when the bottleneck is communication (coordination costs scale polynomial with number of nodes in the graph), the optimal strategy is to have fewer nodes working harder. If AI makes individual nodes more effective, then in terms of people we’re going to eliminate jobs and concentrate effort. The mean job effort (including the unemployed) may be halftime while the median is 996 and the mode is 0. Whether it works is more of a question of policy and how deflationary the technology is - if energy, water, food, shelter becomes too cheap to meter due to rapid technology advances, then i guess maybe that is how governments stave off revolution and survive? Otherwise - civil unrest, etc
The page doesn't load, so I'll just add that we need AI that can clean up after AI.. but this might be difficult, because AI accepting AI output as input seems to be more likely to get "infected" and return the same AI crap.
Yes, this is why it's better to be a contractor/individual right now. You need to capture this efficiency yourself. You have access to the same tools faster than BigCos, so go hard in the paint. If you estimate 3h, and it takes 15m it's up to you how to reconcile this. Personally I bill the 3h.
Page doesn't load for me. I'm in the UK, not sure if geoblocked.
But a response to the title: "_buzzword tech_ is making us work more" - it's rarely the tech making us work more, it's normally the behaviour and attitude of businesses trying to profit from the tech that makes life hard for everyone.
The absolute worst place to be right now is in a B tech startup. Not only do you need to build some kind of app or product, you also need to build some kind of AI feature into the product. The users don't want it and never asked for it. It sucks all the resources out of your actual product that you should be focusing on, doesn't actually work or works non deterministically, but you are held to the same standards if it was another kind of software. And the only lever you have to pull is a lengthy model re-training or fine tuning/development cycle. The suits don't understand AI or what it takes to make it successful. They were sold on the hype that AI is going to save money, and forgot to budget for the team of AI engineers you'll need, infrastructure for training, extensive data annotations and reams of data that most startups don't have.
Tell me again how this isn't pure hell and the cuck chair?
I was never excited for automation. Automation doesn't mean we do less. It means that we do as much work, and now also the work has a higher complexity ceiling; you need to understand the systems that are being automated, and need to maintain the automation. More things are possible, but everything is more complex, and of course, you still need to work 40 hours a week. Products don't get better either, but that's more of a "shareholder value" problem than it is a specific technology problem.
And yes, I only talked about automation, but the same high-level issues apply to LLMs, but with different downsides: you need to check the LLM output which becomes a bigger topic, and then potentially your own skills stagnate as you rely on LLMs more and more.
Site is down to me, but I agree with the argument. I speak for myself: I think AI has removed a lot of small barriers that would naturally slow my work down, increasing dissatisfaction and stress with it. Without barriers, productivity increases and with it, work satisfaction. It's just nice to get things done(tm) faster and at a lower effort due to the quality of the virtual assistant.
I know it's making me work more, and I am thrilled. I have not shipped production code for 20 years, and it was desktop back then.
I am now able to single-handedly create webapp MVPs, one of which is getting traction. If anything actually takes-off, there will certainly be need for a real dev to take over. Also, my commits are not "vibe coded." I have read every single loc, and found so many issues that I am stunned that "vibe coding" is actually a thing. I do let the models run wild on prototypes though.
I think that I happen to be in some magical sweet spot as a person who knows the words, kept up with tech, but not the syntax of framework xyz.
I thought this sweet spot was very transient, and I am very happy that the tools appear to be reaching a plateau for now, so I still have at least another year of being useful.
Since agentic dev tools arrived, I am having the time of my life while gladly working 60hrs per week.
I realize that I am an outlier, but is anyone else in this same boat? If you have product ideas, is this not the best time ever to build? All of our ideas are being indirectly subsidized by billions of VC & FAANG dollars. That is pretty freaking cool.
Sounds like these are hobby projects that you're doing on your own. The point of AI making us work more is referring to people at companies, not for themselves. Employers are capturing the productivity benefits (if they exist), not employees.
Yes. What they've shown me is that I don't like code. Hate it, actually. But I like building things, and can still read and reason about problems and code. So I chose the best stack for what we're doing, but I don't get bogged down in arcania. It's great.
I am in the same boat. But i am getting overwhelm with the code quantity i have to review[0] before i update my code. I don’t let llm touch my code, i ask for code for this or that and then save the code for future review. But ideas flow in faster that i can review then.
I'm seeing amount of changes needed to produce new features when coding with these AI tools constantly increasing, due to the absence of a proper foundation, and due to the willingness of people to accept it, with the idea that 'we can change it quickly'.
It has become acceptable to push those changes in a PR and present them as your own, filled with filler comments that are instant tech debt, because they just repeat the code.
And while I actually don't care who writes the code, I do expect the PR author to properly understand the code and most importantly, the impact on the codebase.
In my role as a mentor I now spend a lot of time looking at things written and wonder: Did the author write this, or did they AI? Because if the code is wrong, this question changes how the conversation goes.
It also impacts the kind of energy I'm willing to put in into educating the other person as to how things can be improved.
The tldr is that AI works like multiplier on both sides of the equation. Not only we will work more but we will get even more stressed because things will be moving at increasing speed - perpetually - until we hit some limit of course .
Just wait till AI starts taking managerial roles. Instead of annual reviews it'll be hourly reviews. "You only have five background AI coding sessions in progress right now. PIP!"
I'd like to know, from people who really believe that we just need to invent the right technology, then we could all do 20 hour weeks and spend the rest of the time in leisure, what keeps them from doing that right now, and what exactly they believe that miracle tech would change.
It seems relatively obvious to me that if a society has work as its cultural core then no amount of productivity increase will get rid of work - it would destabilize the entire society before it could do so.
To be honest the self confidence of ai is rubbing off on the species in a good way. As in suddenly there is somebody there who beliefs in you, unlike that family that constantly sabotage you, the partner that tries to keep your ego in a "handleable due to absence of success" shape. And all those ideologies that want to cripple any one standing out, they can take a hike. So what if its confidently wrong, at least it beliefs more in you then the rest of the species.
47 comments
[ 4.6 ms ] story [ 78.2 ms ] threadBut a response to the title: "_buzzword tech_ is making us work more" - it's rarely the tech making us work more, it's normally the behaviour and attitude of businesses trying to profit from the tech that makes life hard for everyone.
Tell me again how this isn't pure hell and the cuck chair?
#andRepeat
And yes, I only talked about automation, but the same high-level issues apply to LLMs, but with different downsides: you need to check the LLM output which becomes a bigger topic, and then potentially your own skills stagnate as you rely on LLMs more and more.
As jaded as that may be, I believe LLMs for many will become our bigger shovels.
I am now able to single-handedly create webapp MVPs, one of which is getting traction. If anything actually takes-off, there will certainly be need for a real dev to take over. Also, my commits are not "vibe coded." I have read every single loc, and found so many issues that I am stunned that "vibe coding" is actually a thing. I do let the models run wild on prototypes though.
I think that I happen to be in some magical sweet spot as a person who knows the words, kept up with tech, but not the syntax of framework xyz.
I thought this sweet spot was very transient, and I am very happy that the tools appear to be reaching a plateau for now, so I still have at least another year of being useful.
Since agentic dev tools arrived, I am having the time of my life while gladly working 60hrs per week.
I realize that I am an outlier, but is anyone else in this same boat? If you have product ideas, is this not the best time ever to build? All of our ideas are being indirectly subsidized by billions of VC & FAANG dollars. That is pretty freaking cool.
[0] https://medium.com/@xcf.seetan/adventures-on-the-ai-coding-s...
I'm seeing amount of changes needed to produce new features when coding with these AI tools constantly increasing, due to the absence of a proper foundation, and due to the willingness of people to accept it, with the idea that 'we can change it quickly'.
It has become acceptable to push those changes in a PR and present them as your own, filled with filler comments that are instant tech debt, because they just repeat the code.
And while I actually don't care who writes the code, I do expect the PR author to properly understand the code and most importantly, the impact on the codebase.
In my role as a mentor I now spend a lot of time looking at things written and wonder: Did the author write this, or did they AI? Because if the code is wrong, this question changes how the conversation goes.
It also impacts the kind of energy I'm willing to put in into educating the other person as to how things can be improved.
https://chatbotkit.com/reflections/why-ai-coding-agents-crea...
The tldr is that AI works like multiplier on both sides of the equation. Not only we will work more but we will get even more stressed because things will be moving at increasing speed - perpetually - until we hit some limit of course .
It seems relatively obvious to me that if a society has work as its cultural core then no amount of productivity increase will get rid of work - it would destabilize the entire society before it could do so.
I have to believe this is satire.