It's the five layers of product growth between you and the machine that get tiring.
This has always been the hardest part, its just the past few years its gotten exponentially more difficult.
> The hard part now is liking the industry around it.
But now, the industry really kills that passion. I don't believe we're solving real problems, but mostly just solving made up problems that get us money. We've become incredibly dismissive of fixing things and using thought terminating cliques like "I only care that it works"[0] or "don't let perfection be the enemy of good enough"[1]. When finding bugs we end up arguing if it is "valuable"[2] rather than if it actually helps people. I can't tell you how many meetings I've been in discussing if we should fix the problem that were a magnitude more time, per person, than it takes to fix the problem. We've become allergic to deep understanding. We abhor expertise[4]! I thought this was supposed to be a community of nerds? I came to the STEM side because my family was all "business monkeys" and I didn't want any part of that constant BS. There was a real "revenge of the nerds", but the MBAs did strike back.
We're incredibly penny-wise and pound-foolish. We love our sayings, but hate understanding. All to keep that velocity up, but forgot that velocity has direction.
What happened to the time where we could make money AND make meaningful products that make peoples' lives better? (I know, rose colored glasses) That's the real dream, right? That's what we want *an economy* doing, right? Not this bullshit metric hacking. Not this maximizer with complete disregard for the things we're intending to maximize[5]. And for some reason we still look for "passion" in people when hiring, only to beat it out of them when they get hired. And how have we gotten to a point where someone's resume that has "VR + Crypto + AI" is read as impressive rather than the hype chasing giant red flag that it is?
I don't think this is just about computing. It's a bigger cultural phenomena. But without a doubt our field became perverted by whatever this infection is. There's a reason so many are burnt out. The disease creates a negative feedback loop too. We get burnt out, end up just going with the bullshit (tired of fighting), which only creates more bullshit. The feedback has been going on for quite some time now.
I don't know how we push back, but I know I'm not the only one frustrated, and I know if we don't figure out how to make changes soon then we shouldn't be surprised if change happens in an unpredictable and chaotic manner. That steam can only build for so long.
[0] Everyone does you asshole! What we disagree on is if it actually "works"![0.5]
[0.5] https://x.com/mitchellh/status/2066657032938442833
[1] Perfection doesn't exist! Only an idiot thinks there's perfect code. What we disagree on is what is actually "good enough"!
[2] $$$$ not "does this make the product[3] better"
[3] God fuck... https://www.youtube.com/watch?v=YZFTaEenaHM ----> https://www.youtube.com/watch?v=_NeJ3Kg6OUo
[4] https://x.com/yacinemtb/status/1836415592162554121
[4.5] Yes, this is the same guy who said QM and Fluid Dynamics are all easy, "its all einops" (I can, with 100% confidence, tell you that it is not. There's entire subfields of mathematics being lost here that are critical to both these subjects. You can't even get through Griffith's with just Linear Algebra and that's barely scratching the surface of QM!) https://x.com/yacinemtb/status/1836428078999851515
[5] A metric is never perfectly aligned with what we intend to measure https://9gag.com/gag/aQ9GEZ7I still write code by hand. But LLMs have been a legitimately useful tool when I've wanted to dig into a new field like computer graphics, theoretical physics, or numerical analysis. Or even just asking the LLM to write a piece of code and learning from its output. I think it makes me a better programmer because I can bootstrap the knowledge needed for a new project much faster and spend more time programming.
In my opinion you should interpret the usage of "AI" here to mean "the entire business/management/financial/bubble ecosystem surrounding LLMs". The snake oil is much more how LLMs are being weaponized and utilized rather than a specific technical assessment (although that often is an issue too)
Depends how long the RAM correction takes. It is interesting how RAM prices have stifled the creation of cheap laptops capable of running big models. But at the same time, this seems like a second order effect and not the intention.
There's also still the risk of the creation of a new economic underclass, if both a) hardware remains too expensive for local inference and b) subscription or pay-per-token based inference also remains expensive or increases in price, then individuals will largely be locked out of the benefits that having access to AI could bring, leaving it purely in the hands of larger companies. People will only get to use and experience these tools through their employer, for the benefit of their employer.
Choose one:
- You spend 30 hours writing a program to manage data for your hobby. You write it on your personal computer.
- You spend one hour generating a program to manage data for your hobby. You have to lease an H200 behind an API to do it.
Which one will you choose?
I know which one I'm choosing.
I know that many others choose A as well.
A wonderful service known as the web has connected people who choose A with others who choose A and of course with a great many who don’t need to make a choice and benefit from the work of others.
I mourn a world in which few will choose A, because for many to choose B seems to lock us all, tragedy of the commons style, into a worse world.
I swear the anti-AI crowd would all be picking to die if you each had a choice between immortality and living to 85.
This all feels so damned performative. These are irrational decisions.
AI is better at this than you. You just won't admit it. And it's going to get 10,000x better than you in just a short while.
Programming computers is a fad. It's an anachronistic relic.
None of you is writing punch card programs.
None of you are building vaccum tube logic.
None of the things we build today are going to last. Your programs will be meaningless in a hundred years. Probably closer to ten years.
Programs and code and programming languages are as ephemeral as social media.
Get over it. It's not that important.
> AI is better at this than you. You just won't admit it. And it's going to get 10,000x better than you in just a short while.
Where is the 10x (not even 10,000x) revenue? No companies other than those selling the AI itself are seeing it.
It really depends on the cost of immortality. At the very least, it would have a psychological impact that some people may feel is undesirable.
> None of you is writing punch card programs.
> None of you are building vaccum tube logic.
Perhaps none of us, but some people certainly do. We are intellectual creatures. Some of us do things out of pure curiosity. Can we create multinational corporations out of it? Almost certainly not. Can we create businesses out of it? People do so all of the time. There is a market for produce from small farms, hand crafts, heck, even vintage computing.
> None of the things we build today are going to last. Your programs will be meaningless in a hundred years. Probably closer to ten years.
Try telling that to people who are trying to retire legacy systems. Sure, most of them have been modernized. Perhaps they have even been modernized to the point where none of the original code exists. Yet the core ideas still exist since it turns out to be incredibly hard to discard things.
The old ways of writing software will continue, even if they are nowhere near as popular. Call that irrational if you want. I call it human.
Exactly, so why do you care how some people build things? It's not that important.
Maybe so, but I am better at this when I engage with AI in a controlled manner.
It's just not though. Plagiarizing some shit it stole off github does not make it intelligent.
Edit: just because it's amusing, here's something I'm literally running into right now with Opus 4.8 on "Max" settings being dumb. I asked it to add some C++ code to an existing C++ project for Unreal Engine. It did half the work, then balked, because "it doesn't have a way to compile C++". I just had to tell it "yes, actually, you just need to run the fricking extremely-standard-already-generated-build-script." If I had a novel build system it'd be one thing, but it already knows I'm working on an Unreal Engine project and that the build is completely standard and it still couldn't piece together that it could just run the compiler.
I would not employ a C++ developer that could not figure out how to invoke the compiler.
I don't think you understand what code is. What it does is far less important than how it does it.
Software is bureaucracy and always has been. The discipline is just finally maturing into this role like so many other careers have.
Data centers as infrastructure are very different from DSL rollout though. Much, much more expensive to maintain, with a much much shorter timespan.
If the bubble pops and data centers get shut down because there’s no one to pay the bills, there won’t be much left 5-10 years later in terms of infrastructure.
"Snake oil" refers to something sold as a medicament that has no beneficial effect.
It’s hard for me to think of any piece of new tech that hasn’t been over hyped by the people selling it.
This is how it looks in your head, maybe. But in reality since Sonnet 3.5 - when the whole "no need programmers" started - no "years" have passed. Sonnet 3.5 came out on June 20, 2024. We are still 5 days away from the lowest possible "years". So even if you quoted them literally, they could not have possibly proved themselves wrong yet even once, let alone "time and time again".
Not to worry, a fool and his money will soon be parted.
I think the main problem is that there is no definition for "AI".
And from your use case, I don't see any difference between that and a search engine.
faster != better
Oh, not using it right? Not the right model? Insert coin to continue.
Snake oil, total snake oil.
If you don’t like the answer, don’t worry, they’re building more data centers in poor neighborhoods so you can keep submitting the prompt until you get a better one.
AI is something not a colleague (a slot machine), sold as a colleague.
> The comparison suggests that the thing doesn't do what it's marketed to do.
What AI is being sold as is incredibly different than what it actually does. I love AI. I spent years in grad school researching it because I loved it so much (it was never about the money to me). But what it is and what it can do is so different from what it is being sold as.
Snake Oil is an apt comparison because it is being sold as a cure-all. Medical problems? AI. Financial problems? AI. Scientific research? AI. <Insert problem>: AI... It isn't that ML[0] can't help with these problems (it can!), it is that "AI" is being sold as a solution to these problems. As if humans will be obsolete in 6months[1].
LLMs are a fantastic example. We (lossy) compressed the entire internet and build a human language interface into it. That's some real Sci-Fi shit right there. That's an incredible achievement with a lot of utility! But how is it sold? If you call it what it is people will act like you're diminishing its status. We've exaggerated the accomplishments so far out of proportion that we can't even recognize big of an advancement that these machines actually were. LLMs were a huge step forward, but even a giant is small when you compare it to a titan.
So yeah, I do think it is being sold as Snake Oil. And that's been my fear for quite some time (you can dig up my history if you're that passionate). But that's also what we've done with every major tech recently. Hell, even cryptocurrency has real value. The thing that killed it was all the hype built around it when the tech was just in its infancy. Do we really want to do the same thing to AI? It certainly has more utility to it than cryptocurrencies. But it doesn't matter how good the actual product is if people are sold on something else. What matters is how the actual product matches to peoples' expectations. There is such a thing as "overselling", and we're certainly doing that as a community. I know it is an exciting field and there's lots of exciting technology, but we can't promise the moon if we can't deliver.
[0] It wasn't long ago that "AI" was a red flag and "ML" was seen as less likely to be bullshit.
[1] I'm still waiting on my self-driving car...Nowadays people just say “AI” when they mean “LLM,” which is an unrelated thing entirely, but people want people who use it.
Well yeah, because it doesn't. AI is being claimed to be a magical genius intelligence which will solve everything forever, but in reality LLMs are still idiots you can't trust to not screw up without a tight leash. They can't even do the one thing they are supposed to be good at (programming) well, despite all the effort which has been focused on trying to make them good at it. They don't remotely do what they are marketed to do, not even close!
You sound like you have a quote in mind.
> But things feel different now. I can relate to what Chris Person said when he expressed his frustrations about how these slick conmen are using the technology I adore as tools for exploitation and disempowerment. The Internet, built by idealists on a foundation of openness and community, has become a mire of dark patterns and gardens with ever thicker walls, desperate to keep people within an ecosystem where their attention is the prized commodity. I’ve witnessed a nerdy space full of nerds be invaded by marketers, callous capitalists, and “brogrammers”—exaggerating the worst, most toxic, aspects of geek culture in their pursuit of money and power. I’ve poured hundreds of hours of work into open source projects only to have it all be scraped into a plagiarism machine and then aggressively sold back to me. It feels that the hope I had for the future technology could give us, the naïve and starry-eyed fantasies I fostered in my youth, has been eroded when faced with a reality where the thing I love can make a lot of money for people who don’t care for any of it.
You can simultaneously believe that AI is really cool and also that also a lot of companies are degrading the internet, society, and private ownership at large.
Meanwhile, the economy needs software to be written and I need employment, and I'm lucky enough to have a job that hews somewhat close to my interests, whether that be learning the latest JS framework or to prompt Claude. It's all pretty decent and better than chiselling coal out of a pit for 10 hours a day.
This line hits really close to home. :-/
I also love the Computer AND the Internet, allowing me to interact with and experience the world like this.
There's no gate being kept here. It's just someone talking about something they are passionate about and then expressing their opinion about current technology. They're not stopping anyone from also loving the computer. If anything, I could see someone reading this and being MORE excited to get into computing.
You cant really have 90's computing again very easily. Its not as simple as putting a DVD on play. It was a zeitgeist, a vibe.
> Long time fans of a thing are
I definitely think we should be welcoming to new members in any community, but I also think new members should recognize that they're coming into an established space. Not everyone is exactly equal. Not all gatekeeping is bad. Or rather, maybe it is better that we have people that help newbies get involved.
I'm not saying the protection can't get toxic. It definitely can. I was part of the Arch forums awhile ago when there was the push to kill the noob guide (I was pro-noob guide and was a frequent editor). We lost that battle, but hey, it was likely part of the reason we got a bunch more Arch forks like Endeavour and Cachy.
I always want more people to enjoy the things I love. It's great to share and life is so much better with friends. But it is also only natural to get emotional when you're losing that part of your life too. One big problem with big communities is that they become anonymous. Take HN for example. There's a handful of users I recognize, but it is for the most part effectively anonymous. And we're relatively small. The thing I miss the most is small communities, since that's where you get to know people. I think from a broader perspective we've done a great job at destroying those. There's got to be a better balance than what we have now.
Tony Gilroy is not!
People with impure motives are running the show now, and they're the number one reason why technology sucks now. Everything is dumbed down, enshittified, growth hacked, attention grabbing, surveillance capitalism, information brokerage, advertising. It was never about the computers, it was always about money. The computers are not only unloved, they are actively being hidden from the user. The computer does not matter, only the "user experience" matters. Users are not meant to use computers anymore, they are meant to interact with the scripted flows created by corporations, and nothing more. Computers are the enemy in this world, they are too powerful, leave them unchecked and they can wipe out the business models of corporations.
I definitely think someone who genuinely loves computers should have a say in how computers are used. That's the exact sort of person I'd elect as a regulator or lawmaker. There is no substitute for giving a shit. Someone who cares about computers more than money wouldn't have let things degenerate to this point.
Even if it is gatekeeping, what does it matter? It's not a problem. Computers are world changing technology, and they don't deserve to be reduced to vehicles for corporate advertising. They could be more, and should be more. And if gatekeeping is necessary, then so be it.
If someone's gonna gatekeep, their shit needs to smell like roses.
N... no? The guy's pretty clearly anti "attention economy bullshit", anti plagiarism-machine, and of the opinion that it is a very good thing that it is easier than it was when he was young for people to learn to love 'the computer'.
From the end of the "The Smell of Ink on Cheap Paper" section:
> It would be easy to say that it’s just nostalgia that makes me lament what was lost in the transition to the Internet, and it’s not like print was spared the rot of capitalism that has made online geek spaces into ad-ridden, engagement-maximising cesspools. But I am glad that I was able to do my initial discovery in a world devoid of pop-ups, auto-playing ads, click-bait, and incessant pleas to “like and subscribe”. ...
From the final section, entitled "A Life Well Lived?":
> I can relate to what Chris Person said when he expressed his frustrations about how these slick conmen are using the technology I adore as tools for exploitation and disempowerment. The Internet, built by idealists on a foundation of openness and community, has become a mire of dark patterns and gardens with ever thicker walls, desperate to keep people within an ecosystem where their attention is the prized commodity. I’ve witnessed a nerdy space full of nerds be invaded by marketers, callous capitalists, and “brogrammers”—exaggerating the worst, most toxic, aspects of geek culture in their pursuit of money and power. I’ve poured hundreds of hours of work into open source projects only to have it all be scraped into a plagiarism machine and then aggressively sold back to me. It feels that the hope I had for the future technology could give us, the naïve and starry-eyed fantasies I fostered in my youth, has been eroded when faced with a reality where the thing I love can make a lot of money for people who don’t care for any of it.
> Then again, it’s not all bad. We’ve come a long way from the time when computers were seen as expensive and exclusive tools, and the unwelcoming domain of elitist men. Programming—with the empowerment that it brings—is more accessible than ever and there seems to be a strong cultural shift in the techie spaces away from centralised services and onto federated, self-hosted, and in many other ways more free alternatives. The Internet seems to becoming more and more locked down, but us weirdos will just stay in our weird corners and will find means to circumvent any restrictions put on us. My affection towards technology made me an ostracised outsider when I was younger, then it condescendingly made me into a “rockstar”, and now it’s looking like my peers are ushering in the end of civilised society. So I’m ready to go back into being just some strange guy with strange interests, doing silly things people don’t understand and don’t care to.
Perhaps Sir should go back and re-read the article without looking for hidden meaning?
Only later I learned that this, ehat I was missing in the analog world back then, was what made up the core of a turing machine.
So this beauty is still the same for me. Just the sheer amount of state that is available and provided by others make the concept much less powerful than back in the day.
Now AI brings that back a bit, by finding the right items that you can keep and iterate on. But we tend to let AI also do rhe iteration and that introduces that non-deterministic character that the computer had overcome.
So, no wonder, that I, and quite a few others, at the moment, still mostly use AI for finding and typing code to describe the structures for the machine, but keep trying to define the iterations ourselves, guaranteeing clear insight and access to the state we are trying to work with.
Everything else is more like working with an assistant back then (or today), extending your actionable potential instean of your mind. And depending on how you see the world or what the tasl at hand might be, you might prefer one over the other, control over action, insight and perspective over tinkering with the matter to push it somehow in the right direction or implementing a known process.
But that state thing, still priceless, timeless. The right augmentation to our fuzzy brains and better than paper, since, who thought, we can express the algorithms in the same way.
So, I guess I will always see this beauty, even in a simple flip-flop, coin, switch and any array thereof, and anything that can be controlled by that binary configuration.
Nothing will take that away. I have plenty of evidence for it, too.
I think those of us who truly love the mechanism will not only survive the AI wave but will surf that motherfucker because WE KNOW, AT A VERY DEEP LEVEL, THE THING FROM WHENCE IT CAME.
(and honestly, I’ve been waiting for something like this to help filter out all the fucking posers diluting my market value)
then at some point, the balance started tipping.
Things I specifically remember:
- half life required you to register online to install your single-player game from CD
- turbotax would only install on one computer.
- software started rooting around on your system and uploading the findings
- the iphone didn't let you install your own software
- microsoft. enough said.
AI local rigs feel like the pc did in 1980.
Imo 90s sci-fi had a lot of this techno optimism, even if the settings were often dystopian. Cyberpunk was pretty wild and who knows a lot of that could still come to pass.
Now it's the opposite. The computers belong to the corporations. They're just generously allowing us to use them, and only in ways that they profit off of. Computers, arguably humanity's most important invention, reduced to interactive television, a mere vehicle for corporate advertsing and surveillance capitalism.
It's truly disgusting. Makes me sad like nothing else.
I get the way he feels. I remember how special this stuff used to be because of how niche it was. It does feel a bit like the normies co-opted it but that is my personal and selfish view.
This is something LLMs took away from me. I can’t just look at the source code and figure out why a prompt didn’t produce the expected outcome. I have to go with my gut feeling, and with the little I know about LLMs.
On the other hand, LLMs have enabled me to code prototypes that I would have only dreamed about a few years ago.
Do you want your own fancy terminal emulator? Done. A couple of weekends’ worth of work.
How about your own Linux windowing system, running Firefox and a terminal? Done. A couple more weekends.
You always hated KiCad routing, but never had time to go through the code and change it to meet your requirements? No worries. A day’s work.
Of course, none of this is production quality, but it gets you started very fast. And I’m sure you can turn it into a solid, production-quality product in much less time than it would take without using an LLM.
I remember when I was around 10 and we got out first PC - Compaq Presario - that we shared among us 4 siblings. And I was instantly hooked to. And then about a year later, we got internet connected and the first website we visited was Pokemon.
I remember at my high school, the computer room in the library was fitted out with the new colored iMacs. I was shocked! How could a computer look like this. You had to register to use it each day during lunch breaks because so many people wanted to use them.
I remember the first time I came across an Apple magazine, and it was showing screenshots of the new OS X. The Aqua interface got me hooked. I'd read, and re-read, every page, drooling over the screenshots. It wasn't until ~10 years later I got my first Mac and I was obsessed with it!
But this day, I dabble with OpenBSD and Linux (Alpine) and it’s a bit of fresh air. There’s some convenience lost, but you get the freedom of computing back.