AI Makes Me Sad(mondobe.com) |
AI Makes Me Sad(mondobe.com) |
My sad thought: this tech's primary audience is people who have none of those.
Yes, that is part of the inevitability strategy. That is absolutely allowed and encouraged!
The rest of your essay is of course pure heresy! :)
You are allowed to be optimistic; pessimists are "losers" (dhh - https://www.youtube.com/watch?v=vDjW_dRyKXY).
You are just throwing a temper tantrum (https://archive.is/YNxJq).
Do not commit blasphemy against AI. (https://www.reuters.com/business/media-telecom/talk-bubble-i...)
I mean, there are a bunch of reasons one might. Making hugely impactful products can be great for your career, even if you only work there for some time. Maybe you believe you'll be better-positioned to see what's coming. Or cynically you want to make absolute bank so you can hide out in your New Zealand bunker when the AI dictatorship begins or whatever. Or you'd rather be a hammer than a nail, or in this case, the automator rather than the automated.
Or you might believe that superintelligence won't necessarily lead to huge job loss, won't happen that soon, or will bring about the fully automated luxury communist future so it's your moral imperative to be maximally accelerated.
AI doesn't make me sad but this drives me crazy. I don't want racists banned, I want replies to posts that don't make any earthly sense banned. I want twitter to stop you before a post and say "did you even read the post?" "Are you just posting because you want to participate?" "Banned for nonresponsiveness." "Six other commenters submitted comments almost exactly like yours, are you sure yours in necessary?"
Shouldn't AI have solved this? I can't imagine that it's holding people on the site - maybe it pushes up some badly done interaction metric. But there should be a last AI check before submission: "Are you dumb? If you are dumb, press submit, but I have to warn you that the comment you're replying to has showdumb:off."
Why would not using tools that you don’t know how to use that don’t have the negative effects that you don’t have experience with make one sad?
This is just complaining. If you don’t like AI, don’t use it. If it makes you sad, don’t spend time blogging and focusing on it. If you think it entails a loss of control, maybe educate yourself before smearing non-issues with tools you have never tried.
This phase will last like 2 years max I am guessing. What is it with all the linear extrapolation of AI progress these days?
If model progress completely froze here what does that future look like? Chinese labs would continue to bring down the cost of what we have and a lot of companies would lose a lot of money on data center buildout but does software as an industry ever close Pandora's box?
As much as I'm loathe to use it and think criticisms about a loss of capability, quality and skill are valid I also think for the majority of software jobs as-was will remain changed for good. There was always a struggle between 'cheap and quick' offshoring and actual quality software but I don't think the author is necessary overreacting by feeling unsettled.
I hadn't realized people were already looking through inherited-nostalgia glasses back at 2016.
Instead you'd have written a post complaining about the ad-tech space - which is still a deep pit of human depravity and unhappiness mascaraing as a positive on society.
The value-add startup you're imagining still exists. Just not anywhere near the bubble where people are talking about wanting 500k salaries.
I don't doubt this is genuine, but I think it points at a broader rift in engineering where what some consider the craft differs: expressing your craft in hand written code or expressing your craft in making software. I lean heavily towards the latter, which means I have the opposite experience as the OP. My creativity has exploded with AI. I feel my craft has expanded into areas that were previously prohibited by time - whether thats actual typing or doing manual research on SO or Google. My control feels improved because my mind is freed to focus more on my vision.
I doubt I'm the only one that feels this way. It doesn't lessen or make the feelings of others no less legitimate. I just think perhaps there's a communication rift where each side isn't necessarily understanding where the other is coming from. As with most things, understanding and nuance is necessary here.
It’s liberating, as you say, but at the same time I feel guilt for building when I know I’m really standing on the shoulders of giants, maybe undeservedly.
Wanting to use Twitter makes me sad.
Obviously it has been degraded by the switch to algorithm over user curation of their own timeline, literally everything Musk did to it, etc, but old Twitter was good, hence the desire to "go back".
I became very bored with coding early in my career, even though I appreciated well crafted interfaces, beautiful abstractions, etc.
So I moved more towards management of teams, etc. It was far more interesting to craft research directions, figure out how to map intellectual landscapes with experiments and prototypes, etc. I loved the results of coding, appreciated well crafted coding, but doing the actual coding was just boring as hell. It's a means to and end for me.
LLMs have completely reinvigorated my entire career in actually producing code. It takes a fair amount of work to constrain the LLM, rein it in from creating 500 lines of slip where 5 would do, etc. But the drudgery is gone.
I now wake up invigorated to tackle new projects. I fall asleep dreaming about new approaches I can try, because now I really can try them, quickly, without weeks of drudgery. It's a second larval stage.
Additionally the ongoing experience with iPaaS agentic platforms removed the desired to invest into additional programming languages unless they are required for project delivery.
I guess it is ok for those that rather be solution architects or program managers of a little bots team.
You didn't.
I started my career and my first company during the dot-com boom, before YC even existed and the word "startup" took its current narrow shape.
Things were ugly right after the bubble burst, but the years that followed were some of the best for building I can remember.
The tools change, and not necessarily for the better, particularly if you care about the craft for its own sake. But the hard part was never the build. Even back then, before AI, anyone could copy you fast. The hard part was knowing what to build, for whom, and how to reach them.
What anyone calls a startup doesn't matter. Build the things!
The AI hype and AI doom are 2 sides of the same coin. Yes you can make a cool demo quickly with a prompt, but building a real company with moats is still possible, much easier even. The moats have just changed: proprietary data, distribution, access rights, eval sets, integrations, compliance, brand, just not the code.
I mean for me at least, this is not at all what it feels like. About 50% of my working time is split on architecting, planning, designing, testing, automating, reviewing. 20% on meetings and I guess the rest "prompting" but a lot is automated. I don't really mind other than the second order effects of slop grenades hitting my team by less diligent or shitty engineers.
Basically a lot more time spent on "how should this work?" vs before a lot more time spent on "how can this work?"
1) the mass proliferation of low-effort content. A thing can be implemented quickly and correctly but still be badly designed. When people say "oh this will go away as the models improve", I think they are making a category error. The AI cannot read your mind.
2) we really have no idea what to do about how this transforms society. As a result, the debate has been captured by quasireligious doomsday cultists and self-interested accelerationists. Obviously we should improve our cybersecurity and synthetic biology regulations (although I'm a skeptic on the bioterrorism risks). But we had no idea how much time people would pour into social media feeds when they were invented. I am sure we will be careless with this technology as well in this 2nd gilded age.
Some context - I'm a frontend dev and I've lost all interest in work, I just see it as a way to pay my bills now. I'm trying to practice drawing because I want to draw fanart of my favourite characters, and I want it to be meaningful. I've already dabbled with stable diffusion back in the day and I got bored of it because it was just way too easy and didn't require any effort, and thus it didn't feel any rewarding either.
Excitement. Fear of job loss. Fun at vibecoding toys. Worry about falling behind. Apprehension about how society will deal with this
Maybe this is what makes you sad
I'm sorry that you think social media was good before AI came along, but the truth is that it has always been junk food and was ruined by humans a decade before all of this came along. It wasn't a chat bot that was harassing Ariel Waldmen on Twitter in 2007 and it wasn't a chat bot that refused to do anything about it.
If AI finally mercy kills toxic social media platforms it's doing all of us a favor. You might not believe there is a life beyond Twitter, but I promise you that there is and it's a lot better.
I hope it's like those highly paid html coding jobs (where this is your only skill) in the 90s and these people get booted within a few years.
My problem with AI is not the technology. It's people attempting to shortcut knowledge and skill and act like it Trumps years of experience in the industry.
When everyone can prompt, the only high paying jobs left will be the people with actual experience.
As part of my job I host conversations specifically around this, I had a long dinner last night with a diverse group of congressional staffers, engineers, lawyers and managers at the wharf in Washington DC.
I also just got back from Africa and the Middle East talking with student students and educators as well as businessman across those regions.
What’s clear to me is that all of the consistent complaints about AI have nothing to do with AI as a technology. This isn’t new, it’s the history of technological revolution.
To a person, every complaint has to do with the intersection of commerce, government and social structure - and more specifically the management and identification of externalities.
So almost all of the argumentation and debaters are talking past each other
Extremely “pro” AI people talk about all of the real documented progresss but downplaying the risks and challenges and externalities. This is how every revolutionary technology (ICE, spinning jenny, steam engine, etc…) was sold and we live in that.
On the other side the people who are “anti-AI” almost exclusively talk about externalities and never talk about any benefits and any benefits that they do acknowledge are either “not AI” because they aren’t explicitly only LLMs or corrupted by the externalities themselves.
The challenge is both sides are correct in part but they can’t communicate with each other other.
It is all busywork and giving the appearance of an industrial production line to please management.
On the other hand we have spam, blogspam, AI SEO, slop, surveillance, bombing of girls' schools, a bubble, inflation through overspending etc.
It is quite worrisome that you get to influence politicians.
This is a result of the technology being expensive and centralized vs cheap and decentralized. The marketing claiming it’s aggressive and takes from humans vs uplifting and enriching. And on and on
It makes it really clear what the right thing to do is, make the technology widely available, cheap and let humanity do its thing.
I mean I get it, but if the bubble actually bursts, you are not going to be happy about it, lol.
It's a thing inherent to thinking, language, machines and computers.
The thing is, it looks like life but it really isn't.
That fact is being rubbed in our face now. Treat it like life and it will make you sick, sooner or later.
Sooner now, because of AI. So thank AI for making the badness more immediately visible before it can do more damage. So you spit it out immediately instead of swallowing it for years.
BUT. As far as using AI goes, I think it's great. And I'm not just talking about coding. In fact, coding is a small part of it for me. I mainly value AI tools like ChatGPT / Gemini / etc. for their value as:
1. Tutor
2. Lab assistant (more on this in a sec)
3. Research assistant / intern
4. Humor generating interlocutor (more on this to come)
To point 1... ChatGPT is amazing as a tutor and teacher. Pretty much anything I want to learn more about, I can ask it to give me a summary, then drill down on specifics, and then ask it to quiz me. And I can do this over and over and over and over, and it never gets tired, or bored, or frustrated. And I can do this at 2:00 in the morning when some random idea sparks in my head. Even more, if I'm reading a print book (something I still do a lot of) and I encounter, say, a bit of math notation that I'm unfamiliar with (this happens quite often), I can take a picture of the page, put it in the chat, and ask the AI "what's the deal with that notation in the third paragraph on this page?" and BOOM ... I get an explanation. And then I can just randomly have it quiz me on stuff as sort of a "poor man's Anki" (don't worry, I still use actual Anki as well).
To point 2... I've recently been doing some research in the area of neuromorphic computing. Now just to be clear, when I say "research" I mean really entry level stuff, not groundbreaking novel research. Think "lab section of an undergrad science class" level stuff. But in that regard, ChatGPT has been a god-send. It has explained SO many things to me about something as simple as using my oscilloscope, that I'd never bothered learning before. And it helped me with some Python / PyVisa code to control my test equipment to automate some experiments using SCPI over LXI. And I can upload screenshots from my oscilloscope and it can analyze the waveforms and point out interesting details, notice my mistakes, see small things that my eyes can barely pick up on, etc., and MUCH faster than I can. My next plan is to set up an MCP server to let it actually query/control the test equipment using SCPI.
To point 3... I recently gave it a task like this "I was watching a Youtube video where the presenter mentioned that when she first got into neuromorphic computing as a grad student she read 3000 research papers. I want to treat that as the floor for myself, so go find me at least 3000 relevant papers related to neuromorphic computing / neuromorphic electronics / spiking neural networks / etc." And BOOM.. a few minutes later I had a spreadsheet listing thousands of papers, including links to publicly available copies where available.
To point 4... this is harder to explain. When I talk to ChatGPT, I don't just "play it straight". I'm a class clown by nature, and I go ahead and let that enter into the conversations with the AI. I joke around about all sorts of shit, make bizarre references to completely unrelated stuff, constantly refer to the Turing Police as though they are a real agency who are watching us, joke about the poor OpenAI intern who has to read through all the transcripts to watch for inappropriate behavior (I don't think that's a real thing, but the running gag continue regardless). And the bot "knows" when I'm joking and *replies in kind*. And it's genuinely funny. I've literally snorted coffee before over something ChatGPT said to me. And it knows about the various running gags that we always engage in: the Turing Police references, the OpenAI monitor, a mysterious character name "Earl" who may or may not exist, and who may or may not have been onboard the USS Eldridge in 1943 in Philadelphia, etc. etc.
One of my favorite moments was when ChatGPT came back with a fictionalized report from the supposed "OpenAI transcript monitor" that included the phrase:
"User and model both appear to be morons."
So yeah, I find it wildly useful and you basically could not pay me enough to give up these tools.
And as a final word: just to be clear, yes, I know enough about how Generative AI works to know that the AI doesn't "really" have a "sense of humor", or a "personality" at least not the same way a human does. I allow myself to anthropomorphize it a bit, because it makes interacting with it more fun and helps the time go by faster so to speak. But at all times I'm fully aware that what I'm talking to is ultimately a program doing inference on a huge neural network, doing vector looks, etc.
A couple of Fridays ago, I'm sitting at Namu, a cafe in Durham. I'm chatting with ChatGPT about something I'm working on, and casually mention "I've been meaning to learn KiCAD to design PCB's for, like, forever."
And ChatGPT replies "Well I can help you with that." So I figure, "why not bite the bullet now and see what happens?"
About 6 or 7 hours later, including the drive home from Namu, its 4:00'ish in the morning, and I'm generating Gerbers for my first PCB. Sent the files to JLCPCB and by Wednesday of the following week I had boards in hand. Soldered one up and it worked perfectly, first go. Now to be fair, the circuit itself was VERY simple. Just a 555 timer to blink an LED. The goal was to learn the workflow for designing the board and having the boards fabbed, not to create a useful product.
Now you might rightly ask "Couldn't you have done it the old fashioned way by watching Youtube videos, reading books, reading tutorials, and trial and error"? To which the answer is clearly "Yes". But having the AI there to coach me through the process was a tremendous help in finally getting over that hump of getting started and actually doing it.
One might also ask "Did you really learn anything, or did the AI do it all for you?" To which I'd say "I still did all the hands on work in KiCAD myself, by hand. The AI advised and coached, but I did it, so yes, there is some retention. Not 100%, but it also wouldn't be 100% if a human coach had coached me through it." I'll do another board, and have less help from the AI, and then another with even less help, and so on over time. I've already started that process by starting on another board.
I used to think that. I came to the realization that I could have technically built Facebook the product, but would never in a million years have built Facebook the company. But having just finished "Careless People" I feel kinda good about that.
I also just got back from a vacation to Alaska. Lots of gold rush stories from Skagway. One of my favorites was a lady who became massively wealthy by starting baking apple pies for the gold rushers. The classic "sell the shovel" story we hear. And one that also tells me not to lament not being first.
I see people who chased that with crypto or whatever hot trend. None of them struck it rich and feel burned out and have little to show for it. But instead to focus on your values and delivering value. Explorers at the time were suffering scurvy from too much canned and salted food so a pie served not just the soul but the body. She didn't have to be there first even.
I also think that the person I am today might not have even had the skills or gotten into the industry if it were 5 or 10 years earlier. It's the mighty ducks "a half inch and it would have gone in" but a half inch in the other direction could have also completely whiffed it.
I wish that I was born a thousand years ago
I wish that I'd sail the darkened seas,
On a great big clipper ship,
Going from this land here to that,
In a sailor's suit and cap,
Away from the big city
Where a man can not be free
Of all of the evils of this town,
And of himself, and those around.
- The Velvet Underground, "Heroin"
i dont think those exist purely in software anymore- but if you look around at say, the skyrocketing number of elderly and nursing homes (just as an example) there are probably dozens of ways to improve lives and make a good living there
and electronics and hardware has always been so difficult and expensive to get into- but with new infra and workflows and 3d Printing and getting stuff from China- it's vastly cheaper and easier to do a robotics or hardware startup than 20 years ago
there will always be opportunities!
But it makes me happy in such an immediate serotonin spike it’s akin to a drug. It used to be that I’d sit there dreaming every night of new things I wanted to build. It was fun to think, “what if I devoted the next year to this idea?”, but that was always met with the realities of a 9-5 after.
Now I can build those ideas as fast as they hit me. They’re fleeting fancies though. A game here, an app there. It’s fun to see but there’s so much more digital waste.
It does make me happy though. I’d rather be able to live my dreams than let them pass me by. I do wonder about the addictiveness of it though - that hit it gives you is very, very real.
There is a name for this in sociology: anomie [1].
In times of great change, the set of norms and rules that we take for granted as an organizing principle of our lives get uprooted. It's as if all of the pieces and tokens and rules of the board game we played our whole lives are swept off the table to make room for a new game, but the new game has not been set up yet. In that interim, we all feel a sense of unease. It is our turn to move, but we don't know the rules of the game or what moves to make. We feel intense pressure to play the game well, but don't know how.
In the past, these transitions have ended and society has generally returned to some sense of stability. A new board game gets set up, new pieces doled out, we learn to play the new game. I hope we get to that point with AI soon, but the revolutions seem to keep coming faster and faster.
It's a psychologically difficult time.
One semester, I took Intro to Sociology and it was a revelation. I ended up completing all of the courses required to get a minor in sociology within a couple of years. I dropped out before I ever got close to finishing my CS major.
I wish more people knew sociology. It affects our lives more directly and profoundly than possibly any other science, but it's mostly unknown by the masses.
But I gather from reading this that you are young.
The time value of money and the changing nature of our industry tells you a different thing than it tells me, I think.
If you are young enough that you have any period of commanding multiple-hundred-thousand-dollar-per-year jobs: TAKE ONE OF THEM.
This is your opportunity to earn enough money to make you safer in the future. It will be a shitty miserable job but if you keep as much of the money as you possibly can, you give yourself a bit more security in the future (which is bleak, not positive). Just try to keep that in mind as you do a job you might not enjoy. Look for the things you do enjoy and try to keep your soul together.
Jobs paying that kind of money are going to disappear: salaries will have to plunge within two years. There is no way that salaries will stay stable on the other side of this transition. The tech industry is absolutely destroying itself.
(As to what it tells me: get used to cheaper food and think about maybe dying sooner.)
--Michael Caine
You cannot prompt Claude better than your customer can. There is no realistic moat there. You will not survive.
Funny you mention this. I've been building exactly that. gutsy [0] is a small (775MB) local decision model (like Jev). No API calls, runs on CPU. Classifying "does this reply make sense" is basically the use case. Platforms not doing this already isn't a capability gap at all.
As a programmer, I cannot let myself be sad for that. Because making other people jobs’ obsolete has been our bread and butter since the beginning of the field.
As a human being, I cannot watch myself in the mirror loathing about the consequences of stuff I have happily inflicted to so many others, without feeling cynical and egoistic.
I don’t know the answer regarding what to do, besides investing and hoping for policies that would make people standard of living mostly independent from their job *if* we all become redundant.
Just because the job becomes obsolete doesn't mean the employee has to. Some employees have adapted to different jobs and roles many times in their careers.
> I think I could have been a pretty formidable startup founder had I been born five or ten years earlier
I used to think this around 10 years ago. I eventually learned this kind of thinking is defeatist and figured if others are able to succeed in this environment then I can.
> there are only "AI startups", where you build a company that somebody else can write a prompt for in 30 seconds.
Just like how people think ideas are a dime a dozen, software will be. The value will be in polish, ongoing maintenance, feature development, and the incorporation of customer feedback.
> I can't think of a single piece of software I use that has gotten better
Users are notoriously bad at noticing small improvements over time. A more interesting thing way to see the impact of AI is to ask employees at big companies what it has allowed them to work on and release in a previously unimaginable time frame.
But that employee will be on a treadmill and has to reinvent themselves in an ever-increasing pace because the AI is continuously hunting for their job, and its ambitions are growing exponentially.
I notice a lot more people moving to a career where they manage, as opposed to do the programming. I see why.
I'm slowly phasing out of my software role, I don't even have that much experience yet. I just cannot sit behind a PC for 40 hrs a week talking to a chat bot. I would rather do something useful with the time i've been given.
But you could sit behind a PC for 40 hrs a week talking to a REPL (bash, python). Maybe the coding agent is also worthy, but differently.
You need a side project anyway to enjoy true software engineering and hacking. You can still do that with or without AI.
anyway i wrote a suno song about it the other day called BETTER THAN ME (I DON'T CARE) https://suno.com/s/6PdLBvAQwIGuyACr
I don't know how true this is. I've got a handful of vibe-coded projects that are quite interesting to me, but the project that is by far the most interesting to talk about (and most impressive) is the one that I did back in 2021 before I ever used an LLM.
The common wisdom in this industry is that ideas are cheap, and if that's still true, there's now a pretty high bar for a vibe-coded project to be considered impressive. You need to either execute to an unusually high level of polish, or be building in a space where even having an idea would require a high level of experience.
Of course, I might be wrong. Maybe all these vibe-coded projects really are recruitment fodder?
After recently pushing agentic apps to production, my impression is that agentic AI is here to stay and domain knowledge is more important than ever. Real expertise is required for context engineering and harness engineering. You have to think ten steps ahead now if you really want to turn stochastic output of an LLM into something reliable and flexible. Domain expertise is needed to identify workflows which were previously impractical but can now be achieved using the reasoning* capabilities of an LLM.
In my 20 year career I've seen trends come and go. Every few years, some skill upgrade was necessary, and AI is just another push towards that.
I have plenty of vibecoded projects. I didn't know it was that easy.
Dear: startups and FAANG: please send offers.
Not sure I'm gonna make it past paragraph one tbh.
It doesn't sound very different from when I was in undergrad an eon ago, except substitute "AI" with "blatant and rampant cheating" and "openAI offer" with "google" or "facebook"
That's so true!
They drink less, which has been documented extensively. The fact they are asking about intensity is a sign of this, not evidence to the contrary. This is a good thing, imo.
That was true a couple years ago, but now they are picking up. Maybe they were just broke.
I think that the inflation wave has caused cheap beer to bring its prices back to earth Dollar Menu-style. For a long while there was simply no cheap beer.
* But... after months (years?) of despair and anxiety, last few weeks something broke in me, possibly for the better. I am trying to focus on the here and now. I've bought a C64 and I plan to learn it in massive depth with my 8 year old son. I've started water-colouring, because in a world where AI perfection is cheap, I suddenly find value even in my extremely crappy but hard-earned scraps. I restarted playing guitar, poorly. I'm trying to stop myself, in my private life, from calibrating to YouTube Influencers and TikTok Content Creators and despairing that everything I could do, hundreds of people on the internet are doing better. I'm trying (it's a work in progress:) to focus on "What do I actually enjoy?" and do that, rather than focus on the runaway chain reaction our society is apparently hell-bound to follow.
Don't get me wrong, I'm still massively anxious not only about my on job, but also about my kids - they are 5 and 8 and I have NO idea anymore what job market will be like in 15-20 years, now that we are eradicating knowledge workers at massive pace. But now, the hand-made get-well card my cousin made for my wife's recovery, the hand-written piece of super-cool C64 code, and a cool but genuine and sincere clip somebody made in their backyard with a crappy camera are what I'm focusing on rather than massive, massive amount of AI generated stuff out there.
Working with AI tools the way many people seem to be doing it (multiple sessions at once, not even reading the output, rubber-stamping MRs) makes me anxious and paranoid, and it’s definitely NOT because I miss punching in if statements.
And while I don't go all in on the notion of revealed preferences, a few facts stand out to me:
1. Nothing is actually stopping me from writing code by hand for fun, but I don't
2. There is no practical reason to manually write homebrew SNES games in assembly language, but some people do
3. I still have no deficit of fun problems to dig into, to sit and ponder, and then experiment with. They just aren't coding problems, and I have far more powerful tools to help me navigate prior art, test out hypotheses, and so on.
I still wouldn't say im "happy" with the current state of things though, despite mostly fitting in that latter group. AI is still fairly tiring to interact with, and everything feels contested. If its not contested, it feels like putting almost anything out there... means it will become contested when certain people see it and decide "that is a solid idea", and just throw agents at mirroring it.
I keep thinking it would be interesting to do a series of interviews with software engineers on how they’re dealing with this change.
I also like using AI to make things
I am happy
There is more to software engineering than just "coding" or "building," and it's not even correct to oppose the two.
1. The joy of the process itself. If you're a programmer, it's writing code, running it, fixing bugs, making tests pass. If you're a woodworker, it's the feel of wood sliding under your fingers, a deft hand on a chisel.
2. The satisfaction of a completed work. For a programmer, that's an app you can run and click around and see the features up and running. For a woodworker, it's a sanded finished piece of furniture sitting there in the workshop.
3. Seeing the completed work make a difference in someone's life. For a programmer, that's a user telling you how your app helped them achieve a goal. For a woodworker, it's seeing that furniture in a room, watching it accumulate the patina of use.
So much of people talking past each other about AI comes from people who care a lot about 1 and people who never really cared about 1 but care about 2. It turns out there are many programmers who never really liked programming, they just wanted to have made an app. They are delighted about AI because it lets them skip a lot of step 1 and get right to 2.
But what I think is critically missing is that for most people, step 2 is marginal. It's step 3 where most of the real joy is. An app with zero users does not get you up in the morning. A piece of furniture collecting dust in an attic is not a source of meaning.
And the core problem with AI is that it gives you 2 by burning a whole lot of 3. Yes, you can whip up your "exercise tracker for cats" app in a tiny fraction of the time. So can a million other people. Your app is going to drown in a sea of other forgotten apps. Probably none of them will be maintained and none will reach critical mass.
I've done a lot of creative work over the years and time and again, as digital technology has lowered the barrier of entry, I've watched a horde of enthusiastic new people become disillusioned when it turns out they can make things now, but it's harder than ever to actually connect those things to people because everyone else can too.
I remember when DSLRs got a million people excited about being professional photographers, but now almost none are because photos are a mostly worthless commodity. I remember when CD mixers and Serato opened the doors to a million DJs who no longer had to buy expensive vinyl and struggle with beatmatching records. Now DJ mixes are a dime a dozen and it's harder than ever to be a working DJ. Recording an album used to require a room full of very expensive gear. Now anyone with a laptop and a DAW can do it. Publishing a novel required getting one of a small number of publishers to like it enough to print a thousand copies. Now anyone can upload a PDF to Amazon and share it with the world.
What happens when the barrier of entry gets radically lowered is that it requires some other extrinsic attribute to rise above the competition. In practice, that is either being rich and connected already, or getting extremely lucky. Domains where anyone can play have the steepest power law distributions. You have a few stars that everyone wishes they were and a million people doing it in their limited spare time. Unless you're already one of the elites, it's basically a lottery.
Now that's going to happen with software development.
I wonder what is the equivalent in 2026 of what Dave should do to be a human cosmonaut to go against the transformer VAE probability-based machine we call Claude.
It didn’t revolutionize everything but it did quietly take over. Almost every production language has some version of it and pretty much everybody uses it.
LLMs are changing rapidly so I won’t even try to speculate about where they’ll end up, but that’s how it looks now.
That is the problem, why OP is so sad, you spam with your great ideas taking up the oxygen around him. AI is absolutely amazing when you use it for your own ideas and projects, and terrible when others do the same, raising the bar and making your own efforts look smaller. Have you read the open letter written by Fields medalists? They complain "the collection of good, fruitful open problems is now being mined in a non-renewable fashion".
"Ah, I see you are trying to build an amateur rocket, but what do you think of the space elevator I just made?"
They will just prompt their own version into existence without any obvious heritage to yours.
I've wanted to build something rather specific for a long time.
With AI i'm close to release, but had AI not come along it would not have happened for at least another 5 or 10 years depending on what life throws in my way.
The current state of AI seems to be more about automation of well defined tasks rather than things like creativity and challenging design that bucks convention, or requires human sensibility, so I think there is still room to get satisfaction from that aspect of it (you do the design, AI does the coding).
People point out how computer chess hasn't killed human chess, but I don't see much satisfaction from doing any development work by hand that could equally well (at least as far as the resulting artifact goes) be done by one/few-shot prompting, even though part of the fun of being a developer is being able to conjure dreams into reality - the end result is part of the satisfaction, but the challenge in getting there also matters.
Design is still hard. People are very good at sniffing out shortcuts and inconsistencies and AI “smell”. The last 10% of designing a product is more like the last 400% (because you’ll probably need to throw it away or start over). It’s not spending time on design that cheapens the product
If you're building it to sell.. I don't see how this doesn't commodify most products if a competitor can just clone what you're selling with minimal effort
This is why SAAS is here to stay. Yes, I can one shot it, but then I've got to maintain it, and I want to get on and play with the next shiny thing!
Doing things for me, awesome, that works. Doing things that have an expectation from a third party, not so fun.
If you build something, sit with your users, listen, build community etc. I don't think that cheapens the final product.
It feels like a very very well done exhaustive PoC.
Like executing an Idea is a 1000 times more fun than just thinking it and rethinking it a few month later again and then typing a few things down, again, and then thinking about it again and then thinking "i should do some sabatical to work on that stuff" Just to talk to friends who did sabaticals and never done anything with their time and realizing that you wouldn't just do that just because you took some sabatical.
You're discovering that effort doesn't matter, only results.
I bought an old house and started renovating it. Countless hours of back breaking work to get things up to a modern standard of living. Yet bringing people over to see the finished result felt so hollow. There's no "gosh this must have been so hard, great work!". To them it is just a house, like any other. The incredible amount of effort on my part to get it to that point meant absolutely nothing. No one cares how much work you put into something. It ultimately only matters what something is or isn't, and competence is the expected baseline.
And even if there ends up being some kind of new future SWE work that AI is bad at, how long would that last? People aren’t realizing that AI isn’t just a way to replace current human work, it’s a framework to replace all future human work too. Wouldn’t companies be tripping over themselves to feed AI as much data on this new task as possible, so that they can replace those human workers ASAP?
I don’t see a future where humans do long term meaningful software work unless AI improvement plateaus quickly or there’s some kind of regulation slowing down future AI training.
A company, a product or job seeker the all need the same thing - to be picked - why would you pick me? That question is harder to answer today than in 2020. So it becomes a "musical chairs" moment. Everyone is panicking and scrambling for their own seat. This is how we get "new opportunities that didn't exist before" - we will strive to create new ways to differentiate ourselves. We could spend an enormous energy on that goal. But we can't predict what they will be no more than we can predict where the ball will be 1 minute later on the football field, it depends on what everyone else is doing.
I think this is very true but I'm not excited about it!
This feels like cope to me. So much of what "knowledge workers" do at all falls into a few buckets: we shuffle data around, we serve ads, or we provide tooling for those who do those things.
When the car came, the buggy drivers were screwed, but there were ultimately more cars (and hence work for people who supported them) than there could ever have been buggies.
What does AI give us "more" of, exactly? It makes everything simpler, and thus requiring fewer "knowledge workers" to plumb things in. After all, the need to shuffle data around is finite.
I actually think you still can. We're still in the "centaur chess" phase where a knowledgeable human sometimes needs to know the right things to ask for.
Will you be able to in 6 months? 2 years? I don't know, but it's not looking great.
I don't think the person you are replying to meant everyone won't survive as of literally today.
We're still in a somewhat calm period where something fundamental changed in the world but it changed so fast most people have barely reacted to it as of yet so things feel relatively normal, but with a growing sense that they aren't.
In the beginning of 2025 if anyone suggested "You cannot prompt Claude better than your customer can. There is no realistic moat there. You will not survive" most people on HN would have laughed at it, even the people who were considered AI-pilled for the time.
Today that sort of claim may spark some debate and some people are clinging to denial, but nobody who actually uses this stuff is laughing it off anymore. Even the staunchest deniers recognize there is enough to be concerned about there that it isn't a joke anymore.
Steve Jobs built an empire around the idea that the customer is generally terrible at specifying a particular solution.
A well designed, well built physical product has a far higher barrier to entry
Even if your customers won't do it, there are endless people who are willing to compete with you if they see you making any money
When iteration had a real costs, you couldn't afford too many mistakes.
That is simply and obviously not true. Technical expertise is still highly relevant when prompting, reviewing, understanding consequences, and approving.
The main reason Claude is "cheap" because it has to compete with OpenAI, Deepseek, GLM, Qwen, Muse (who is also massively subsidised their model to get training data), etc. If they increased prices 10x, people would instantly jump ship to the next competitor. There's no moat for Anthropic either, short of constantly training new models in a red queen's race.
> That's why Claude is so cheap for what it does.
That's probably the competition, and
> You cannot prompt Claude better than your customer can.
You may currently have enough technical knowledge about computers and services to collaborate with a customer better than either of you working alone.
Thinking back to a previous discussion, a veterinary clinic likely does know better than you what they need in their CRM, but they may wildly overestimate the hardware required to do that and just accept the first thing Claude comes up with that's scaled for a city hospital with 4000 beds occupied at any given time, when they're actually just one person in a village with 100 pets.
Or the other way around, which is even worse.
Right now, experienced engineers can help with this kind of problem. I don't know how long this will remain true, but right now it seems to be. I think.
I fully acknowledge this is a mess right now, but vibe-fixing or vibe-rebuiliding-from-scratch is only going to get better.
More realistically, I think other companies will vibe code and sell these solutions, but the cheapness of building will drive down costs so much that it'll be effectively the same result.
But yeah the majority of startups don't have well thought out reasoning either. In the end, there's no substitute for hard work. AI can reduce the iteration time maybe.
I suspect we'll increasingly see vibe coding targeting assembly language. If humans aren't going to read or write the code then using higher level languages becomes kind of pointless, except for being able to use the toolchains.
Now I derive no satisfaction from anything 'I' have built with it. At some point I am just stuck asking myself "Why am I doing this if someone can prompt an LLM the same way I have and get a similar result?". I can have the LLM write documentation, only to think "who am 'I' writing this for?". Why bother if someone isnt even going to read it, its just going to become context for someone else's LLM session.
I know there are other ways to look at but this is where I am at currently.
Many people would choose not to be a plumber for various reasons. A shocking number of people seem to deem manual work as beneath them. Many others just prefer to do other work.
I have no problem doing plumbing or building. I've installed floors, ceiling fixtures, drywall, whatever. But I am best at software and would prefer to do that for my career.
Product design is mostly saying no to ideas. And occasionally yes, but only to a very select cohesive set.
Most people are terrible at that - though I think it's learnable to some extend. And at least current-day agents will generally just do whatever dumb shit the human says it wants.
Beware of survivor bias though.