site banner
PaperclipPerfector  4d ago (text post)   16471 thread views

Culture War Roundup for the week of October 19, 2026

This weekly roundup thread is intended for all culture war posts. 'Culture war' is vaguely defined, but it basically means controversial issues that fall along set tribal lines. Arguments over culture war issues generate a lot of heat and little light, and few deeply entrenched people ever change their minds. This thread is for voicing opinions and analyzing the state of the discussion while trying to optimize for light over heat.

Optimistically, we think that engaging with people you disagree with is worth your time, and so is being nice! Pessimistically, there are many dynamics that can lead discussions on Culture War topics to become unproductive. There's a human tendency to divide along tribal lines, praising your ingroup and vilifying your outgroup - and if you think you find it easy to criticize your ingroup, then it may be that your outgroup is not who you think it is. Extremists with opposing positions can feed off each other, highlighting each other's worst points to justify their own angry rhetoric, which becomes in turn a new example of bad behavior for the other side to highlight.

We would like to avoid these negative dynamics. Accordingly, we ask that you do not use this thread for waging the Culture War. Examples of waging the Culture War:

  • Shaming.

  • Attempting to 'build consensus' or enforce ideological conformity.

  • Making sweeping generalizations to vilify a group you dislike.

  • Recruiting for a cause.

  • Posting links that could be summarized as 'Boo outgroup!' Basically, if your content is 'Can you believe what Those People did this week?' then you should either refrain from posting, or do some very patient work to contextualize and/or steel-man the relevant viewpoint.

In general, you should argue to understand, not to win. This thread is not territory to be claimed by one group or another; indeed, the aim is to have many different viewpoints represented here. Thus, we also ask that you follow some guidelines:

  • Speak plainly. Avoid sarcasm and mockery. When disagreeing with someone, state your objections explicitly.

  • Be as precise and charitable as you can. Don't paraphrase unflatteringly.

  • Don't imply that someone said something they did not say, even if you think it follows from what they said.

  • Write like everyone is reading and you want them to be included in the discussion.

On an ad hoc basis, the mods will try to compile a list of the best posts/comments from the previous week, posted in Quality Contribution threads and archived at /r/TheThread. You may nominate a comment for this list by clicking on 'report' at the bottom of the post and typing 'Actually a quality contribution' as the report reason.

4
Top Bottom Old Controversial Comments
Jump in the discussion.

No email address required.

I don't know if it's because I skip most AI conversations, but I think you guys might have failed to discuss some things that have been happening for the last several weeks.

Reverse engineering with AI has gotten really, really good, somehow. It's so good that you can reverse engineer Microsoft Office and host it online for free. It's so good that you can submit your tokens to the Machine God and get World of Warcraft open source. You can now play the original Xbox Halo, but with 32 player coop (seemingly most of the players being Mexican). On your Nintendo 64. In fact, you can play Halo in your browser now. You can play lots of stuff in your browser now. I looked up an old game called Apeiron for old Macintosh, released in 1990 something, just a Centipede clone developed by Ambrosia Software, long since abandonware. There are two ports of it that I found in comments sections on Macintosh Garden. The classic decompilation site is shutting down due to sheer volume. And yes, there's also mods. Flying my Banshee across the Mojave almost makes me wish for a nuclear winter. My Foundry in Warcraft 3 just told me that my AWP research has completed, so I went on a tear across Lordaeron.

On the other side of this, AI can now shut down 70% of the Internet. This bug was in PHP for a long time, and only now discovered, because someone pointed a technological superweapon at it.

What the hell even happens from all this? Can Apple really continue to have a closed ecosystem? Are we going to wake up with no money in our bank accounts? Does proprietary code even exist anymore? How is Steam going to cope with having tens of thousands more games submitted to it? And why now? AI was good for several months up to now. Was it the release of tools that work with agents effectively? This seems a bit more significant than hundreds of mathematical proofs being published at once to me, especially because you can verify it works yourself, on your machine, and see the effects on the world right in front of you.

The economy and way-of-life of the first world as we know it might be ending. But at least we can play Bloodborne natively on PC.

I'm going to echo one of the below comments and state that reverse engineering has gotten good, but that's a long way from perfect. The various photoshop and office clones out there do a great job of looking convincing, but the reverse engineering is only skin deep.

And none of this should be a surprise, as these tools have been developed over decades with hundreds if not thousands of people contributing. That you can run Astra or Opus for a weekend and get 80% of the way there is pretty incredible, and some skeptics using the fact that they aren't 100% to say AI is rubbish is an incredibly stupid argument.

The pareto principle still applies though, and I don't think (assuming AI just stopped advancing, which it won't) this would change software that much. Maybe 1-2 months of tinkering with your agents would get you basically feature complete for all the big stuff on a photoshop, but then the rest is going to be 10000 little things that have been added in over the decades and will take ages to discover and implement. 99% of these vibe projects are going to lose interest after the first week and go nowhere. Enterprises will continue to use the big SaaS programs, and you'll see some more niche stuff switch more to internal builds. (again, assuming AI didn't keep advancing. What will actually happen is in 6 months - 1year they will be pretty much perfect and software will cease to exist)

The big impact is likely going to be the cloud coming for even more software, especially in gaming. Reverse engineering is difficult, because you only have the software's surface to inspect. Decompiling a local application is a different ballgame and I think, even if direct game streaming remains a niche, a lot of games are going to shift some parts of their functions into the cloud, to tie games to internet access and prevent them from simply being copied straight away.

"Can the results convince experts," that is the real test, imho. So for math, can AI-assisted papers actually get published in journals by non-experts using the tools. Can non-experts make art or writing using AI that is likewise well-received.

Enterprises will continue to use the big SaaS programs, and you'll see some more niche stuff switch more to internal builds. (again, assuming AI didn't keep advancing. What will actually happen is in 6 months - 1year they will be pretty much perfect and software will cease to exist)

This is why AI is so big right now: it's software and other analytical stuff, not art. This is literally worth trillions given the valuations of Anthropic + Open IA. Soon every company will be using AI, such as payroll.

In a comment down the thread, I mentioned that (1) my business makes use of proprietary software; (2) the software company wanted to switch to one of those monthly subscription extortionate pricing schemes; (3) I used AI to build a stripped down version of the software with only the features I need; and (4) so far it's working fine.

So while I agree that it would be a challenge to get 100% of the way to a reverse engineered copy, a lot of the time 50% of the way is more than good enough.

Based on my experience, I would say that AI poses a big threat to these software companies that try to lock you into their system and then rent-seek.

Yeah, also for every nifty feature that maybe somebody's added 5 years ago in the enterprise software there's also a metric ton of random bloat, go-nowhere features and weird implementations cluttering everything up. If your business only needs 15% of the grand scope of whatever Enterprise Bloatware and it's a non-core functionality which you're using the bloatware for since it's the closest thing in market, you're also going to massively struggle to get any adaptions made unless you are a strategic-scale client.

Probably better bashing together with AI at that point.

Counterargument: Google Docs. It's a 80% copy of MS Office, but it has managed to carve out a niche. Why can't a small, fast and free copy of MS Office do the same?

Because Google Docs already exists.

There are already free versions of all these. All of these are probably going to get better a lot quicker, but they still won't do the enterprise stuff that gets the big subscriptions

edit: this comment was made without seeing Brainwavez reply. Great minds, etc.

When Google Docs came out, Microsoft Office already existed. And these projects don't have to get massive enterprise subscriptions to have an impact, they just need to prevent other people from subscribing to existing suites.

Because Google Docs already exists.

Or there are other alternatives: LibreOffice, OnlyOffice…this site list contains 7 more. And don’t forget LaTeX, Typst, and 100+ free markdown editors.

LibreOffice sucks, because no one wants to work for two months to fix a minor difference between it and Excel. LLM agents do not get tired and can get asymptotically closer and closer to the software they are cloning.

That's true. And AI could fork LibreOffice, but I suspect it would be faster and better to do a fresh port (and someone could test both, with barely any effort...but I'm not paying for the credits).

How is Steam going to cope with having tens of thousands more games submitted to it?

Presumably by using AI to filter and handle submissions, as far as possible.

There's a $100 fee (well, deposit, I suppose, you get it back if you cross some threshold of sales) to upload a game. Spamming steam with games no one buys just means you're losing money.

I'm reminded of a prediction we had just 4 months ago: https://www.themotte.org/post/3758/culture-war-roundup-for-the-week/446278?context=8#context

The TLDR of that post was of AI being mostly fake tricks and hype, costs rising faster than intelligence (Opus 5 drops in 2029)... Remember the discourse we had about the AI plateau? Don't hear so much about that anymore.

And why now? AI was good for several months up to now.

Kind of? It gets better over time though, whips out new tricks easier and cheaper than before.

We had AI images that were kind of good since the end of 2022, if you did a lot of tricks with them and good prompting. Back in 2023 you could say images were good, albeit text was impossible and certain detailed direction impossible. Then there was that Ghibli phase when GPT image rolled out. And then everyone learned to accept what images could be made and everyone understood that AI could make cartoons with text and images in them, learned to recognize and resent that specific artstyle of GPT-image and nano-banana.

And so AI images are still 'kind of good' in 2026 even though they're much better than 2023.

It's the same with coding I think, except replace 2023 with 2024. Trying to code by copy pasting snippets back and forth was possible. I did it. It was a huge pain though and stuff would just not work. Then we had coding agents. Then the agents get better and better. They start being able to review and check if what they've done works. They get better and better at that. They can see, they can operate all kinds of other software for sound and images.

Progress is constant I think but we suck at gauging steady progress as humans.

2030: Someone proves the Riemann hypothesis after spending $30,000 worth of tokens over 2 years.

given how the navier stokes took $25 million , 10k agents, and a private model to prove the easier version of the problem (which some argue is not the true problem) , for an individual to prove the Riemann hypothesis for only $30k is an optimistic forecast, if anything

The models really hit prime time in mid 2026. The disproof of the Jacobian Conjecture really is what set things in motion, in July 2026. Before then it was assumed mathematicians would always have an edge.

AI images and writing are still kinda bad, but the attention has moved to coding and math, where it really shines. It took 4 years for AI to find its true purpose as a computational tool rather than a creative one.

Spending $30K tokens over 2 years (2029-2030) is the Terence Tao-approved 'mathematician with AI grad student who does the work' kind of thing. He means some mathematician does the principal work, works at human speeds and the AI largely executes directions. It's the status quo but with slightly less tedium for the mathematician.

It's not '2026 agent swarm starts and finishes in a few days.' That's a different kind of scenario, where the AI agents can do all the work and it's principally a matter of the AI's speed, the AI's price, the AI's intelligence. There was some discourse about whether the agent swarm used someone's approach, then some counter-discourse saying that was unlikely, the approach was materially different... Anyway it's not taking years and Terence Tao is upset.

Even in creative works, AI gets better and better. With Opus 5.5 release we had those music videos pop up across AI twitter, they weren't bad. Some human curation of course, some human direction. But a clear capabilities improvement.

I think its writing is fine, with good prompting. I can easily spend a couple of hours building out a story and a setting with it. I have to come up with the premise and that's a bit creatively taxing. But I'm confident that its creative abilities will keep improving every three months like it has been doing. Humans are not going to outwrite entities trained on the whole corpus of writing forever.

Nah it's still terrible at writing essays, imho. It inserts too much hedging language and it often comes off as oddly formal. People complain about AI writing because it has these tells, unless your audience are youtube, Instagram, or tiktok, I suppose . But where it can be useful is if you're trying to write a sentence, but you cannot get the turn of phrase quite right, so you can feed it to Ai and it generates many suggestions and you pick the best sounding one.

Yeah, there is an issue with information density and it not getting right to the heart of the matter.

But also - people don't notice. Not the majority of people: https://x.com/SamaHoole/status/2113126506457186451/photo/1

I see it all the time online:

The animal that does the invisible job is always the one on the ledger as a cost.

yeah most people may not notice AI, but all it takes are a few people to make a stink about it and it costs you invaluable credibility, then others see it , too "cannot be unseen effect". Someone taking a screenshot of a pangram score overshadows any good will the author had.

A browser is like a mini computer ,except everything is in virtual memory . the ability to run applications , or play games , such as flash, long predates modern AI. Same for porting and reverse engineering stuff.

The walled garden aspect is not going away. Apple and Google still have a duopoly on apps. Same for the monopolies on payments. You cannot just vibe code your merchant, as orders route though the same credit card providers. Or vice code the interoperability between the icloud and your mac/phone. or the social network aspect of facebook/instgram.

Does proprietary code even exist anymore?

"For my first act, I'm going to obliterate the source of revenue of my most obvious customer!"

Maybe the next big release from the primes will be an Opus-performing LLM you can run on your laptop.

It was bound to happen. If vibecoding is so easy that software companies can dump all their programmers, then why do we need software companies to begin with? For that matter, why do we need software? If ChatGTP can generate complex code off of vibes, then certainly it can manage spreadsheet functions on its own, or obviate the need for Photoshop (see me comment below), or do whatever SalesForce does, or run engineering simulations, or anything else you want software to do except maybe play games.

It was bound to happen. If vibecoding is so easy that software companies can dump all their programmers, then why do we need software companies to begin with?

Yeah, it's sort of like that bank robbery scene in The Dark Knight where each thug does his job and then gets shot by the next thug who thinks that he's increasing his share of the loot, until finally at the end the Joker ends up with everything.

I could easily see something like that happening with AI. i.e. a company uses AI, fires most of its workers; then the company's customers start using AI and dump the company; and then the same thing happens to the customers; and so on. Probably that's part of the reason the big AI companies are going so crazy right now - they are hoping to be among the last companies standing after the chain of dominoes goes down.

then why do we need software companies to begin with? For that matter, why do we need software?

It has to run somewhere? Runtime environments are still a huge deal.

The "somewhere" is increasingly "somebody else's computer". With how much software companies are pushing web-based cloud models that don't run anything locally, I don't see why this would be a problem.

It's a "turtles all the way down" type problem. the infrastructure has to be hosted and run somewhere

If vibecoding is so easy that software companies can dump all their programmers, then why do we need software companies to begin with?

Legal liability, and the policy control to punish/exclude/extort the organizations that don't play along.

Certified regulated entities perform a variety of useful functions, among them being a legal entity to offset legal/monetary blame onto. This is one reason that legal personhood corporations overtook personal-liability ownership models in the first place. An extension of this is that the regulatory/legal superstructure that permits such legal constructs can also use failures to comply/submit as a basis to move against businesses that want to cut those legal-compliance department costs. Typically this is done for deterrence, but it can also be used for other government purposes, ranging from seizing assets (for profit) to strategic advantages (pretext for targeting foreign interests).

What I’m really excited for is the actual Holodeck / Tron / what people thought future virtual worlds would be like in the 80s / type stuff where it just generates fully immersive, coherent, endless, well-written virtual worlds on the fly. I don’t even need VR, just an endless Red Dead II tier world

What I’m really excited for is the actual Holodeck / Tron / what people thought future virtual worlds would be like in the 80s / type stuff where it just generates fully immersive, coherent, endless, well-written virtual worlds on the fly. I don’t even need VR, just an endless Red Dead II tier world

In theory I agree with this but I'm a little nervous about holodiction.

the single wheel bikes are sorta similar to Tron i guess. Maybe in 10 years VR will finally catch on, but they said that 10 years ago...

Same here. I do think we're on the fast track to gaming experiences that truly would have seemed like ridiculous sci-fi ten years ago. Not quite there yet, of course, but there are signs. I consider it a significant milestone that there are now tons of bespoke AI music videos on Youtube (and, I'm sure it's just my lack of taste and discernment, but I find a lot of them very very good). Going from niche personalized music videos to niche personalized movies to niche personalized videogames doesn't seem like it's going to take too long (though it's possible we'll hit a compute wall somewhere).

Man that LGTM video is a banger.

I've recently posted about the Helldivers 2 modding Renaissance enabled by AI. As I had feared, it only lasted a few weeks before the developers cracked down and added Easy AntiCheat to their game. The primary impetus, of course, was the fact that their wallets were being hurt.

I left a suitably negative Steam review.

The good news is that barely a few hours after the patch, mods that bypassed EAC showed up. The bad news is that the overall difficulty of creating new mods has shot up drastically.

Why? My understanding is that the devs removed a lot of the hooks that allowed for easy scripting, and encrypted some of the code. Most modders using OAI or Anthropic models found that they'd no longer engage in intentional circumventing of DRM, leaving aside the classifiers kicking in. Previously, the devs had simply made no effort to obfuscate things, unfortunately the merest act of putting on a lock (no matter how shoddy) means many AI models suddenly develop a newfound respect for property rights.

Clearly people are finding their ways around it. I'd imagine some have resorted to using Chinese models like GLM 5.3 that are more gung-ho about greyhat territory, or found other tricks.

I make no strong claims on the balance between attack and defense here, or if it's going to turn into an endless cat-and-mouse game. We shall see.

The prop software dudes are on a losing battle. Qwen 4 and Kimi 3.1 are coming in just a few days and the previews once again show they are only a low single digit number of months behind the western frontier. We'll have Opus 5.5 tier open models by the new year. We now live in a world where you can Reverse Engineer Anything and western software companies need to make peace with that.

The devs have access to the same tools, as I've noted before. It's also an online live service game, so even though people have already made their own offline forks (with co-op, because why not), it's not impossible to lock down official servers.

That makes it somewhat different from reverse engineering Photoshop or making a cheap and cheerful clone of a game some dev made the mistake of teasing on Steam.

I still don't know how it'll work out over the medium term. Eventually, we'll also have communities making their own feature-complete and free alternatives to proprietary games, but for now it's mostly reproducing or extending existing works.

I honestly don't get why devs in general don't allow you to choose between starting the game in anti-cheat mode or free-for-all mode, like Halo: Master-Chief collection does. Seems like the fairest solution.

(Putting aside the financial incentive in this case. The Helldivers 2 micropayments model is trash).

You've acknowledged and set aside the biggest issue, which would already explain everything: money.

Arrowhead are also particularly stubborn about responding to community feedback. Well before this all blew up.

I was speaking more broadly. It’s annoying to have to hack Elden Ring and disable online play, risking a ban, because I want to try it with DLSS5 and anti-cheat prevents all modding.

Previously, the devs had simply made no effort to obfuscate things, unfortunately the merest act of putting on a lock (no matter how shoddy) means many AI models suddenly develop a newfound respect for property rights.

I mean, that seems fair. I like the various reimplementation projects, but I don't see what good a decompilation free for all is supposed to bring. Just make your own, it's never been easier.

Is there a list of all these "play old game X in your browser" games somewhere? I could stand to play Wing Commander 2 again.

You know what would truly make my Glorious AI Future? AI creating a Freespace 3 that satisfyingly finishes the story.

I personally hope AI gets to the point where it can make a Final Fantasy VII remake which is actually good. That would be something good to come out of all this, at least.

I don't think so, it's just an adventure of discovery for everyone. A common complaint on Github is that nobody is really contributing to other repos right now, they just start up their new reversing project and fire away. I didn't even know an Oni web port existed until I read about it from some other random user. Also EV Override.

Garrett Petersen, who's been working on a from-scratch retro game with AI assistance, put it as:

AI only really became good enough to take a high quality project to completion in 2026. There were lots of projects started in 2020-2024 that are finishing now. The creators say, "look at this thing I made without AI." But of course they didn't use something that barely worked!

Anyone starting a big ambitious solo project now has to choose between using the new tools and finishing in 2026-2027, or not using them and finishing around 2031.

I don't think that's absolutely true -- there's some solo projects that are ambitious in the Linux ethos of doing something small really really well where the time tradeoff doesn't exist. But it probably tells you why a lot of things seem to be happening all at once, even as the models or AI tools haven't had a quantum leap for a bit.

I started my little hobby project in 2013, and I expect to be done with it about ten years after I retire, so around 2075, if life should last me that long. Not because I refuse to use AI, but because not even current-gen AI can overcome my convoluted working process. At least not for the the kind of budget I can afford.

FWIW I use proprietary software for my business and the provider recently switched from a modest one-time-for-life fee for the software to a subscription model which was far more expensive. (I guess they were counting on the fact that for most users it would be time consuming, inconvenient, and expensive to switch over to something else.)

Anyway, I asked the AI to create a system for me which had all the features I needed and which could upload and incorporate my current data. It took less than an hour to get it running and so far everything seems to be working perfectly fine.

So yeah, I would say that AI poses a threat to software providers who try to engage in rent-seeking behavior with subscription fees.

Stallman-level intensifies...

This bug was in PHP for a long time, and only now discovered

What people don't realize is that this is good actually, because "only now discovered" just means "only now made public". This PHP vuln was getting exploited for months, people have retroactively found log evidence. Democratization of security research is a massive win for everyone.

yeah like the firefox tor exploit the us govt. secretly used

This PHP vuln was getting exploited for months, people have retroactively found log evidence

Did they hit anything particularly big? The tone I got from the revelation that this was like a doom-level exploit

The media always exaggerates these things

Yeah, I was thinking of starting a post about all this, but couldn't quite muster the motivation to collect all the links, so thanks! I don't recall how it was phrased, but I remember someone saying "Everybody is using AI, and everybody's saying how capable it is, so howcome nobody has done anything with it", at the time I agreed and decided that was going to be my personal benchmark / canary. Well gents, I think we're here, even if you're skeptical of these examples, those clones of the Microsoft/Adobe suite are insane. If you're calling that canary alive, I'm going to say you're re-enacting Monthy Python's dead parrot sketch.

I think we're going to to see a lot of disruption, but despite my pessimistic nature, I somehow don't feel like "the economy of the first world is ending". Somehow this feels more like the Industrial Revolution, or how Photoshop wrecked the world of hand-drawn art, but we're still in the early stages so it's hard to tell how things will develop. I'm also pleasantly surprised by centralization not immediately elevating Big Tech to all-knowing dictators. Even though technology favors centralized control, it seems like the pressure from open source / open weight models is real, and puts a hard limit on how hard the Silicon Valley creeps can cuck us.

I can't make sense of the Culture War angle though. When @DaseindustriesLtd said that AI is going to be the next big Culture War issue, I thought he was nuts. I still think he's nuts, he has no right to be right about this! This is the dumbest culture war issue I've ever seen, and yet... Not that I'm complaining, watching the same people who have been meticulously taking over institution after institution, and purging all dissidents, self-quarantine themselves from tech that can 10x your productivity, and thus leaving themselves and their institutions in the dust, is quite satisfying. On the other hand, I might need to intensify Aldeaposting. Even the innovations of the digital / social media age had a tangible negative impact on me, that I haven't figured out how to fully mitigate yet, and with AI the temptation to outsource my thinking to the quartz god is real.

If you're calling that canary alive, I'm going to say you're re-enacting Monthy Python's dead parrot sketch.

The problem with most of these examples is that they boil down to plagiarism in one form or another. The AI are not being used to create new stuff, they're being used to automatically reconfigure existing stuff either for the purposes of IP evasion (the Photoshop clone stuff) or for cheap gimmicks (juxtaposing consumer icons together in video games). Chomsky had dubbed chatbots plagiarism machines all the way back in 2023, so yes, I'd make the claim that in a sense nothing's really changed since then. They're still just prediction software recreating what they're trained on with increasing fidelity, which isn't the same thing as designing or engineering things that are new. The utility of this is overstated. If something already exists, why would you need an AI to create clones of it? It circles back to IP evasion and gimmickry.

Automatically reconfiguring existing stuff is an incredible demonstration of capability. Smashing Counter Strike and Warcraft 3 together would never have been done before this because it's really difficult. Decompilations had been done before this, but they took years of effort. That Diablo 1 decompile was done by one guy, who said it took him over 1000 hours. This isn't something you can wave away as "gimmickry" and "plagiarism". If it can do these really difficult things, it can certainly create really good new things, too.

Also how many genuine original ideas exist? People stand on the shoulders of giants doing their best to jam together information that exists previously, for the most part, if AI can just do that at scale then it kinda wins.

A bulldozer can do really difficult feats of moving vast masses of earth about, too, but its potentials are still limited.

If all AI ever is is a digital bulldozer, then, not unlike real bulldozers, it's still going to reshape the whole world.

The problem with most of these examples is that they boil down to plagiarism in one form or another.

How is that a problem? All software that competes with other software boils down to "plagiarism" under that criterion. We've been doing that artisenally for ages.

Chomsky had dubbed chatbots plagiarism machines all the way back in 2023, so yes, I'd make the claim that in a sense nothing's really changed since then.

What's changed since then is that back then you couldn't use it to compete with mainstream industry. Now you can.

The utility of this is overstated. If something already exists, why would you need an AI to create clones of it?

This is absurd. The reason you would want a clone is because you want something similar, but changed in some key ways to better fit your use case. Or because you want an older version of something that you were very happy with, but the provider refuses to support it anymore, or because you want to increase it's performance, or use it on hardware that was not intended by the original creator, or, or, or...

It circles back to IP evasion and gimmickry.

What if it's a clone of a an open source program, where's the IP evasion then? How does the entire entertainment industry not fall under your definition of "gimmickry"

Photoshop was a triumph of compounded design inputs over the course of its existence, which were very costly to work out and optimize. The purpose of IP law at least ideally is to encourage creative design by allowing designers to monopolize their abstract goods for some period of years and thus secure remuneration for their efforts. It is not equivalent to have prediction software clone a finished design-based good in order to avoid paying a fee or to alter minor aspects of it. It is not equivalent because it does not create the same utility as encouraging people to pour large amounts of creative effort into designing new products in the first place. What you are doing is trying to build a false equivalency between using know-how and techniques to build something from a principled standpoint of design and using predictive software to perform something more akin to a copy-paste function. AI does not put very good design thought into its outputs, that much is observable from said outputs (however opaque its internal workings may be).

You can't copyright an idea, this is basic fundamental IP law. You can protect designs yes, but only for a few years. Copyright over software only covers the exact source code itself and as the claimant you must prove the defendant copied from your source code to hold them liable. If they take 100% of the functionality of your code and make their own copy from observing just the functionality itself that's completely fine and you can't do a single iota of stuff about it.

From Pumfrey J back in the day:

Take the example of a chef who invents a new pudding. After a lot of work he gets a satisfactory result, and, thereafter, his puddings are always made using his written recipe, undoubtedly a literary work. Along comes a competitor who likes the pudding and resolves to make it himself. Ultimately, after much culinary labour, he succeeds in emulating the earlier result, and he records his recipe. Is the later recipe an infringement of the earlier, as the end result, the plot and purpose of both (the pudding) is the same? I believe the answer is no

IP protection was never as broad as you're claiming it to be. If it was, all the Open Source clones of mainstream software would be illegal, and yet we have Gimp, Libre Office, and 7 zillion open source browsers that actually outcompeted their proprietary counterparts. What you are calling for is a complete abolition of all competition. Also, I will note, you never answered where is the "IP evasion" is taking place when the program being clones is open source to begin with.

What you are doing is trying to build a false equivalency between using know-how and techniques to build something from a principled standpoint of design and using predictive software to perform something more akin to a copy-paste function.

I'm doing no such thing. First of all, I made no claim about whether AI is "building something from a principled standpoint", the only argument was about whether it can be used quickly and cheaply to catch up to the industry standard. Even if the underlying mechanism is akin to copy-pasting, the technology will have a massive disruptive impact on the industry.

Secondly, if you read my "Aldea" link, you'd see that I actually share many of your concerns about AI and the effects it will have on innovative thinking, or even thinking in general. It's definitely something I believe we'll need to address and mitigate, but the one thing that will do worse than nothing to address this problem is pretending that nothing changed since 2023.

AI does not put very good design thought into its outputs, that much is observable from said outputs

I can agree with the former, I don't think it's putting thoughts into it at all. But if your argument was correct "not putting good thoughts into outputs" would mean that the outputs are of inferior quality, which in turn would mean they are of no threat the artisenally created software. Since they are, something in the logical chain between the first and second part of your sentence is broken.

Software is protected by copyright, not patent, which means that while the term is longer, the protections aren't as strong. Copyright doesn't cover the software itself but the source code, which means you could develop a program with identical functionality to an existing program as long as you use different code. What this means in practice, however, is that you have to develop the product independent of the existing product, i.e. you can't just take the existing code base and make superficial changes. Courts generally require that the code itself be developed entirely independently. All the programs you listed (Gimp especially) aren't even clones of existing products, just products with similar functionality. Gimp wasn't developed by someone taking Photoshop's code base and modifying it. If what the AI is doing is taking the existing product and reverse-engineering or reverse-compiling or whatever it is to figure out what the code is, then as far as the law is concerned it's no different than if they copied and pasted the executable file. And no, it's not a strong defense that their versions don't have the account control features that verify that the user is a paying customer.

Yeah, seriously, it's not really even a debate at this point, is it? I even forgot an example, Modern Warfare 2, and in the last hour my attention was brought to Diablo 2 that was uploaded... yesterday. Someone needs to tell the years-overdue Kickstarter scam developers that the jig is up.

Wow, I'm glad I found Emdless Sky and Naev when I wanted an updated EV style game. They may not be perfect, but at least they're free and playable.

Endless Sky was great! I really liked EV Nova but I don't want to start up Sheepshaver just to play EV Override, I want a real port. It hopefully happens soon. You can already play it in your web browser, but it's not been reverse engineered yet. This site just plays the compiled binary in the browser.

years-overdue Kickstarter scam developers that the jig is up.

I think one big issue is having a project without raising enough money to work on it full time. If that's the case, it's still just a hobby project, with the money raised versus not. People like this should just keep working on their hobby and release it when it's done.

Honestly scamstarter should set a minimum of like 200k or something just to raise money, as anything less isn't enough to actually fund the development of a new product. And if they want to raise less just to do a production run at a factory for a product that's ready to make, they'd better have a final quote from the factory before raising the money.

It's okay, at least the remaining developer is still getting $376 per month on Patreon to do, uh, nothing.

An example I've been meaning to list is a youtuber randomly deciding to make his own Linux distro. I thought it's a meme, but apparently it's halfway decent.

Hardware exploits being much easier to find is also true. PS5 had several discovered and patched, and now even Switch 2.

Personally what I look forward to is retro games being translated. Imagine opening up a mook listing Top 100 FC or SFC games, and now being able to play evey single one knowing only English. Romhacking now no longer being a bottleneck. While path dependence would still cloud the judgement, arbitrary choices which games would be localized made by Americans in the 80s and 90s, now the barrier wpuld be lowered for those interested in game from those times to come to their own conclusion.

Perhaps a situation analogous to discovery of new manuscripts of Ancient authors will occur. Some text may have been only found in the late 19th century, thus it was unable to exert influence on thought about its writer, until its rediscovery, but now this is no longer the case. A more complete tradition now forms.

in an amusing coincidence, my "for you" twitter feed has been not-stop shit-flinging between Japanese users and everyone else over a mod - not even a prated copy, but just a mod - of Tekken 6 that lets you play online. When there's finally translations of Mother 3, the shitshow is gonna be glorious.

Mother 3 was translated years ago pre-AI.

Yeah aside from game slowdown during text scrolling, it seemed about perfect to me. You can also play it natively on your 3DS or DS, which is what I did.

I’m still in the “nothing ever happens” camp unless there’s a literal apocalypse (either ASI scifi world or the literal extinction of humanity). Everything else just washes out in the end and adds up to normal.

I don't know about office but the games weren't reverse engineered, they were decompiled, as far as I'm aware anyway.

Presumably the solution will be some sort of run-time encryption, possibly combined with requirements of being online and some parts of programs running off-device?

And as far as I'm aware, decompilation is a kind of reverse engineering. You use Ghidra or some other decompiler and get the assembly code, and then you write something that does the same thing that the original functions did. The best ones might match byte-for-byte when compiled with the same compiler, but otherwise, a good goal is to get each function to output the same things given the same inputs. The guy who reversed Diablo seems to use the terms interchangeably.

lol you're just now hearing about it? 😄 Pretty sure i mentioned it about these gamecube games being decompiled and 100% matched, such a thing could be used for d2 as well, though it's harder to split and match PC binaries

Edited to add: on your solutions point, there is only so much you can do. Even without AI tools, Red Dead Redemption got cracked eventually, despite its DRM protections. You can always set up a server emulator or your own server, which happened with Club Penguin, and probably will happen with World of Warcraft.

Reverse engineering seems like the wrong word when the decompiled code essentially amounts to an extremely detailed instruction manual.

I'm not saying this isn't impressive, it's just that the office and Photoshop things seem much more like reverse engineering. Decompiling is something different.

And as far as I'm aware, decompilation is a kind of reverse engineering.

It's pretty different as used here though -- for one thing "super-decompiler" is pretty neat, but not nearly as impressive as "can code office from scratch"* in terms of broad applicability of the tech.

Decompiled software is also still covered by copyright, so while new forms of piracy are also pretty neat I don't really see much of a business impact here?

*More like "can code an office demo from scratch" AIUI at this point, but still pretty impressive

On the other side of this, AI can now shut down 70% of the Internet.

The fears of AI superpowering the script kiddies were overblown, at least so far: https://eliastaylor.com/noexploit/

I think that a lot of people had assumed that LLMs will be able to build an exploit chain instantly from reading the diff, and that even if this was harder to exploit than other exploits, LLMs would've levelled the playing field. Well. There's no working exploit. Now what?

That's not to say that the exploit isn't bad, but it may have been overhyped. The actual "hack any website instantly" button remains in the hands of only a few elite hackers, if it exists at all.

you can reverse engineer Microsoft Office and host it online for free.

This is bullshit. According to anyone who actually claims to have used these vibe coded clones, they are buggy and unusable. They are created by grifters who want attention and do not even use the tools they are cloning.

I unfortunately have to admit that there are some (not many) quite useful vibe coded apps starting to roll out, but this isn't it. As far as I know, all of the useful ones were created to solve some of the creator's needs, rather than just built to be something flashy.

This is bullshit. According to anyone who actually claims to have used these vibe coded clones, they are buggy and unusable

I use the vibecoded photoshop clone. I don't know how it compares to actual Photoshop because I'm not deep enough into graphics to justify paying Adobe, but it sure beats Gimp for my use cases. Call it damning with faint praise if you want, but a small army of driven and intelligent people worked on that project for decades.

If I ever get the time (a big if), I'd like to install the clone on a VM and see how it compares. I say this as someone who pays for Photoshop, but who isn't a graphic designer or using it for anything work-related. The problem with a lot of this testing is that, even if you do have the full version of Photoshop for comparison, the relative quality isn't going to be apparent from testing alone. You actually have to rely on the thing in question for some time and see how it goes, and see how the frustrations build. To give an example, ChatGTP's image generation is actually fairly good at doing things that Photoshop can't, notably upscaling low-resolution images. The first few times I tried it, the results were impressive, and I started making it part of my PS workflow for stubborn images that would take too long to edit the traditional way. The problem (if you can call it that) is that it only works about half the time, and that's being generous. It's not merely editing the image but generating an entirely new one based on what it thinks the old one should look like. When it gets that wrong, which is often, it generates something that is not only an incorrect representation but one that can't easily be rectified in Photoshop. It also works better if the images are in good shape going in, so there's a lot of preprocessing involved.

The upshot is that, a not insignificant percentage of the time, AI actually increases the amount of work involved. To be clear, I'm using this primarily to edit album covers for which clean versions don't exist, and the work involved can range from minor editing of a decent-sized image to working magic so a small, poorly taken one looks presentable, to completely reconstructing an image from scratch using the original as a guide. I now need to add a brief extra step of deciding whether it's worth attempting to use AI. Then I need to find an image to use, and prepare it. The extra work involved comes from the fact that the best image to use for an AI reconstruction isn't necessarily the easiest one to edit manually. I would also add that even successful AI generations have a tendency to oversaturate the colors or otherwise screw them up, so there's an extra step of loading the finished image into Photoshop and dialing things back (I have to do this anyway because ChatGTP generates 1200×1200 images and I want 1000×1000). When it works well, the results are nonpareil. But getting it to work well means doing 75% of the work for a 50/50 shot at something usable. It isn't so much a time-saver as it is an overall value-add.

The whole experience kind of sums up my opinion of AI workflows in general and why I'm not convinced that it will dent employment in any significant numbers. It seems impressive when you first use it, but when you actually try to implement it, it doesn't produce results with the consistency one can expect from doing the work manually. Yes, humans make mistakes and do shoddy work to, but humans screw up in predictable ways. The mistakes AI makes aren't ones that humans typically make. For example, people talk about AI use in my own field, law, as though it could eliminate 90% of our work. Aside from the fact that people who say this don't seem to have much of an idea about what lawyers actually do, the well-publicized boo-boos of lawyers submitting filings to the court that contain hallucinated cases that happen to perfectly fit with the case at hand aren't the kinds of mistakes lawyers typically make. In fact, if I fabricated a case in a motion because my argument was a loser or there wasn't anything on-point, I'd be sanctioned by the judge and fired from my job. Even if a first-year attorney submitted something to me for review, I'd assume from the outset that the cases he cites are real and that he's not misrepresenting them. Having to fact check every citation does not cut down on the amount of work involved. It may mitigate it somewhat, but that assumes that we're drafting from scratch; most motions I file are similar to ones that I've filed before, and I'm not going to reinvent the wheel every time I need to file one. AI drafting could theoretically give me scaffolding from which to work, but even then I can probably find a motion someone else has filed if I'm looking for a bare template. And that doesn't even get to the problem that I need to be prepared to argue the motion and am thus going to have to do the research equivalent anyway. So AI may shave some time off of drafting motions if I need to draft something I haven't done before, and even then it's not like I spend a ton of time drafting motions to begin with.

The upshot is that, a not insignificant percentage of the time, AI actually increases the amount of work involved.

I know what you mean, I've run into this myself a few times, but I don't think all workflows are affected the same way. For images what you said harkens back to the old graphic designer's lament, when a customer really likes what you did, they just want a "small adjustment" like having a character face the camera, instead of facing sideways. With AI maybe the changes you want to apply aren't quite so drastic, but when the output is a flat unlayered image, making a small adjustment will indeed add work.

I don't think this applies to text output, though. Text is easy to modify, so if the AI gets off track you can correct it cheaply. Code comes with it's own set of traps, where it's possible to write something that makes sense when you read it, works, but contains so fatal flaw. Early on developers reported that whatever gains in code output are made up by the extra time spent on reviewing it. It's still an issue to some extent today, but in my experience nowadays AI just comes up with dumb ideas you never asked for, rather than doing what you wanted, with fatal flaws attached. The former is a lot easier to filter.

Either way, the proof is in the pudding, and I'm telling you reproducing the entire Adobe and Microsoft suite would have been a gargantuan undertaking as recently as last year, or even a few months ago. And it's not theoretical, we have actual examples of intelligent, driven people, with an ideological hate-on for big corporations taking their best shot at reproducing these suites to give them away for free. They've put countless man-years into it with laudable results, but since they don't have the same amount of resources they usually couldn't keep up with industry. Their work is now getting btfo'd by AI, which you can say is a far cry from btfo'ing Adobe itself, but I think it shows that all these arguments about how AI only seems impressive when you first use it, or that it makes errors humans wouldn't and this somehow costs more work than it saves, are all flawed.

The whole experience kind of sums up my opinion of AI workflows in general and why I'm not convinced that it will dent employment in any significant numbers.

I don't know about employment in particular either, people adapt, and it's not like the introduction of power tools caused massive unemployment in the construction industry, but we are definitely looking at a similarly sized disruption.

This is bullshit.

Yes, probably it is, but we're not far off from an actual reverse engineered project being hosted somewhere. That was what I was trying to get across with my numerous examples of video games getting reversed or having extremely in-depth mods being slotted into them. This has implications far greater than just free and easily moddable video games.

I'll caveat that while that post isn't entirely wrong, it's a little overstated. PHP-FPM will generally terminate naive attempts to exploit the bug unless the rest of the infrastructure is also fucked up, but Apache's PHP module won't. While this exploit chain is difficult and fairly time-consuming per-site, it's available well before you get to 'elite' level.

I guess there's an argument that running mod_php instead of php-fpm is itself a misconfiguration (or, frankly, using Apache), but that's a level where you're arguing definitions.

It's not a Heartbleed-level incident or as bad as 4, but it's still pretty high up there.

https://www.php.net/manual/en/install.unix.apache2.php

Do not use mod_php for new installations. The instructions on this page are retained for historical reference and for the rare cases where embedding PHP directly in the Apache httpd process is specifically required.

At this point they should really stop shipping it

or, frankly, using Apache

What's wrong with apache?

At this point they should really stop shipping it

That'd be nice, but there's a pretty big asterisk there. A lot of software ships with support for mod_php, up to and including mainstay software like Nextcloud and some (paid!) web hosts, and warnings about performance never stopped people from just using it.

What's wrong with apache?

I'll admit that this is partly an artifact of my specific use cases where there's a lot of reverse proxy or wonky configuration going on, but I've almost always found nginx a better (if slightly more annoying in syntax) option, performance, reliability, and consistency of behavior under load. The only really big selling point I've ever felt for Apache is the .htaccess concept, but unless you're running a VPS that's always seemed more a source of footguns than a benefit.

(I guess Apache and one-button LAMP stacks are easier in Windows compared to LEMP? But speaking of footguns...)

A lot of software ships with support for mod_php, up to and including mainstay software like Nextcloud

True but I feel like very little php software at this point requires mod_php and doesn't run better on fpm. Nextcloud prefers fpm.

They should yank the module so that incompetent shared hosting providers are forced to update their setups.

What is this place?

This website is a place for people who want to move past shady thinking and test their ideas in a court of people who don't all share the same biases. Our goal is to optimize for light, not heat; this is a group effort, and all commentators are asked to do their part.

The weekly Culture War threads host the most controversial topics and are the most visible aspect of The Motte. However, many other topics are appropriate here. We encourage people to post anything related to science, politics, or philosophy; if in doubt, post!

Check out The Vault for an archive of old quality posts. You are encouraged to crosspost these elsewhere.


Why are you called The Motte?

A motte is a stone keep on a raised earthwork common in early medieval fortifications. More pertinently, it's an element in a rhetorical move called a "Motte-and-Bailey", originally identified by philosopher Nicholas Shackel. It describes the tendency in discourse for people to move from a controversial but high value claim to a defensible but less exciting one upon any resistance to the former. He likens this to the medieval fortification, where a desirable land (the bailey) is abandoned when in danger for the more easily defended motte. In Shackel's words, "The Motte represents the defensible but undesired propositions to which one retreats when hard pressed."

On The Motte, always attempt to remain inside your defensible territory, even if you are not being pressed.


New post guidelines

If you're posting something that isn't related to the culture war, we encourage you to post a thread for it. A submission statement is highly appreciated, but isn't necessary for text posts or links to largely-text posts such as blogs or news articles; if we're unsure of the value of your post, we might remove it until you add a submission statement. A submission statement is required for non-text sources (videos, podcasts, images).

Culture war posts go in the culture war thread; all links must either include a submission statement or significant commentary. Bare links without those will be removed.

If in doubt, please post it!


Rules


Recommended Posts And Communities

Recommended Realtime Chats