A classic blog post of the "I ran an expensive prompt in Claude, included a table of token usage, and wrapped up with a reflection on how machines once took jobs away from wheelbarrow pushers - yet chess players are still playing after Deep Blue" variety.
> I watched numerous YouTube videos and read many social media posts and blog posts by programmers about a profound sense of alienation, loss of human dignity, and anger over stolen independence.
Yes, it's funny to see programmers of all people, uber-technologists, shoot "i'm walking around in the woods videos" decrying the end of human civilisation merely because one specific craft practice is being automated.
AI isnt taking 'human creativity' away, this is an incredibly parochial and self-interested take. Its taking human creativity, as expressed in certain 20th C. crafts, away. Just as textile art has gone, and we don't have lace everywhere any more -- so it goes, that handcrafted programming goes to.
Engineers are those people who do that to human civilisation. That's always been the job. Replace people with machines. So it goes.
Everyone always wants to compare this revolution to the Industrial Revolution but there's a saying "history doesn't repeat itself but it rhymes". There are similarities, and reading how people coped back then can be helpful, but also this is a completely different revolution!
The only themes that rhyme imo are the loss of power to a smaller subset of humans (owners) and technological advancement enabling us to produce more (not necessarily better quality) with smaller costs. But people think that the history will repeat itself, and Jevons paradox will hold true, because it held in the Industrial Revolution; quote from the article: "Eventually, everything should be fine. The technology will be more powerful, and the job will still be there, it’ll just look different."
In this revolution, what if instead of being the human that invented the steam engine, we are the horses? "What happened to all the horses replaced by machines after industrialization? Humans keep a few around for entertainment but that’s about it." [1]
Also it took 80 years of events with “massacre” in name to estabilish basic limitations on capital class, like not exploiting children or allowing people to work less than 6x12 hours per week.
Sure, but horses today have unbelievably good lives compared to horses of the past. The Industrial Revolution basically introduced post-scarcity for horses.
The output isn’t interesting to me. I clicked around in both and it seemed kind of neat but like a portfolio project for school and certainly not art that would engage me.
I’m not sure how this causes an identity crisis. But if my purpose in life was to churn out random stuff that wasn’t meaningful to people, I would assume that I would be crisising much earlier than now.
I think alignment is the missing part of this post.
Creating something visually interesting Invisible-Cities-ish is basically a task with a billion or more satisfactory outcomes. You don't have a precise view on what makes an invisible city a good one as long as it has all the fancy looking bells and whistles. This means the AI can generate a bunch of tokens which as long as they technically work (which it seems the max-thinking did not fully) produce a "good" result. This makes you think - wow AI is impressive. However if many things in the system really mattered, it might have been much less successful as the AI may have made poor assumptions. This was basically a task that given a large volume of tokens and technical competence it's fairly easy for the AI to do a good job at because we don't evaluate it with any depth/precision. I think the same is true for music, where there are basically an infinite number of combinations of sounds, beats, melodies, etc. which are pleasant to hear - and songs aren't mission critical - so alignment is not a big issue. Generative AI is good at these tasks.
> The max effort + ultracode version is even more stunning, but interestingly it never got mouse input working without errors. If you want to test it out, you need to stick to keyboard controls - see the Use chapter in the project’s README file.
god is this a perfect encapsulation of the jagged nature of LLMs. incredible and kind of broken. for now at least
Makes sense though! It's really tough to test touch control programmatically, and that's all LLMs have. Same reason ChatGPT can't play Super Mario Bros.
Sometimes I think about that. In the LLM era, what is the meaning of my existence? Compared to the past, I’ve become able to do so many things, but there isn’t really anything I want to make. I don’t know where to place my identity.
This is a guess (correct me if I'm wrong) but probably you valued the mastery of a skill to create, and now the mastery is worthless as anyone can create anything, be it music, art or software if they just have access to latest AI?
It could be that, or it could not be—but rather than that, maybe it’s because work has decreased. Normally around this time of year is a season when freelancers have a lot of work, but there isn’t any, so maybe that’s why. People say I should just do creative work for myself, but maybe I’m feeling rushed because I don’t have any ideas.
And what’s wrong with that? Spending years to understand the intricacies of technology or to master a science or an art field, and as a result, among other things, gaining a competitive advantage in life over those who haven’t done that - is that a bad thing?
I think everyone enjoys that kind of feeling of intellectual superiority to some degree. But doesn’t using an LLM make it easier to feel that way even more? As you know, current LLMs produce different answer quality depending on the input, so using an LLM actually makes it easier to feel a sense of omnipotence. Papers that would normally be hard for me to understand—if I just put them into an LLM, it explains them clearly enough that I can understand them. I can’t say that the feeling you’re talking about doesn’t exist… but, well, I’m not sure.
Rather, I tend to feel that sensation even more now, while LLMs are still in their early stages.
For me; no identity crisis. Or at least a very short lived one until I realised what makes me tick.
I’ve been coding since I was 6. In my 40s now. Making stuff. Making stuff in meatspace, too.
I always told myself that I enjoyed coding, that I enjoyed making stuff.
I was wrong. I hate coding. I hate standing in a ditch with heavy cables slung over me. I hate neck and eye strain from hours bent over a soldering iron.
What I like, what drives me - the outcomes. The neat bit of software that I conceptualised and now have in my hands. The mad Heath Robinson contraption that somehow makes the machine that goes “ping!”, in fact, go ping.
I’ve always recognised that I am a lazy asshole - but there has always been a carve-out for creating - yet the giant heap of unfinished projects tells me my “eh, amanhã” attitude does in fact persist throughout many avenues of my life, even those I thought I enjoyed.
Me, I fucking love vibe coding. I get to go to bed at a sensible hour, and wake up in the morning like it’s bloody Christmas Day, as I know there’s going to be something cracking to play with. The ideas I have, realised, without me having to slog through the weeds. Nobody sucking their teeth at the scope. No “madaxe got distracted by another shiny object”.
I’ve executed on basically every digital must and should that I’ve come up with over the last decade, in the last nine months. I’m now onto the “could haves” and “would like to haves”.
And… I’ll be honest. Maintainability, understanding what it does under the hood - for the first time in my life I barely care about, as next year I can have a better model do it again, and better.
So, I like the outcome. I like having an idea and seeing it come to life. What happens between A and B isn’t terribly important, as it turns out that not all paths are the goal.
And I think the same may apply to a lot of creative people - I mean, artists had and have artisans for a reason - and now we can all have an ever-improving artisan.
waste of time.
Yes, it's funny to see programmers of all people, uber-technologists, shoot "i'm walking around in the woods videos" decrying the end of human civilisation merely because one specific craft practice is being automated.
AI isnt taking 'human creativity' away, this is an incredibly parochial and self-interested take. Its taking human creativity, as expressed in certain 20th C. crafts, away. Just as textile art has gone, and we don't have lace everywhere any more -- so it goes, that handcrafted programming goes to.
Engineers are those people who do that to human civilisation. That's always been the job. Replace people with machines. So it goes.
The only themes that rhyme imo are the loss of power to a smaller subset of humans (owners) and technological advancement enabling us to produce more (not necessarily better quality) with smaller costs. But people think that the history will repeat itself, and Jevons paradox will hold true, because it held in the Industrial Revolution; quote from the article: "Eventually, everything should be fine. The technology will be more powerful, and the job will still be there, it’ll just look different."
In this revolution, what if instead of being the human that invented the steam engine, we are the horses? "What happened to all the horses replaced by machines after industrialization? Humans keep a few around for entertainment but that’s about it." [1]
[1] https://news.ycombinator.com/item?id=50147755
I’m not sure how this causes an identity crisis. But if my purpose in life was to churn out random stuff that wasn’t meaningful to people, I would assume that I would be crisising much earlier than now.
Creating something visually interesting Invisible-Cities-ish is basically a task with a billion or more satisfactory outcomes. You don't have a precise view on what makes an invisible city a good one as long as it has all the fancy looking bells and whistles. This means the AI can generate a bunch of tokens which as long as they technically work (which it seems the max-thinking did not fully) produce a "good" result. This makes you think - wow AI is impressive. However if many things in the system really mattered, it might have been much less successful as the AI may have made poor assumptions. This was basically a task that given a large volume of tokens and technical competence it's fairly easy for the AI to do a good job at because we don't evaluate it with any depth/precision. I think the same is true for music, where there are basically an infinite number of combinations of sounds, beats, melodies, etc. which are pleasant to hear - and songs aren't mission critical - so alignment is not a big issue. Generative AI is good at these tasks.
god is this a perfect encapsulation of the jagged nature of LLMs. incredible and kind of broken. for now at least
Rather, I tend to feel that sensation even more now, while LLMs are still in their early stages.
I’ve been coding since I was 6. In my 40s now. Making stuff. Making stuff in meatspace, too.
I always told myself that I enjoyed coding, that I enjoyed making stuff.
I was wrong. I hate coding. I hate standing in a ditch with heavy cables slung over me. I hate neck and eye strain from hours bent over a soldering iron.
What I like, what drives me - the outcomes. The neat bit of software that I conceptualised and now have in my hands. The mad Heath Robinson contraption that somehow makes the machine that goes “ping!”, in fact, go ping.
I’ve always recognised that I am a lazy asshole - but there has always been a carve-out for creating - yet the giant heap of unfinished projects tells me my “eh, amanhã” attitude does in fact persist throughout many avenues of my life, even those I thought I enjoyed.
Me, I fucking love vibe coding. I get to go to bed at a sensible hour, and wake up in the morning like it’s bloody Christmas Day, as I know there’s going to be something cracking to play with. The ideas I have, realised, without me having to slog through the weeds. Nobody sucking their teeth at the scope. No “madaxe got distracted by another shiny object”.
I’ve executed on basically every digital must and should that I’ve come up with over the last decade, in the last nine months. I’m now onto the “could haves” and “would like to haves”.
And… I’ll be honest. Maintainability, understanding what it does under the hood - for the first time in my life I barely care about, as next year I can have a better model do it again, and better.
So, I like the outcome. I like having an idea and seeing it come to life. What happens between A and B isn’t terribly important, as it turns out that not all paths are the goal.
And I think the same may apply to a lot of creative people - I mean, artists had and have artisans for a reason - and now we can all have an ever-improving artisan.