> I've seen people supposedly smarter than I advocate for just giving in, conceding to AI coding as it's the future. But doing so means tossing out my friends who make art or the people who work their asses off to properly test and review code or the writers pouring all of their energy into even mundane dialogue. It means throwing out my dignity
Conflating things in this inflammatory way is a big mistake. Using a technology employers want you to use is not betraying your friends. Not everything has to be a culture war front.
The author is coming from the games industry. As an arts based industry, AI is EXTREMELY divisive. The author really will lose friends over even so much as touching AI in certain ways, because the artists that built the games industry were already badly abused, and now, they're being squeezed out entirely. For business developers, AI is somewhat less existentially terrifying, as it can be seen to be really empowering to an experienced user.
In the games industry, AI usage immediately eliminates a human job. Why pay a pixel artist if AI can generate 100's of unique little people pixels in seconds, and output them in the right format? Hollywood is going through the same thing: the companies that are building AI for Hollywood have to do so in the bushes, hiding. You don't see them advertising or flashing cash. That's because no one involved in using their wares wants anyone to know they're using them, lest they alienate the highly-talented people they still need to fill the gap between concept AI and full theatrical release AI.
In the software world, we are worried about AI. In the creative industries, they are absolutely pants on fire, screaming at the sky, burning down the village terrified of AI.
Everyone in the games industry is badly abused. Mostly because so many want to go into it. And yet game publishers continue to struggle with profitability because budgets have ballooned. I honestly think there's just not enough room in the market for the number of game studios we have today, at least not unless management improves and gets costs under control.
For one small idea, we probably don't really need Ron Perlman, Keanu Reeves, Kevin Spacey, Kit Harrington, Gary Oldman, Kiefer Sutherland, Willem Dafoe, Peter Dinklage, Samuel L. Jackon and Ray Liotta doing voice acting.
A friend of mine is an artist working in the games industry. So far nobody has been fired. They all simply started using AI to be more productive.
They care more about doing their job well than some artistic ideal that only works in practice if you don't have to care about things like food and shelter.
> Why pay a pixel artist if AI can generate 100's of unique little people pixels in seconds
Because someone who knows something about pixels needs to make a judgement. It is rare to see a machine-generated artifact (picture, video, text or code) that's good on the first try. And not always a non-specialist can see the issue.
Same thing why we still need human software engineers, even though a machine can generate code. Someone with actual understanding of the problem needs to make a management decision. Just like engineers see code slop (design or implementation) that laypeople vibe-coding don't recognize, artists see the visual slop where layman eye glances over.
Honestly, IMHO, this whole panic is artists' own creation. Instead of educating others on how to spot the issues (and thus reaffirming that expertise still matters - nothing had changed, and probably nothing ever will), a notable fraction went all-in on neo-luddite ideas, as if they don't know the history of their own craft and adjacent creative industries (I guess many really don't, or at least it doesn't click). Evaluate new tools, make use of them when they provide value, skip them where they fall short, and most importantly reaffirm that fancier brushes don't replace the artist in a human society - this is an already well-tested and proven strategy. Ring the existential bells when we'll get to the question of machine cognition rather than just intelligence.
Same for the engineering. Don't shy away from new tools, use them where they're a good fit, don't waste time when they are't (but periodically check out if something changes), explain everyone why you still matter - just to push back on unfortunate misconceptions.
The fact that a lot of companies' upper management went delusional and decided they want to replace humans witch machines (as if don't need responsibility anymore) doesn't help. But - hey - already plenty of stories how it bites them back, so while this period sucks, it's not exactly fucked, just in a state of (a pretty much expected) confusion.
Dishing out pixels or lines of code got somewhat cheaper. Expertise cost remains the same, though.
>Because someone who knows something about pixels needs to make a judgement. It is rare to see a machine-generated artifact (picture, video, text or code) that's good on the first try.
Not if nobody cares for the end product that finely.
And even if it was true, one person can make the judgement, while automation erasing 5 others that would have worked in both the judgement and the graphics wrok.
> Not if nobody cares for the end product that finely.
Yes. But a non-negligible number of people do care. If nobody would've cared, we wouldn't have this drama.
> one person can make the judgement, while automation erasing 5 others that would have worked in both the judgement and the graphics wrok.
Yes. In other words, new tools had increased people's performance for mechanical work - individual units of that can be arguably done faster than before. So hypothetically one person can do a work of five. Note of that erases those people and their skills. The real underlying issue is that demand for it doesn't catch up because world's coincidentally fucked (through a series of unrelated issues, such as a quite few global conflicts) and economies aren't exactly thriving outside of a few niches du jour.
I strongly suspect that if we would've had a flourishing economies around the world, the demand would ramp up and artists (and engineers, and writers, and everyone else whose performance could've been positively affected by new tools) would be in greater position than ever before.
>So hypothetically one person can do a work of five. Note of that erases those people and their skills.
It just doesn't magically take them out of the universe or turns them into unskilled persons.
But it does magically erases those people with those skills as needed employees.
>The real underlying issue is that demand for it doesn't catch up because world's coincidentally fucked (through a series of unrelated issues, such as a quite few global conflicts) and economies aren't exactly thriving outside of a few niches du jour.
Why would that demand have to "catch up"? Just because we can do something faster or automate parts of it doesn't guarantee demand will go up, even in a good economy. No shortage of jobs that vanished forever in a similar even, despite the economy going otherwise up. Even more so now, where it's fucked up anyway, of course.
> Just because we can do something faster or automate parts of it doesn't guarantee demand will go up
It's not guaranteed, but why wouldn't it? It's digital entertainment and art, and the market is not saturated, there's still tons of ideas unexplored. If more people would have spare money they can spend on digital purchases instead of making ends meet - I'm pretty sure there'd be more people seeking ways to earn that money. Especially if production just got cheaper.
Correct me if I'm wrong, please, but every time production got cheaper for something that isn't obsolete (and art and stories are never obsolete), the markets became more and more flooded.
>It's not guaranteed, but why wouldn't it? It's digital entertainment and art, and the market is not saturated, there's still tons of ideas unexplored.
Because there's only so much "digital entertainment and art" people practically care for, and already people were saturated and bored with all the crap out there pre-AI too.
I surely can be, but I don’t see it so far. How the latter suggests the former? The demand doesn’t have to scale infinitely - I don’t believe the demand is or was anywhere near saturation.
Tons of music that nobody listened to is certainly nothing new. Predates “AI”, and even Internet for sure. What would be indicative are shifts in distribution of listeners outside of extremities (excluding top hits and low audience both), but I’m struggling to formulate a good criteria to watch out for (and then we probably don’t have the data). If not a distribution, maybe a dynamic of total listening hours over time adjusted for income and inflation fluctuations may suggest if there was ceiling or not.
There’s something to the series induced boredom, though. Increase in quantity paired with a feeling there’s not much to actually watch is something I can relate to. I’m not sure how to tell oversaturation fatigue from actual lack of interesting (at specific moments) media. Thanks, that’s something to think about.
What gets me is that LLM writing has an inhuman voice. Of course we know the tells. Not only this, but that. You're exactly right! —
The woman and man AI voice endemic to YouTube and ilk is also tremendously off putting. M5Stack has a bunch of these videos, and it devalues what they're doing.
And then... Transformer "art". It is some of the worst drek I've seen. I smell it a mile away. It's easily seen by slop english-like characters. Or too glowy humans. Or overall fake feel. For pixel art, I can perhaps see it. But for anything it just feels... Gross.
I'm completely sure management LOVES it cause its cheap and devalues humans.
> And then... Transformer "art". It is some of the worst drek I've seen. I smell it a mile away.
Hating that every third or fourth ad now is AI-generated. So much worse than what entry-level graphic designers can put out while putting them out of work.
> I'm completely sure management LOVES it cause its cheap and devalues humans.
And the tragic bit is that instead of educating them about the pitfalls so everyone's on the same page, a lot of brave hotheads are literally calling for a class war
> educating them about the pitfalls so everyone's on the same page
What? They know what they're doing, they have access to the same information we do. There's just too much money in it for them to care about irrelevant things like product quality or the continued employment of anyone but themselves. Their prosperity and usefulness in their position is fully reliant on them not being on the same page. Their incentives are completely different from ours.
> They know what they're doing, they have access to the same information we do.
Who? I'm not talking about a few well-known sociopaths that run few well-known transnational corporations and have means to pump and squeeze biggest markets for profit (though I doubt they really know what they're doing). I'm talking about your casual folks in management roles in tons of companies of all sizes out there. Our peers. Just like us, on average, they, put frankly, don't know shit.
There is no "here's how we cook now, when new tools are invented" vetted guidebook anywhere. Can't be - no one had a head start long enough to prove anything. Almost everyone out there are more or less blind kittens from the same litter, applying some heuristics to poke around to see what works and what doesn't, learning what's not to do as it bites us in the ass.
Always been like that - humans learn from their failures. Most - we do during our training and experimentation days (which directly translated into our professional knowledge of things to be avoided), but this time we got caught with our pants down, when many say we gotta try to find something ourselves, or lose out to those who tried and found something of value, and there's something to it.
One can argue the writing was on the wall for years, but ask around and be surprised how many people actually had time and mental capacity to do the predictions, and if they had - how accurate those predictions were. In my bubble, most people are still trying to make sense of things past the very basics. Slightly past "all things AI now", onto "okay, for real, how do we learn to use this where it's a benefit, and don't use this where it's a liability" stage.
Even if you potentially have access to the information (others' experiences and discoveries), you gotta dig it up from all the surrounding noise. Figure out how to validate it. Actually validate it. That's a ton of unplanned effort.
Spend some time learning about the concept of the excluded middle[1]. It is something that will be useful on your life going forward, it might even save you a lot of trouble.
I am writing a post about that for my blog but it's still not there so ... check back later? Lol.
Anyway, in this particular scenario, a co-worker not being my friend does not mean we are enemies.
This is seemingly spoken from an ignorant and insulated position. The victims of invasion don't get to decide whether or not they live on a war front, nor do the countless skilled and creative individuals losing their entire careers almost overnight.
By the way it's class war you're talking about, not culture war.
Just to +1 the sentiment here; AI in the gaming industry is absolutely a culture war front. Especially in the indie space, you will lose advertising deals with many sponsors & influencers, for example, if you appear to endorse AI. Probably won't be that way forever, but it's a hotbed right now.
Meanwhile, Asian studios are much more open to experimenting with AI workflows, and will eventually build the next generation of Niantics and AppLovins if this neo-ludditism continues.
Heck, 50% of all Japanese game studios [0] along with all of China's largest studios [1] now use AI within their development pipeline - often with explicit state backing.
You may not like Blizzard or Ubisoft but Tencent, Sony, miHoYo, and even Nintendo are much worse from a work culture, compensation, and work expectation perspective.
> Why should that be a relevant concern? It's not like any of us will be working in game dev if AI wins.
Becuase AI is not taking all jobs. Yes a lot of redundant work will go away, but there is still a real need for human intervention, monkeypatching, and ingenuity.
The North American gaming industry only exists because the entire ecosystem from AdTech to Engines to Marketplaces exists to develop, finance, and distribute IP.
If you stay frozen in the past, you eventually get outcompeted and the ecosystem will leave. And unlike the automotive industry, game devs aren't a core voting bloc.
This is what happened to the entire animation industry and is what is happening to the film and television industry. Gaming will be the next IP driven industry to leave if everyone remains frozen and opposed to innovation.
> Maybe (sic) Chines sell tools for artists pitching them as useful rather then "haha this will make you obsolete you looser look at slop I made" middle finger pitch?
The obsolence and cost saving message is true though and used all over Asia - even in China [0][1].
Either you innovate and compete, or you will get trounced. THIS is the cultural mindset back in Asia.
I dunno, maybe OpenAI and Anthromorphic should learn from Chinese companies how to not act in ridiculously off-putting ways?
Maybe Chines sell tools for artists pitching them as useful rather then "haha this will make you obsolete you looser look at slop I made" middle finger pitch?
Not sure the technology is the issue here. He isn't debating using Java v C#, and one is a betrayal of all that is holy, so he refuses to use it.
I think everyone kind of feels that AI is sucking up all the content that we have all created collectively, and we all know that the bell is tolling for thee, no matter how much you adapt. So if you see friends being fed into a meat grinder, you can have a 'culture war' take on it. He's posting on line to vent, something everyone is venting about.
Of course, for a job, to get a paycheck, we'd use any technology, even if we are the ones running the meat grinder.
I think you might not get the context this is coming from, it’s about a game developer laid off from Blizzard not wanting to partake in the technology that will justify layoffs of their friends. Pretty straight forward to me.
Social implications are really hard to shake, but I agree it’s basically a lost cause fighting at this point. I’m hoping like Ed Zitron that it fails via some more durable reason.
You will not have a job in software in the future if you refuse to use AI.
I fully stand behind that prediction.
edit: despite the downvotes, I'll double down: most of today's software jobs will disappear. Your job, if it is in software, will disappear. It might transform into something new, if you're lucky. Or it might just go away entirely.
I guess I'm just confused by this sentiment. Are you making these conclusions while considering the fact that AI is still heavily subsidized? The economics of AI isn't quite the same as other software/tech.
I don't think it's going anywhere, but I don't know what happens when prices start to rise because these companies need to start turning a profit.
I'm happy using GLM 5.2 at API pricing. There will never be a point where essentially unlimited usage (i.e. what I, as a normal developer, can use at peak effort) of extremely good models will become financially unattainable.
> There will never be a point where essentially unlimited usage (i.e. what I, as a normal developer, can use at peak effort) of extremely good models will become financially unattainable.
The way you phrase this prevents refutation. But there will be a point where ordinary individuals cannot participate in the majority of the upside.
In the future, some class of models will require enormous compute that is outside the financial capital capability of ordinary individual contributors, middle class, and upper middle class. This will be sold as a capability to well-funded companies.
Yes, I don't disagree with what you're saying. I'm responding to a specific claim: that the capabilities we have now will become much more expensive in the future.
> For the vast majority of people who currently have a job in software... you will not have a job in software in the future whether you use AI or not.
The tide is coming for almost all of us.
I'll agree with you as soon as all video games are in the metaverse and run natively on the blockchain.
AI does your job. Robots do your job. These are real and substantial and actually provide enormous value. You can get more done per unit of time.
I want them to do my job and everyone else's jobs, I just want to make sure there's a functioning economy we can participate in and benefit from at the other side.
I've just built a $3M run rate company in five months on mostly AI outputs in Rust, which is rock solid code.
In my career I've been a six nines systems engineer. Stuff like billion dollar daily volume payments platforms, and my work was always situated in critical flows - so I know a thing or two about solid engineering work.
Just a moment ago I asked AI for a complicated multithreaded state machine with variational queuing logic. Claude Code Opus delivered. I'm reviewing it now. It's phenomenal code.
You need to reassess the world. Your priors are deeply flawed. These models are absolutely incredible.
I mean it - you're going to be hurt badly if you don't reevaluate what is happening and plan accordingly.
I think you'd be the first one, so I highly doubt that. If it's true, good on you.
I use Opus daily. It can take some typing off my hands, as long as I keep it to highly specific, limited, straight-forward things. And as long as I spend a long time preparing everything in the most minute detail. Then I have a chance of it being a slight efficiency gain. Veer slightly outside of those preconditions and the output is invariably impressive at first glance, garbage at second glance. Not to mention that it's barely any help for the 80-90% of the job that isn't writing code.
Even Jensen Huang expects the LLM efficiency boost to be about 30% for software engineers. Jensen freaking Huang, who has every reason to exaggerate the benefits! So I'm realistically taking that as an upper bound.
Run rate is easy to get while not meaning much (which is why GenAI vendors love talking about it). Report back in a year about how things are going, and how much of your code you had to rewrite from scratch.
>I mean it - you're going to be hurt badly if you don't reevaluate what is happening and plan accordingly.
What plans can most people reasonably come up with?
If AI is taking over software, it's taking over lots of other computer related jobs. If you're not a highly paid engineer, or come from money, most people can't just re-skill for entirely different careers.
Propertly testing and reviewing code is a must with AI, it's actually more important. I don't know why the author feels like it's assumed that this goes out the window with AI
yeah, the question is what would you do with $100 million? I mean I don't have $100 million but I'd give it to the arts after I've taken care of myself well and science. Am I gonna make $100 million playing with AI? I don't even know if I can take care of myself, shit, but the government ain't gonna do that. What are the new jobs after the Saaspocalypse? Well all the money went to those companies and what do those companies want or what do those newly minted millionaires want? Hopefully something I have but shit yo. So they're gonna go back to where they came from and be millionaires but how much is Dave Chappelle changed the city that he moved to? Is there a Dave Chappelle center for comedy? some of them want ASI so I look forward to studies on human intelligence. Maybe they'll pay human subjects to model supply two digit numbers while in an MRI machine hopefully enough of those studies to pay my mortgage.
> I mean I don't have $100 million but I'd give it to the arts
I might be naive, but rather than giving money "to the arts", I would much prefer to give money "to the people" en masse, and then leave it up to everyone to decide whether they want to make art for art's sake.
Corporations are the organizational equivalent of AI, they one thing well asmass money, their output aside from that is total shit. Not only are you betraying your friends/community/loved ones but such a setup is necessary to prevent the total collapse of society.
