe not the best selling texts. Maybe I'm just using it wrong.

*So far* as a developer it doesn't seem able to replace me: it's great to kickstart a new project but unable to manage an old, big project.
I've read a guy who tested Bolt.new and when he would ask it for a new feature it would break a previously working one.

I know it will get better though.

[ProxyRetired]

It will replace the lower, if not mid-level.

Help desk employees? Gone in three years, tops. I don't need someone to help reset passwords, report a system outage, or order a new mouse.

Programmers? First level programmers gone; can have seniors edit and tune code churned out by AI.

Security?  Co-Pilot is working hard to replace first level security responders. It can automate disabling computers faster during a ransomware event than any human ever could.

Are there yet? No, but five years from now, tech is going to be very different.

[Ron Scott]

There is a tendency for people to equate AI with LLMs the or generative AIs of today. This is like comparing the adult Einstein to a newborn. Christ, we just started.

I think we’ll be moving more toward world models that in addition to training on text, become adept at other correlations in the real world. They’ll be trained on decisions made with the IoT, sensors, cameras, navigation of the physical environment (walking and task-driven robots, self-driving cars—that develop an working understanding of how to engage successfully with the laws of physics. etc.), reactions to it’s activities by humans and other animals, and the overall process of scientific experimentation.

In the future AI’s “training data” will only be limited by what the world contains and humans know. Freed from scouring the internet for training data and capable of exploring the world around it with better sensory perception than us, equipped with many of the analytic skills humans developed over the millennia, real progress beyond quaint ChatGPTs will begin. The initial set of these AI/robots will learn like a baby but with amazing speed, and subsequent generations will be “born” with all that knowledge and learn from there.

I expect humanoids on steroids that can get their energy from the sun or some download, that can travel almost anywhere (with few of the physical limitations we have), and accomplish more than we expect from our best and brightest.

Maybe not in the next decade tho…

[ProxyRetired]

I expect humanoids on steroids that can get their energy from the sun or some download, that can travel almost anywhere (with few of the physical limitations we have), and accomplish more than we expect from our best and brightest.

Maybe not in the next decade tho…

We already have robots in agriculture which are solar powered, and spend their days roaming the field, weeding plants with small lasers. 24/7, with minimal downtime. It's already begun.

[obstinate]

Doubt it, but I won't care because I won't have a tech job much longer.

[Dave1442397]

At the moment it's a mostly-useful programming aid, but the code it generates for me is usually just a starting point. I do find it handy for help with algorithms and syntax, and for generating comments. I'm learning (more) Java right now, and I find it handy for checking code or telling me what a piece of code does.

I have three years to go before we have FU money and I don't really care anymore, so hopefully AI doesn't replace me immediately :)

[moneytaichi]

Could you say a little more about how it has replaced a marketing team? I understand it can write articles yes but do they really convert leads to clients better than a marketing team?
ChatGPT seems to generate texts that are politically correct, maybe not the best selling texts. Maybe I'm just using it wrong.

I use a licensed version of ChatGPT from my company so I assume it's more advanced than free version. I just asked ChatGPT a question: "which roles in a marketing team can AI replace most easily? ".

It gave me the following response in less than 2 seconds, which I agree. Also, the remaining team members are expected to do any gaps that laid-off colleagues left, with aids of AI.

AI can easily replace or assist in marketing roles that involve repetitive or data-driven tasks, such as:

1. **Content Production**: Generating written content and simple graphic designs.
2. **Digital Marketing**: Automating SEO, analytics, and campaign management.
3. **Social Media**: Scheduling posts and analyzing engagement metrics.
4. **Email Marketing**: Automating campaigns, segmentation, and tracking.
5. **Ad Campaigns**: Optimizing bids and targeting.

**Harder to Replace Roles** include strategic and creative tasks like leadership, content strategy, customer relationship management, and product marketing, which require human judgment, creativity, and interpersonal skills.

[lthenderson]

I asked ChatGPT and it said, "Artificial intelligence (AI) is expected to significantly impact many technology jobs, but it is unlikely to completely replace most of them. Instead, AI will reshape the tech job landscape, automating certain tasks while also creating new roles and responsibilities."

I agree completely. As a student of history, you can look back throughout our history and see many revolutionary technologies that have altered our job landscape significantly (factories/assembly lines, electricity, computers, robots, etc.). A.I. is just one of a long line of them. Like all the others in the past, we will adapt and keep on plugging away as humans. People will lose jobs. New jobs will be created.

[YttriumNitrate]

I asked Chat GPT a pretty basic prompt in my area of work - I asked it to assess whether a particular case in my field of law was likely to be appealed and if so on what grounds of appeal, and it had no idea whatsoever. So if it can't even figure out the likelihood of an appeal or the grounds of appeal, it's not going to be able to draft appeal submissions, and that means trial lawyers' jobs are safe.

While I don't believe AI will entirely eliminate lawyers, it may reduce their numbers. As an example, if it previously took two associates and a partner to draft the appeal brief, with AI it may only take one associate and a partner.

[Paul der Krake]

Humans will never run out of desire for new products or experiences, so we'll never run out of of new tech jobs to invent for ourselves.

[Baguettestache]

I use a licensed version of ChatGPT from my company so I assume it's more advanced than free version. I just asked ChatGPT a question: "which roles in a marketing team can AI replace most easily? ".

It gave me the following response in less than 2 seconds, which I agree. Also, the remaining team members are expected to do any gaps that laid-off colleagues left, with aids of AI.

AI can easily replace or assist in marketing roles that involve repetitive or data-driven tasks, such as:

1. **Content Production**: Generating written content and simple graphic designs.
2. **Digital Marketing**: Automating SEO, analytics, and campaign management.
3. **Social Media**: Scheduling posts and analyzing engagement metrics.
4. **Email Marketing**: Automating campaigns, segmentation, and tracking.
5. **Ad Campaigns**: Optimizing bids and targeting.

**Harder to Replace Roles** include strategic and creative tasks like leadership, content strategy, customer relationship management, and product marketing, which require human judgment, creativity, and interpersonal skills.

Again will the AI generated content outperform the human one?
LinkedIn seemed full of "AI will write my posts for me" posts when ChatGPT came out but it doesn't seem to produce any creative though that a specialist in his domain can produce. AI generated content doesn't seem so interesting so far.

Yes I understand that it's only the beginning and it will get better but so far we haven't heard of companies laying off their whole marketing team.

"Automating SEO, analytics, and campaign management"
Really? ChatGPT can't edit your website to enhance the SEO, it can't buy paid links either, can it? You still need a human to start the campaign, follow and analyze it... Sure he can get ideas from ChatGPT but ChatGPT still plays the role of the copilot not the pilot.

Maybe we'll need less people for the same amount of work but we still need those people *for now*

[Baguettestache]

I have three years to go before we have FU money and I don't really care anymore, so hopefully AI doesn't replace me immediately :)

If we experience massive layouts won't the stock market be massively impacted? Will you still have FU money?
Who's going to buy stuff when no one has a job anymore?

[obstinate]

If we experience massive layouts won't the stock market be massively impacted? Will you still have FU money?
Who's going to buy stuff when no one has a job anymore?

They'll still have jobs (they'll have to), it just won't be jobs that AI can do.

Layoffs don't necessarily cause stocks to go down. Often they cause stocks to go up.

[Ron Scott]

Not coding per se, but I couldn’t resist this one in which the judges of a major international photography contest were not only fooled by AI-generated art, but gave it “the creative open category” award last year.

https://www.bbc.com/news/entertainment-arts-65296763

[Wolfpack Mustachian]

I have three years to go before we have FU money and I don't really care anymore, so hopefully AI doesn't replace me immediately :)

If we experience massive layouts won't the stock market be massively impacted? Will you still have FU money?
Who's going to buy stuff when no one has a job anymore?

If that happens, some sort of UBI will be a must. I'm more thinking that society will just invent new jobs, though, needed or not.

[obstinate]

If that happens, some sort of UBI will be a must. I'm more thinking that society will just invent new jobs, though, needed or not.

There's still a significant shortage of people to do work with their hands. (General contractors, plumbers, welders, painters, all kinds of service work.) There ain't gonna be UBI for desk-workers as long as the folks who work out in the real world still have to turn up for business.

[Wolfpack Mustachian]

If that happens, some sort of UBI will be a must. I'm more thinking that society will just invent new jobs, though, needed or not.

There's still a significant shortage of people to do work with their hands. (General contractors, plumbers, welders, painters, all kinds of service work.) There ain't gonna be UBI for desk-workers as long as the folks who work out in the real world still have to turn up for business.

Exactly. There have always been other jobs even if they're radically different from what people are used too. If for some reason it changes, sure UBI with the very basics covered could happen. I'm not thinking that is likely, though.

[Ron Scott]

I have three years to go before we have FU money and I don't really care anymore, so hopefully AI doesn't replace me immediately :)

If we experience massive layouts won't the stock market be massively impacted? Will you still have FU money?
Who's going to buy stuff when no one has a job anymore?

If that happens, some sort of UBI will be a must. I'm more thinking that society will just invent new jobs, though, needed or not.

It’ll be tricky.

We probably wont be able to create enough “unnecessary jobs” in the private sector, but I suppose there’s always government work.

But if UBI is the proposal, the key will be getting through the transition from where we are now to actually paying UBI (a disaster just there) and managing a world in which billions of people don’t have day jobs. My guess is this is unworkable.

[twinstudy]

I have three years to go before we have FU money and I don't really care anymore, so hopefully AI doesn't replace me immediately :)

If we experience massive layouts won't the stock market be massively impacted? Will you still have FU money?
Who's going to buy stuff when no one has a job anymore?

If that happens, some sort of UBI will be a must. I'm more thinking that society will just invent new jobs, though, needed or not.

There will always be new jobs. People will learn to adapt. Think of all the changes that globalisation, offshoring and automation have wrought - yet people still have jobs, and it's not like the unemployment rate has steadily increased despite all these factors. I think what AI will do is tip the balance of power in the job market even further towards those with particular skillsets. But it won't lead to mass unemployment.

[Ron Scott]

l
I think what AI will do is tip the balance of power in the job market even further towards those with particular skillsets. But it won't lead to mass unemployment.

You could be right. But what uniquely human skillsets do you think will be required to keep the masses employed?

[Wolfpack Mustachian]

I have three years to go before we have FU money and I don't really care anymore, so hopefully AI doesn't replace me immediately :)

If we experience massive layouts won't the stock market be massively impacted? Will you still have FU money?
Who's going to buy stuff when no one has a job anymore?

If that happens, some sort of UBI will be a must. I'm more thinking that society will just invent new jobs, though, needed or not.

There will always be new jobs. People will learn to adapt. Think of all the changes that globalisation, offshoring and automation have wrought - yet people still have jobs, and it's not like the unemployment rate has steadily increased despite all these factors. I think what AI will do is tip the balance of power in the job market even further towards those with particular skillsets. But it won't lead to mass unemployment.

I agree with this mostly. I do think it could change the nature of employment to further move towards the part time hustle style of employment.

[twinstudy]

l
I think what AI will do is tip the balance of power in the job market even further towards those with particular skillsets. But it won't lead to mass unemployment.

You could be right. But what uniquely human skillsets do you think will be required to keep the masses employed?

Some laborious jobs will always be needed. I mean, put bluntly, someone has to service the machines. And there are probably other jobs that are more cheaply done by a human than by AI. Remember we only have to pay humans $15/hour.

At some stage, maybe in 30-40 years, we will have true AI (general intelligence) and maybe it will find a way to finally supplant human labour, or the need for humanity entirely.

[Ron Scott]

l
I think what AI will do is tip the balance of power in the job market even further towards those with particular skillsets. But it won't lead to mass unemployment.

You could be right. But what uniquely human skillsets do you think will be required to keep the masses employed?

Some laborious jobs will always be needed. I mean, put bluntly, someone has to service the machines. And there are probably other jobs that are more cheaply done by a human than by AI. Remember we only have to pay humans $15/hour.

At some stage, maybe in 30-40 years, we will have true AI (general intelligence) and maybe it will find a way to finally supplant human labour, or the need for humanity entirely.

That seems reasonable. And plumbers, carpenters, etc.

I’m worried about handling the jobs displaced before we have workarounds. For example, as above, what happens when half the accountants, lawyers, and so forth, can do the jobs of the other half with an AI assistant?  There’s that gap and they can’t all build houses.

[twinstudy]

l
I think what AI will do is tip the balance of power in the job market even further towards those with particular skillsets. But it won't lead to mass unemployment.

You could be right. But what uniquely human skillsets do you think will be required to keep the masses employed?

Some laborious jobs will always be needed. I mean, put bluntly, someone has to service the machines. And there are probably other jobs that are more cheaply done by a human than by AI. Remember we only have to pay humans $15/hour.

At some stage, maybe in 30-40 years, we will have true AI (general intelligence) and maybe it will find a way to finally supplant human labour, or the need for humanity entirely.

That seems reasonable. And plumbers, carpenters, etc.

I’m worried about handling the jobs displaced before we have workarounds. For example, as above, what happens when half the accountants, lawyers, and so forth, can do the jobs of the other half with an AI assistant?  There’s that gap and they can’t all build houses.

If 2 accountants/lawyers can do the job of 3, I expect that instead of reducing head count by 1/3, the firm will just increase productivity by 50% and charge clients accordingly. There's always work to be done, and most professions have no real ceiling to what can be delivered to clients. To give you an example, the advent of email (much more efficient than faxes/letters) and internet databases (vastly better than going to a law library) hasn't meant that lawyers charge less in fees. It hasn't even meant that lawyers spend less time working up a case. Rather, it's just led to more complex discovery disputes and other interlocutory steps, as each side has more weapons at its disposal.

[Ron Scott]

l
I think what AI will do is tip the balance of power in the job market even further towards those with particular skillsets. But it won't lead to mass unemployment.

You could be right. But what uniquely human skillsets do you think will be required to keep the masses employed?

Some laborious jobs will always be needed. I mean, put bluntly, someone has to service the machines. And there are probably other jobs that are more cheaply done by a human than by AI. Remember we only have to pay humans $15/hour.

At some stage, maybe in 30-40 years, we will have true AI (general intelligence) and maybe it will find a way to finally supplant human labour, or the need for humanity entirely.

That seems reasonable. And plumbers, carpenters, etc.

I’m worried about handling the jobs displaced before we have workarounds. For example, as above, what happens when half the accountants, lawyers, and so forth, can do the jobs of the other half with an AI assistant?  There’s that gap and they can’t all build houses.

If 2 accountants/lawyers can do the job of 3, I expect that instead of reducing head count by 1/3, the firm will just increase productivity by 50% and charge clients accordingly.s disposal.

This is a complex topic and I think I need to clarify my point.

What I intended to comment on is the fact that AI will specifically increase labor productivity. Similar to the initial phases of industrialization in which the population was fed with a dramatically smaller percentage of us working on farms, we will now see (for example) all our legal and accounting needs met with far fewer lawyers and accountants. In a competitive market, the cost savings will be reflected in pricing.

During industrialization, lives were certainly disrupted, but eventually jobs were created for the majority those who wanted to work. We produced more and consumed more—an abundance of plenty.

We do not know how the AI revolution will play out or if the next phase of dramatic productivity growth will come with massive job creation. I personally am not confident that there will be enough increased demand (consumption) for additional goods and services to generate full employment. I could certainly be wrong if developing and poor countries pick up the slack and supercharge consumption. But is logically ends at some point.

[twinstudy]

What I intended to comment on is the fact that AI will specifically increase labor productivity. Similar to the initial phases of industrialization in which the population was fed with a dramatically smaller percentage of us working on farms, we will now see (for example) all our legal and accounting needs met with far fewer lawyers and accountants. In a competitive market, the cost savings will be reflected in pricing.

That's what I'm saying though. There won't be any cost savings.

In law, which is adversarial, greater productivity just means greater weapons on both sides. It's a stalemate arms race.

In accounting, we already have ample tools (resources online + web forms) for any smart person to do her own taxes, in most cases. Yet small accountants still prosper. And as AI becomes more utilised, I suspect it'll just increase the complexity of accountancy practices without necessarily bringing the price down.

[Metalcat]

What I intended to comment on is the fact that AI will specifically increase labor productivity. Similar to the initial phases of industrialization in which the population was fed with a dramatically smaller percentage of us working on farms, we will now see (for example) all our legal and accounting needs met with far fewer lawyers and accountants. In a competitive market, the cost savings will be reflected in pricing.

That's what I'm saying though. There won't be any cost savings.

In law, which is adversarial, greater productivity just means greater weapons on both sides. It's a stalemate arms race.

In accounting, we already have ample tools (resources online + web forms) for any smart person to do her own taxes, in most cases. Yet small accountants still prosper. And as AI becomes more utilised, I suspect it'll just increase the complexity of accountancy practices without necessarily bringing the price down.

I fully agree with this that it could even raise the cost of accounting and legal services.

In my previous healthcare profession, we had an explosion of technology over the past 30 years, particularly software to save time, and as you said, it just increased costs for the patients in the end because it increased complexity.

Increased efficiency in technical work doesn't necessarily translate to lower costs.

[Fru-Gal]

Wow, I find these insights from @Metalcat and @twinstudy to be really exciting and reassuring. It reminds me of the desktop publishing revolution, and the personal computer revolution (I guess I have these in backwards order.) My parents had personal computers at home, and then years later when desktop publishing started to get so big they embraced that as well and then years after that self-publishing started to be a thing and they got on board with that!

Once again, I will mention ephemeralism in the eternal hope that someone on this form will go and run with it. Lol. Buckminster Fuller concept.

While there are definitely negative outcomes that will happen with AI, we can’t also separate out the evil marketing genius of that asshole Sam Altman. In order to gain his regulatory moat, he thought it was a great idea to tell us that his product was really great, but also could accidentally result in human extinction.

[roomtempmayo]

If that happens, some sort of UBI will be a must. I'm more thinking that society will just invent new jobs, though, needed or not.

There's still a significant shortage of people to do work with their hands. (General contractors, plumbers, welders, painters, all kinds of service work.) There ain't gonna be UBI for desk-workers as long as the folks who work out in the real world still have to turn up for business.

This brings to mind an interview that David Brooks did with Steve Bannon that ran in the Times a few months ago.

One thing Bannon said about his right flank of the MAGA base (and I trust him to know it better than anyone) is that they're rabidly against AI.  Even the guys who work with their hands tend to realize that, broadly speaking, their wives are exposed, and their wives' desk jobs are often providing the family's health insurance.  That's obviously a whole lot of stereotyping, but that doesn't mean Bannon isn't on to something.

That doesn't really speak directly to the UBI question, but I suspect there'll be more pushback from the MAGA right than might immediately be apparent.  For every tradesman in that coalition, there are probably ten people answering phones.  This won't shake out as neatly along party lines as it might appear on first glance.

[Metalcat]

Wow, I find these insights from @Metalcat and @twinstudy to be really exciting and reassuring. It reminds me of the desktop publishing revolution, and the personal computer revolution (I guess I have these in backwards order.) My parents had personal computers at home, and then years later when desktop publishing started to get so big they embraced that as well and then years after that self-publishing started to be a thing and they got on board with that!

Once again, I will mention ephemeralism in the eternal hope that someone on this form will go and run with it. Lol. Buckminster Fuller concept.

While there are definitely negative outcomes that will happen with AI, we can’t also separate out the evil marketing genius of that asshole Sam Altman. In order to gain his regulatory moat, he thought it was a great idea to tell us that his product was really great, but also could accidentally result in human extinction.

It's good and bad, like everything. But you can look at industries that have undergone massive technological advancements or over saturation/drop in demand to see how the workforce within them responds in different fields.

The changes may in the end result in higher performance and higher fees, but it depends on the industry and the process to get there can be brutal with many casualties along the way.

This is my second time entering a profession that's in the early waves of oversaturation thanks to some policy changes and tech changes that rapidly altered the landscape. I've had the privilege of coming into each with advanced skills from a previous profession that leapfrog me well ahead of my entry-level colleagues who got pummelled by a precipitous drop in demand for their entry-level skills. I'm currently watching former classmates struggle to get any traction in a field that had they graduated into it a decade ago would have been a immediate high 5 to low 6 figure income for just competent level work.

Professions adapt, but that doesn't mean that individuals do. A lot of people hurt when the tides in a profession change. And that's where the fear comes in, because being in the wrong career phase at the wrong time of a tide change can be devastating.

Now, being in the right phase at the right time can be spectacularly beneficial, so there's that flip side too.

[roomtempmayo]

Two names I didn’t need to read in a sentence, David Brooks and Steve Bannon

It's actually a pretty interesting interview.  Looks like it was published last summer (paywall): https://www.nytimes.com/2024/07/01/opinion/steve-bannon-trump.html

[YttriumNitrate]

That's what I'm saying though. There won't be any cost savings. In law, which is adversarial, greater productivity just means greater weapons on both sides. It's a stalemate arms race.

At least among big law firms, only about a quarter of lawyers' time is spent on litigation.[1] I would expect most of the time savings from AI to come from this like drafting contracts, wills/estate planning, and government compliance.

[Metalcat]

That's what I'm saying though. There won't be any cost savings. In law, which is adversarial, greater productivity just means greater weapons on both sides. It's a stalemate arms race.

At least among big law firms, only about a quarter of lawyers' time is spent on litigation.[1] I would expect most of the time savings from AI to come from this like drafting contracts, wills/estate planning, and government compliance.

Yes, this is like DH's job. He's in policy development and there are a few lawyers in his team and what they do is much like law in that it's processing enormous amounts of data and reports and creating more condensed reports and recommendations.

So much of what DH does could be done by AI. It's taken 2 decades of this kind of work for him to even be able to do the kind of work he does, to even know where to look for the kind of data he needs.

At least 50% of his job could be AI automated, which would be very, very interesting.

[Baguettestache]

Professions adapt, but that doesn't mean that individuals do. A lot of people hurt when the tides in a profession change. And that's where the fear comes in, because being in the wrong career phase at the wrong time of a tide change can be devastating.

That's my main fear. I'm a software developer, if AI replaces software developer will I really be able to adapt to a new job?
How much lower will my entry salary be in that new job compared to my senior software developer salary?

People who say "there will be new jobs" scare me. 😁

[Metalcat]

Professions adapt, but that doesn't mean that individuals do. A lot of people hurt when the tides in a profession change. And that's where the fear comes in, because being in the wrong career phase at the wrong time of a tide change can be devastating.

That's my main fear. I'm a software developer, if AI replaces software developer will I really be able to adapt to a new job?
How much lower will my entry salary be in that new job compared to my senior software developer salary?

People who say "there will be new jobs" scare me. 😁

Yep, when major changes happen, there is always collateral damage. There's never a guarantee that you will be able to predictably continue your job at your current income.

Industries can change overnight, personal circumstances can change overnight. That's why we focus so much on FI, to have the freedom and flexibility to adapt to changing circumstances.

My career was destroyed overnight by changing circumstances. I trained for 11 years for my previous profession for it to literally disappear on me over the course of a single year because something changed.

I had to completely retrain for an entirely new career. I just started in my new industry, which is actively going through massive upheaval, so I'm actively responding to those changes.

If you want security, the best thing you can do is be maximally adaptable.

I personally grew up with parents who pivoted careers multiple times due to pretty extreme changing circumstances, so I never had a view of a job being something you just got and could expect to keep indefinitely until it was convenient for you to leave.

I'm currently in a tiny fishing village in Newfoundland where the generational cod fishery was shut down suddenly in the 90s decimating hundreds of years of way of life here with astronomical collateral damage for the population.

These things happen all the time. Whole industries go sideways, local economies collapse, personal circumstances make a chosen career impossible. Plans change.

The more nimble and adaptable you are, the more you can pivot when change happens, and the less you will be caught up in the collateral damage.

I started taking courses on AI a few years ago to better understand how to be responsive to it as it comes in. I now have a much better understanding of what it's utility could be in my profession, and from day 1 I've built a brand that is robust to that new reality. I've invested thousands in advanced training and worked extra side jobs to build expertise that would be compatible with the advent of very effective AI tech.

I'm not scared because I put in a lot of energy to understand it and to prepare to work within an ecosystem that includes it. If you see a change coming, start adapting early. Be proactive, not just reactive.

[kenner]

Professions adapt, but that doesn't mean that individuals do. A lot of people hurt when the tides in a profession change. And that's where the fear comes in, because being in the wrong career phase at the wrong time of a tide change can be devastating.

That's my main fear. I'm a software developer, if AI replaces software developer will I really be able to adapt to a new job?
How much lower will my entry salary be in that new job compared to my senior software developer salary?

People who say "there will be new jobs" scare me. 😁

Speaking from hardware design perspective I'm not sure the mass replacement is eminent just yet, but I am seeing people changing how they do things.  As of now AI spits out generally functional code that does generally what's asked for, but the person doing the asking has to be pretty damn specific/detailed unless the code involved is pretty simple.  So we're starting to see people who are getting very good at writing what amount to spec variants and then dealing with the iterations involved when the thing you've written isn't being interpreted by the way you might expect.  I fully expect AI to keep improving, in particular with regards to the current input and processing limitations which are a major factor in what we deal with, and I wouldn't be surprised to see humans fully removed from the actual coding process in the fairly near future, but I think knowing how to present a design to an AI is going to be an in-demand skillset.  On the entry level side I can see it being pretty brutal, but most design and verification people I know spend a lot more time on the what rather than the how anyway.

[YttriumNitrate]

Professions adapt, but that doesn't mean that individuals do. A lot of people hurt when the tides in a profession change. And that's where the fear comes in, because being in the wrong career phase at the wrong time of a tide change can be devastating.

That's my main fear. I'm a software developer, if AI replaces software developer will I really be able to adapt to a new job? How much lower will my entry salary be in that new job compared to my senior software developer salary? People who say "there will be new jobs" scare me. 😁

One advantage to any AI upheaval in the job market compared to many previous upheavals is that the job losses will likely be spread out geographically.

In the past, when a major factory became automated, or was off-shored, not only did a worker lose their job but their neighbors did too causing the whole area to become economically depressed. When a software developer loses their job to AI, they'll probably have the good fortune of being in an area that is still economically vibrant.

[Metalcat]

Professions adapt, but that doesn't mean that individuals do. A lot of people hurt when the tides in a profession change. And that's where the fear comes in, because being in the wrong career phase at the wrong time of a tide change can be devastating.

That's my main fear. I'm a software developer, if AI replaces software developer will I really be able to adapt to a new job? How much lower will my entry salary be in that new job compared to my senior software developer salary? People who say "there will be new jobs" scare me. 😁

One advantage to any AI upheaval in the job market compared to many previous upheavals is that the job losses will likely be spread out geographically.

In the past, when a major factory became automated, or was off-shored, not only did a worker lose their job but their neighbors did too causing the whole area to become economically depressed. When a software developer loses their job to AI, they'll probably have the good fortune of being in an area that is still economically vibrant.

Yeah, as I mentioned above, living in an area that was geographically devastated by the cod fishery moratorium, it's brutal how universal the local devastation is to all industries.

[Must_ache]

How about quantum computing?

Google scientists have created a new quantum processor that, in five minutes, cracked a problem that would have taken the world's best supercomputer 10 septillion years to solve.

[obstinate]

How about quantum computing?

Google scientists have created a new quantum processor that, in five minutes, cracked a problem that would have taken the world's best supercomputer 10 septillion years to solve.

Been hearing about how quantum computing is going to change the world immanently for my entire career (20 years). Maybe this time it will actually happen, but my pattern-matching monkey brain is not panicked yet.

[GuitarStv]

I expect humanoids on steroids that can get their energy from the sun or some download, that can travel almost anywhere (with few of the physical limitations we have), and accomplish more than we expect from our best and brightest.

Maybe not in the next decade tho…

We already have robots in agriculture which are solar powered, and spend their days roaming the field, weeding plants with small lasers. 24/7, with minimal downtime. It's already begun.

Good to know we're already giving them weapons and instructions to kill biological life.  :P

[joemandadman189]

I think many if not all jobs will be impacted in the next 5-10 years. AI abilities are growing exponentially, right now AI is said to have an IQ of 115 across all topics. What happens when its 150 or 200 across all topics?  Robots will continue to improve at an astonishing rate, augmenting a number of menial/labor intensive jobs. Like others have said not all jobs will disappear but far fewer workers will be required to achieve the same or greater output.

In the end, i hypothesize that college will still be necessary as we will need phd level experts in nearly every field to review and fact check AI outputs, we may need double or triple the current number and the need for most other degree levels would be drastically reduced.

Google's Willow quantum chip will only accelerate these outcomes.

Imagine if AI is able to displace 1% of the work force per year over the next 10 years. The current USA workforce is ~169 million, 1% a year for 10 years would equate to ~154 million, a loss of 15 million jobs. The current population growth rate in the USA is 0.5%, if we take half that number 0.25% and apply it to the current USA work force of ~169 million, in 10 years we hypothetically could have ~172 million workers. in this scenario we have 172-154= ~18 million people of working age that dont have jobs. A 2% displacement from AI would result in ~31 million jobs not available, considering population growth. A 10% displacement rate per year for 10 years would result in a future work force of ~65.5 million, a net job loss of ~106.5 million. What would a UBI look like in this scenario?

I work at a matured start-up. AI just replaced the entire content marketing team a couple of months ago. I'm using AI to research and publish the marketing content with a jaw-dropping speed. The first draft done by AI is better than my 2nd or 3rd draft in the old time. I think AI is going to replace LOTS of jobs in many fields really quickly. Even if it's not fully matured in certain fields right now, AI can learn and improve. That's both scary (from job POV) and exciting (from productivity POV). I also wonder if young people would ever learn how to write with AI's aid.

This first comment stuck out to me, i think most folks have no idea how big the tsunami barreling towards us. One positive may be that the time frame needed to achieve large scale project could be greatly reduced, furthering human progress.

[GuitarStv]

Maybe AI will one day even be powerful enough to solve the environmental problems that encorporating AI into everything has caused.