How many of you have lost work because of AI?

Now that we have some distance from the initial AI boom, can you say whether AI took your job or improved it?

I definitely felt the impact in the first half of 2024.

I was working as a copywriter when a client told me I should either produce more articles for the same amount of money using AI, or lower my price per article.

As you can probably guess, I lost that client.

Paradoxically, I knew about AI roughly 6 months before it became widely known in my home country (I discovered AI through Product Hunt, btw).

And just two weeks ago, a programmer friend, who has been coding for 15 years, told me that he used to have 5 different income streams.

Since AI became mainstream, he has lost 4 of them.

Personally, I think that in this era, creating your own product with the help of AI can be an important safety net.

Have you experienced AI taking away a type of work you used to do (and did it make you less employable)?

181 views

Add a comment

Replies

Best

Honestly, I haven’t lost work to AI but I’ve definitely seen parts of dev work get automated faster than I expected The bigger shift for me is how much more important it is to actually understand what the code is doing

 People underestimate skills :)

 Exactly having a skill and knowing how to use it are two different things Building your own product can be a safety net but it also comes with a whole new set of skills to learn beyond just development

Your client asking for more articles at the same price is the tell. What got cheaper is the drafting, and drafting was never the expensive part. Deciding what a page should claim still needs someone who watched people misread the last version, and no tool has sat in that room.

 Agreed, and the version of this I see is that the work which vanished wasn't anyone's job, it was a line item. We used to buy stock footage and basic voiceover and neither is a purchase now. Nobody got fired, a supplier just stopped receiving an order. On Nika's safety net point I'd push back a bit, since building your own product now means competing with everyone else who can also build one in a weekend, and distribution is the part that didn't get cheaper. Mine launched a fortnight ago and did 3 points on day one.

 I am afraid that it is a matter of time before AI will be able to do everything, even good fact-checking.

  Maybe. What I'd watch is that checking a claim is the easy half, and knowing which claim will cost you is the half that lives in whoever remembers the last time one went wrong.

the part about losing a client despite learning about ai early is a tough one. knowing the tool exists doesn’t always mean you get to benefit from it.
do you think building your own product is becoming a real safety net, or just another thing people have to learn to survive?

 Actually, building one's own product can work only for a while. Building with AI becomes easier, and more online products will be available; at the same time, we will become more self-sufficient. So we will not be so motivated to purchase from others.

I haven’t. And I don’t know anyone who has.

 That's a good case, thankfully :)

Hmm I haven't lose my work because of AI yet ..
But my close friend who has been coding for a long time and has fixed clients started to worry about losing his job because of AI. Since Claude code is really convenient to do coding without professional coding knowledge, I can feel him.

So does this is the reason nowadays I'm keep thinking about what work AI cannot replace.

 + I can see more and more people on Reddit posting discussions where they are asking what other work they can do. It is such a paradox that the most desired and well-paid positions are becoming "useless" in terms of easy AI replacement.

 yeah yeah I know right :((

There's a lot of argument about whether AI takes your work. My own view is that this is a paradigm shift, and you can't go back up the river now.

A while ago there was news about Hollywood actors striking over films being made from AI copies of their own footage. Watching it, what I thought was: what a waste.

Say you'd have been in ten shows. Before, you got paid for ten. Now you shoot one and the other nine get made with AI. The fee drops, obviously. But I'd guess it doesn't drop to a tenth. More like 30 or 40 percent. So your time is a tenth of what it was, and 20 or 30 percent of the money arrived without you working for it. And all the time you'd have spent shooting the other nine is now free. Spend it doing the same thing again and you're in ten shows anyway.

That's arithmetic on paper and the real numbers won't line up that neatly. But I'm doing a version of it in web development. I use AI for everything, so even when my rate per job halves, the work I put in is something like a tenth of what it was. My hourly rate went up and so did the total.

It mostly comes down to whether you can take the shift as an opportunity instead of something to hold off.

 The thing is... will be people okay to pay you more when they have:
a) other competitors who offer the service/product cheaper
b) the option to try it for cheaper on their own with the help of AI?

 (a) is a real fight, yes. There will always be someone cheaper, so the thing that matters is shipping a better service before they do. When a new tool shows up and everyone gets more done, the price per unit falls. You can't stand against that, and I don't try to. Someone who could only write code can now cover design too, and the client gets more from the same person.

(b) is smaller than it looks, I think, because the client has to learn it first. We're used to IT so we assume it's easy. For someone who isn't, the bar is high. Even when AI makes the thing for them, deciding whether the output is any good takes knowledge and experience, and that is the part they don't have.

I've been doing system development for 22 years on that basis, with no outbound sales, and the only contracts that ended were ones where the client's own business stopped.

Ah, I think I was lucky. My current contract started right when the Claude boom was getting prepped. It only helped boost my role and make me more useful. None of the devs in the company have had any problems because of AI. Nobody has complained about wanting more productivity now that everyone has their own agent at work. I think as long as the people working are smart and can use AI tools, they will be employable and will have what to do. Maybe start-ups will cut down the number of devs, because one smart employee can vibe code all the features needed by a start-up, but once the company is a big bigger, they will actually need more employees because they will be able to build more faster.

 I still think that the capacity of the people they will hire will be fewer than before.

I chose translation as my major. AI has changed the industry hugely. It is not about whether AI can actually replace translation anymore—it is about price. The cheaper option wins, and society does not really care about the end result.

I like the convenience AI gives me, yet I feel sad about how people treat translation and translators these days. Being a full-time translator is no longer my goal.

 When I was younger, I wanted to become a translator, but I decided quite well because now, AI can translate, communitcate in foreign languages without any problem.

 How did you make your choice at that time? Did you notice the future already back then?

I think basic communications are solved by AI, but for publication industry, I think that's a disaster. People cut their budget, the readers have to read books that translated and proofed by AI. (They can learn the language or read the original books with the help of AI though.)

Watching translation major falling down is really hurtful, I still believe translation shows the fundamental principles of human brain, but I guess no one is willing to invest it in a pretty long time.

 To be honest, I didn't believe in myself so much, I thought it would be difficult for me. It seems that my weakness helped me.

 haha, A loss may turn out to be a gain.

im on the other side of this. we build the kind of tool that client wanted u to use. so pinch of salt.

what surprised me is that volume stopped counting almost at once. we can draft 10 posts a day and it did nothing on its own. replies moved things. and someone still has to read the thread first.

 as there are so many AI-generated things, I do not have much time to read everything in the end.

The framing misses where the leverage actually moved. Handing execution to an agent raises the price of whoever can tell it what actually matters and catch it when it's confidently wrong. Most teams doing this well kept the same headcount and picked up a much faster pace instead, which is its own risk if nobody's watching for burnout. Judgment got more valuable. Attention is the real bottleneck now.

 do you feel that your job is safe?

 

I do, actually I'm the employer;) I've got developers, designers, marketers on the team, and the more we use AI (and we use it a lot) the more I value what these people bring on top of it. AI's not going to be the reason I say goodbye to any of them.

 aaaa, okay :D in this particular case, you have an upper hand :D