Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

AI doesn’t reduce work, it intensifies it

simonwillison.net

Post reply on HN
Print view View original

131–140 of 329 posts

Re: AI doesn’t reduce work, it intensifies it

#131

Post by FeteCommuniste »

jillesvangurp wrote: post #126

That's the bane of all productivity increasing tools, any time you free up immediately gets consumed by more work. People keep on making the same naive assumption that the total amount of work is a constant when you mess with the cost of that work. The reality is that if you make something cheaper, people will want more of it. And it adds up to way more than what was asked before. That's why I'm not worried about los…

> That's why I'm not worried about losing my job. The whole notion is based on a closed world assumption, which is always a bad assumption.

You might be right, but some of us haven't quite warmed to the idea that our new job description will be something like "high-level planner and bot-wrangler," with nary a line of code in sight.

Re: AI doesn’t reduce work, it intensifies it

#132

Post by sawmurai »

co_king_3 wrote: post #100
sawmurai wrote: post #93

Earlier quoted context omitted.

I am all for labor organization. I just don’t see how it would be of benefit in this particular case.

If I'm not mistaken it would appear that you're saying that you are in fact *not* for labor organization in this case.

No, absolutely not. I would even be for labor organization if it had no impact on this matter primarily because I don't see why it would be a negative.

Re: AI doesn’t reduce work, it intensifies it

#133

Post by simonw »

co_king_3 wrote: post #125
latexr wrote: post #97

> the productivity boost these things can provide is exhausting . What I personally find exhausting is Simon¹ constantly discovering the obvious. Time after time after time it’s just “insights” every person who smoked one blunt in college has arrived at. Stop for a minute! You don’t have to keep churning out multiple blog posts a day, every day. Just stop and reflect . Sit back in your chair and let your mind rest. W…

> You don’t have to keep churning out multiple blog posts a day, every day. How do you know that? You don't think he's being paid for all this marketing work?

I'm paid by my GitHub sponsors, who get a monthly summary of what I've been writing about on the basis that I don't want to put a paywall on my content but I'm happy for people to pay me to send them less stuff.

I also make ~$600/month from the ads on my site - run by EthicalAds.

I don't take payment to write about anything. That goes against my principals. It would also be illegal in the USA (FTC rules) if I didn't disclose it - and most importantly it would damage my credibility as a writer, which is the thing I value most.

The big potential money maker here is private consulting based on the expertise (and credibility) I've developed and demonstrated over time. I should do more of that!

I have a set of disclosures here: https://simonwillison.net/about/#disclosures

Re: AI doesn’t reduce work, it intensifies it

#134

Post by Lalabadie »

jasode wrote: post #89
falloutx wrote: post #29

I am becoming more and more convinced that AI cant be used to make something better than what could have built before AI. You never needed 1000s of engineers to build software anyway, Winamp & VLC were build by less than four people. You only needed 1000s of people because the executive vision is always to add more useless junk into each product. And now with AI that might be even harder to avoid. This would mean the…

, Winamp & VLC were build by less than four people. You only needed 1000s of people because the executive vision is always to add more useless junk into each product. Many types of software have essential complexity and minimal features that still require hundreds/thousands of software engineers. Having just 4 people is simply not enough man-hours to build the capabilities customers desire. Complex software like 3D m…

But big projects are where the quality of LLM contributions fall the most, and require (continuous, exhausting, thankless) supervision!

Re: AI doesn’t reduce work, it intensifies it

#135

Post by aetherspawn »

mentalgear wrote: post #48

I'm also coming to the conclusion that LLMs have basically the same value as when I tried them out with GPT-3 : good for semantic search / debugging. Bad for generation as you constantly have to check it, correct it, and the parts you trust it to get "right" are often those that are biting you afterwards - or if right introduce gaps in your own knowledge that make you slowly inefficient in your "generation controller…

Yes, and I’m convinced AI companies either pay or brainwash these people to put out blog posts like this to spread the idea that it actually works. It doesn’t.

Re: AI doesn’t reduce work, it intensifies it

#136

Post by simonw »

latexr wrote: post #97

> the productivity boost these things can provide is exhausting . What I personally find exhausting is Simon¹ constantly discovering the obvious. Time after time after time it’s just “insights” every person who smoked one blunt in college has arrived at. Stop for a minute! You don’t have to keep churning out multiple blog posts a day, every day. Just stop and reflect . Sit back in your chair and let your mind rest. W…

Writing about "the obvious" is a useful service. Often people doubt what their own experience is telling them until someone else helps confirm their suspicions and put them into words.

Re: AI doesn’t reduce work, it intensifies it

#137

Post by fendy3002 »

teekert wrote: post #129
fendy3002 wrote: post #127

Earlier quoted context omitted.

Now that you get accustomed with Hugo, I wonder if the way you plan & prompting now will produce better result or not

I probably will, I use AI extensively, but mostly when I can't remember tedious syntax or suspect something can be done in a better way. And that work well for me... If I go too much towards vibe coding, the fun is sucked away for me.

thought so. I find that too much vibe coding (less spec) will make the AI perform worse, even with 4.6 opus. Pseudocode is obviously the best they perform, having a good lower-level specs usually provide a good result too.

Re: AI doesn’t reduce work, it intensifies it

#138

Post by mfru »

latexr wrote: post #106
darkwater wrote: post #72

Earlier quoted context omitted.

> Previously, I'd have an idea, sit on it for a while. > With agentic development, I have an idea, waste a few hours chasing it, What's the difference between these 2 periods? Weren't you wasting time when sitting on it and thinking about your idea?

Sitting on an idea doesn’t have to mean literally sitting and staring at the ceiling, thinking about it. It means you have an idea and let it stew for a while, your mind coming back to it on its own while you’re taking a shower, doing the dishes, going for a walk… The idea which never comes back is the one you abandon and would’ve been a waste of time to pursue. The idea which continues to be interesting and popping…

They for sure weren't vaporating hundreds of litres of water and wasting a bunch of electricity while doing it

Re: AI doesn’t reduce work, it intensifies it

#140

Post by zozbot234 »

Lalabadie wrote: post #134
jasode wrote: post #89

Earlier quoted context omitted.

, Winamp & VLC were build by less than four people. You only needed 1000s of people because the executive vision is always to add more useless junk into each product. Many types of software have essential complexity and minimal features that still require hundreds/thousands of software engineers. Having just 4 people is simply not enough man-hours to build the capabilities customers desire. Complex software like 3D m…

But big projects are where the quality of LLM contributions fall the most, and require (continuous, exhausting, thankless) supervision!

Big projects can still be highly modular, and projects built by "1000s of devs" typically are. If your desired change can be described clearly without needing too much unrelated context, the LLM will probably get it right.
Post reply on HN