The AI boom is screwing over Gen Z | ChatGPT is commandeering the mundane tasks that young employees have relied on to advance their careers.::ChatGPT is commandeering the tasks that young employees rely on to advance their careers. That’s going to crush Gen Z’s career path.
I like to compare modern LLM to Excel or calculators in the past. Some years ago a company would have an in-house team of accountants. Then came Excel and now a single accountant can do the job for 10 companies. Let’s now consider programmer: currently a project manager oversees a team of programmers, most of whom are only responsible for mundane work of typing out code. With AI a single worker will be able to perform more productive than that team of programmers, because they will offload the boring work to AI and focus all their attention to what AI is perhaps incapable of.
What this article is really saying, which I agree with, is that AI improves productivity ,just like perhaps the steam engines did in the 1800’s. But this time the problem is we won’t increase the output and let the workers work more efficiently and earn more money, because it’s not manufacturing jobs which were limited by technology that this is influencing. It’s office jobs, which the economy has a pretty much fixed demand for. Workers will not improve their productivity, they will just be replaced because their work can be offloaded to a machine capable of doing that same jobs better in every significant way.
Can you elaborate on the “fixed demand” aspect?
From what I know as a software engineer, companies would simply make twice as much software, if their software engineers were twice as efficient. There are always requirements pushed out of scope because the complexity of the solution is growing and growing. The ability to make more complex software solutions with the same amount of engineers is not going to result in less engineers, it is just going to cause more complex software products.
Also note that more engineers has deminishing results due to communication losses. This, along with a fixed supply of engineers seems the biggest limitation to the industry to me.
Only if there’s demand for twice as much software. Otherwise, you make the same software twice as fast and with half as much work. Let’s go back to the example of accountants. Sure, the demand for accounting work may be somewhat increasing, but with productivity per worker increasing orders of magnitude faster than demand, the overall number of accountants shall decrease. A piece of code a junior programmer writes within a week can be obtained immediately with a tool like chatgpt simply by formulating a clear prompt – it’s not like we’re talking about better keyboards which improve your typing speed and therefore increase your productivity by 10% by letting you type code faster, it’s actually orders of magnitude!
Again, sure, but wouldn’t you agree the technology will some time reach a point where more complexity is redundant? I would argue it’s sooner than later, see how smartphones and computers keep improving in their performance, but there are no technology breakthroughts anymore. Is infinite growth even possible?
Not sure what you’re getting at here, so let’s go back to your original question: what do I mean by fixed demand of the office jobs.
Doing accounting faster will not land you more gigs anymore, unless you steal some other accountants’ clients. Writing longer reports will not make your employer require you to write more of them, unless they fire your colleague who does that too. Going through motions and legal documents faster will not magically give you more legal work, unless a different legal counsel changes industry. Unlike manufacturing in the 1800s, the supply and productivity of modern jobs are not limited by technological disadvantages so much, but instead, the demand for this work is corelated with other branches of economy.
One could argue, in fact it’s an ongoing debate where I’m from (Poland): “yeah sure but when we started switching away from coal then miners were supposed to be off work as well and yet they mostly managed to find previously non-existing jobs in newly created industries and the unemployment remained low”. Right, but in that case one industry was replaced by another, workers’ productivity could be moved to doing something else. This time it’s different, because the jobs don’t change, the demand doesn’t change, instead the supply of labour (via increased, AI-fueled productivity) increases so much, that large part of the workforce is found to be straight up redundant.
deleted by creator
Interesting point, I guess you are right