I spoke to a CTO of a firm doing relatively cookie-cutter work (custom software for things like waste management). They were able to lay off a significant portion of their workforce due to AI. I think it's eating in more at the bottom-end of coding right now.
It definitely makes me more productive in high-end, by more than what I'd call marginal, but not by enough that it's risking jobs.
However, the article, and my comment, were more general than coding. The impact is different industry-by-industry, from zero to very high.
Whats interesting is I think there might be a disconnect between AI capabilities and tech executives. I'm guessing executives believe AI will catch up to be good enough to multiply engineers productivity say 120% within the next few years - hence the 20% layoffs everywhere.
Maybe this will be the case. But for me on the ground its producing a lot more work. I now have to code review everything starting at a very high level working my way down and there's just so much MORE code now.
I'm well aware of this and in those sorts of jobs, most of the work isn't programming, it's sitting in meetings and communicating. While AI might help with this simply by shrinking the amount of people needed to do a job, I'm very dubious about a 70% efficiency gain when you look at the entire scope of what these workers do.
The entire concept that if you make coding 70% more efficient that you'll make coders 70% more efficient is simply wrong.
This is pretty muchy experience as well. AI is a fantastic helper, and it will make devs more productive, but it is not going to put all software devs out of work. Probably a tiny fraction of them at best.
However, with recent grads flooding into IT for remote work and high pay, they could be hurting as AI reduces the need for entry level roles. Entry level was already saturated, and now it will be more saturated, with AI reducing the need for jobs.
AI is coming from software engineers, and increasing their impact.
Yeah, it makes it take less specialized knowledge and effort to produce code of any given scale, and if the quantity of impact for programming were to be static, that would mean fewer programmers.
But that’s true of every advancement in the history of programming, which just keeps causing employment of more programmers at higher wages in the long term.
One way to think about ai programming is that it's another step on the continual ladder up from writing machine code to writing more in more 'human readable' programming languages. It's essentially a non-deterministic compiler from natural language to an intermediate representation like python. Even assuming it was able to near perfectly translate intent to code, that doesn't mean it will replace programmers, necessarily.
The way to think about the impact on employment is comparative advantage. https://www.investopedia.com/terms/c/comparativeadvantage.as... Even if AIs are better than humans at every programming task, there is still a limited amount of compute in the world and unlimited amount of potential work to do, so there will always be tasks available for humans to work on. If the cost of writing code becomes cheaper, there will be tasks to automate that aren't worth the effort to automate now, that will _become_ worth the effort to automate in the future.
The modern developer tooling already had a much larger impact on productivity than AI probably will and the only thing it did is increase even more the number of developers.
If I was the CEO of a company making headcount reduction with AI, I would be more worried about my company itself than the job of the ones I'm firing.
But has anyone here seen any evidence of AI impacting software jobs? Maybe my company is just behind the times, but I'm seeing no indication that AI is going to displace anyone. Copilot and GPT-4 are convenient and speed up certain parts of the job, but the parts they impact aren't where the bulk of my time was spent anyway.
Just take AI assisted programming, if it makes devs twice as effective at their jobs then its massively increased the output of our most profitable industry.
This is just one field, AI will impact the output of most fields
I’d believe if you did nothing with AI and just fired a big chunk of your bottom tier cheap outsourced programming “talent” your productivity would significantly increase as a result of your actual talent no longer being occupied with managing and reacting to low quality code contribution and contributor hand holding.
I've been working as a programmer for 29 years now, and almost since the beginning there has been a story ciruculating about how new technology ('4GLs', then 'Rules Engines' and now 'AI') would put programmers out of business.
It's never happened, of course. But if AI does ramp up and take over a bunch of programming space, it may drive IT jobs ever lower on the social/pay scale, perhaps even elevating manual labor.
My impression is that AI has done more for the accessibility of coding than the efficiency of coding.
Sure, it's much more possible for bob from accounting to throw together some trivial python script than it was a few years ago, but it seems the main benefit more skilled programmers get from it seems to be helping with syntax and pumping out boilerplate. Which is quite a bit of programming work but well less than 70% of it.
When a problem is tricky, it seems people turn off the autopilot.
Strong disagree. Look at the crunchiest of people who are doing very low level work. I wouldn't say they are really making more than anyone else. In a lot of ways they are at a disadvantage because their skills are very niche and very specific. If they find work it can be good. But it can be a frustrating exercise to find that work.
People pay for productivity, making a product of some sort that has value. These AI tools will increase productivity. And they will only get better with time. That translates into the productivity of people who have mastered these tools as more valuable. They just get more done.
Will syntax issues and other problems be an issue with early AI? Sure, but it will get better. The AI will start to catch more issues and move farther up the stack. We still need to know enough to guide it, but the details required will be fewer.
I think it also has hurt a lot of software engineers in a number of different ways. AI assisted programming tools are an experiment and the jury is still very much out.
Technology always has a weakening effect. AI copilots will make us lazier and produce more buggy code. It can increase productivity for certain boring tasks (generate a bunch of empty classes, etc.). It can increase productivity for early on in projects. Otherwise it's just a crutch.
That sounds like a good thing. There are many people paid to write software who absolutely cannot write original code and have no idea how things actually work. A lot of that can be, and probably should be, eliminated by AI.
I'm talking about time spent. I completely agree that all the work you've mentioned is stuff that takes time and that AI won't replace any time soon, but I strongly doubt that it's the bulk of the work.
I'm talking about the programmers that a friend of mine works with who take several quarters to produce a dashboard, or the entry level PHP jobs that make barely above minimum wage and sell custom websites and hosting to real-estate companies in small towns, or the outsourced programmers that other friends interact with, who spent ~10 eng over 2 years building a "PaaS" style system in shell scripts that no one in their company actually uses because surprise surprise no one ever thought to ask whether it was necessary or had the right feature set or even worked (but management paid for it).
These people are spending a significant amount of their time coding, but they're not spending time automating, or improving the process for next time, and so the work they're doing is fairly trivial and is most likely to be impacted by AI.
It definitely makes me more productive in high-end, by more than what I'd call marginal, but not by enough that it's risking jobs.
However, the article, and my comment, were more general than coding. The impact is different industry-by-industry, from zero to very high.
reply