AI is mass decompiling games. That's both awesome and scary
popcar.bearblog.dev
popcar.bearblog.dev
This is the future I always wanted, full control over the hardware and software I "bought" but I think we will end up lamenting our moment of empowerment in one way as the time we lost even more agency in the end.
Personal computing is increasingly under threat by government and big corporations, we all need to be vigilant and fight back.
When people with no prior experience, including teenagers, can produce a polished game by investing a few hundred hours of their free time to do the creative part, what do we need Ubisoft and co. for?
There is zero indication that that will ever happen. It takes a lot more to create a polished game like Assassin's Creed than telling a chat bot to "make it fun" while using stolen game assets. It's cute to watch mario run around in call of duty (https://www.dexerto.com/call-of-duty/call-of-duty-fan-create...) but most of the AI mashups people are putting out like that recently are basically unplayable garbage. It's fine, it's mostly for laughs and likes/views anyway, but the idea that people with zero experience in making games are going to be vibe coding games that are worth playing and on par with the best Ubisoft titles is ridiculous
The idea that this will next year be possible for more polished 3D games with a larger scope is far from ridiculous.
If a community of users get together on Discord to collaborate on a project, I see no reason why this should not work.
I'm thinking more of the technical capability. I tested some Chinese browser game that played like an old GTA and it worked decently well on my iPhone.
Evidently the tools are good enough now.
I believe the issue is that there's a lot of people who just release slop. Both for games and regular software.
To build a decent product you need to invest months into testing and refining. Clearly many skip this in order to ship barely working garbage.
For the capabilities of eg. Opus 5.5 to diffuse into actual products rather than just demos, it's going to take some months.
I'm sure some readers are thinking that I'm just moving the goalposts into the future here, but it's the truth. For the games that were made with AI and which already are making money, I found that many are projects started already last year.
I myself have been working on a commercial mobile app for the last six months, targeting a release this week.
Building a good product still takes time.
The answer would be no, obviously not.
The question is whether AI enables the general public to build fun games with no prior experience and with a fraction of the previous effort.
What I said is that some of the successful and therefore likely fun games made with AI are projects already started like late last year.
Considering that it still takes months to build a good product, these capability improvements are diffusing relatively slowly.
It is imaginable that models available in early 2027 can allow individuals or small teams to produce games of AAA scope, taking care of the technical aspects.
You still wouldn't expect to see a finished product until later that year, unless the AI developed something resembling taste and creativity, which today still needs to be injected by us. Not to mention stronger capabilities for evaluating visual work, motion, and physics.
Will you prompt the movies you want instead of watching a movie made by someone else? Will you prompt the novel you want to read? It’s nonsensical in other art form, it’s nonsensical in video games too unless you go for the lowest common denominator friendship or similar.
If it’s functional, good enough, and the price is right people will prompt generate games and share them with or without the gate keeping game industry.
There’s a lot more to this world than software developers.
Quality free game engines have been available for years. Being a game developer has gotten so accessible that the market is completely jammed with titles and discoverability is very difficult.
I get the feeling from your post that you feel “ability to write code” was something put in place by “gatekeepers” but not only is that not what “gatekeepers” means, but software developers make up a small part of a game’s production team.
Also, it should be really clear from the past decade or so that the games industry is not especially gatekept. Not only are there bazillions of indie games on itch or steam, many of the most loved games of the past decade are tiny indie games!
Trying to vibe code such algorithms generally yields results that are not suitable for realtime gaming performance. I’m sure they’ll get there, but saying “make it fast” without carefully directing the algorithm design causes LLMs to add hack on hack on hack, creating a slow, spaghetti code monstrosity.
Is this your first hand experience?
If you consider that "producing a polished game" then I can produce a polished game with a DVD burner and a Steam account.
I mean to be more clear I have a friend all excited about making really nice games and now realzing he’s one of a bajillion others so no one is gonna pay for it (less you’re awesome at marketing in which case call me )
Children especially are extremely tolerant of complete garbage (I have similar concerns with animation) but I think targeting older audiences, respecting players by avoiding typical player-hostile bullshit (ads, in-game purchases, time gates, etc.) and really putting in the effort to make something special will still attract players.
It's going to be very rough for anyone trying to make simple arcade/flash/mobile style games. People will probably be able to vibe-code those games themselves well enough, but games with a well crafted story, amazing art and music, and satisfying gameplay will still offer unique experiences that you won't be able to just prompt your way into creating.
AI has the potential to give gamers something to play with for a while, but eventually people will get tired of poorly made AI reiterations of games they've already played and they'll want something new and actually creative. AI games aren't going to kill that demand for the same reasons AI won't be the end of books or movies. Making something that's really good takes talent and creativity and AI isn't capable of that. Most people generally aren't either. Even talented and creative people aren't interchangeable. If A Song of Ice and Fire is never finished I can ask AI to write the ending for me, but it's not going to be satisfying. Nothing stops me from writing the end for myself right now, and I can even make anything I want happen, customize the story and text for my exact preferences but it still wouldn't stop me from wanting to read whatever GRRM finally manages to release.
I agree. You need to constantly inject taste and creativity via a lot of testing, feedback, and iteration.
With current technology, this still takes months to years of effort, depending on the scope of the project.
I feel that when you say AI game, it sounds like slop produced near autonomously in a single weekend.
There are other ways to fully use AI for software implementation and asset generation, but still apply standards.
One can spend a full day refining a single gameplay element to tune it into something enjoyable, or to explore and select assets that look great. This can happen with just prompting and testing.
Many skip the effort, and it currently of course results in a bad, barely working product.
That's most of the project I've seen people post about at least. Either actual clones of simple games, or just dreams of saying to AI "Make me red dead redemption but with pirates!" and expecting something playable.
If you spend enough time with AI you likely could come up with a great game. It'd probably take longer to explain to AI what you want, correct what it gave you, explain how you want to teak it, then correct that, etc. than it would just to code it yourself (for those who can code) and as the project grows there's a risk of it becoming unmanageable unless you have AI hyper-compartmentalize everything to fit each function within the limit of what it can keep track of, but that's going to have impacts on performance and maintainability too.
It may not be the most efficient way to do it, but as long as you put in enough time and effort on design, are willing to spend the time forcing the AI to get it right, and don't settle for less than what you envision, it could work.
That things come out perfectly working without need for QA and refinement cycles when writing the code manually seems very unlikely, and the case where coding it yourself is faster than explaining the task to the AI does practically not exist.
> as the project grows there's a risk of it becoming unmanageable
I suspect that what you said there is not based on actual experience but rather something you imagine. In my personal experience AI has shown no issues working efficiently on projects with even 500k LOC.
This is a very important observation. This is fundamentally about personal computing and the power to control the hardware and software we use for our work.
I recently reread "What The Dormouse Said", by John Markoff. Superb book that describes the tension between the developers of prototype personal computing on one side and distributed computing on the other. Interestingly, the need for distributed computing was motivated in part by early artificial intelligence research.
What we're seeing today IMO, is the reemergence of distributed computing. We saw it first with storage but we're now seeing it with compute itself. Given that excellent personal hardware has been available for several decades, it seems like madness to change course.
It's exciting and kind of scary.
Naturally, decompiling a binary would be simpler comparatively, but for models of sufficient capacity it seems the 'simpler/complex' axis has been shrinking.
It is more likely that we will see a cambrian explosion of games that obviates the need for big company games in the first place. Big gaming companies are big because they need lots of developers, producers, artists, QA, and management and executive suites to manage those. You can even today get code, assets and QA for a tiny fraction of the price you have to pay humans, and that will apparently only get 'worse' for humans over time. It's going to be a very intriguing next few decades.
Why would this be the future you want?
At least as far as games are impacted, that wouldn't matter to me whatsoever.
I mean triple a productions have been complete dumbsterfires for over 10 years now. I don't recall a single high budget game since 2015 that was just good, and while gta6 will break records, it's still almost certainly gonna be a mid game overall.
And saas in software has already been the default for way over 10 years, so not sure what you man by that.
Last weekend I had it decompile the "Kitten Cannon" Flash game and then ran (what I believe was) an exhaustive check of the 36 trillion reachable launch+power+rng combinations to see what the longest run was (I got a little over 11k feet, successfully replayed in the Flash version).
It's so cool to see modifications/enhancements to nostalgic things appear for next to no effort as much as I want rather than be limited by the time I have for it.
Maybe the next time I spend some tokens on it I'll ask for it to convert the implementation to a "proper" runtime rip where the user can provide the source ROM and it extracts those assets live.
:)
For everything else .. well, it's hugely useful to be on one side of it. I was just today getting codex to explain why the Mac and Windows versions of a library I am using behave differently, which it does by decompiling them. We've yet to see how this pans out downstream, but it's quite likely there will be some sort of retaliatory closing of hardware to attempt to fight it.
https://flashpointarchive.org/view?id=54d974f5-51f8-42c8-a43...
I look forward to AI decompiling and making available on modern MacOS all the Intel 32-bit games that aren't compatible with Intel 64-bit, let alone Apple Silicon.
Heck, extend that to PowerPC and even 68K games. Yes, emulators have been available for such for a long time. But that's not the same thing as recompiling natively onto Apple Silicon (or whatever comes along in the future) and adding support for modern resolutions along the way.
It traditionally involves investing incredible amounts of time and effort and intelligence, but what I think AI does most encouragingly is allows people to access more complex, more modern, more high-performance technologies and interact with them in the same creative and explorational way.
Over the weekend, I was curious about the state of open source touch-based HID support as it compares to the state of proprietary market leaders. I tasked an agent with decompiling and analyzing firmware blobs from the systems I was interested in, and now I have an early alpha version of the most robust source-available multi touch implementation.
There are obviously some legal and perhaps even ethical considerations worth reviewing before considering publishing anything like this, but we have reached the point that proprietary software is essentially open source.
I ended up coding coding up a fully interactive version stream on twitch. Unfortunately the Twitch extension I made to make it interactive is still pending review by Twitch, but I can manually add anything keen enough to the allow list. https://www.twitch.tv/medic_911/ (edit, I've switched the stream to use the existing twitch interactive ext, GreenHeat, so you can play the game now, it's just not optimised for this game like my ext is)
I also converted one of the existing decodes of Simtower out there into a native macOS app, which is pretty cool.
I find it (AI) really exciting, but also scary how this will shape the future. People still need to make money somehow to be able to spend it.
It is great in terms of what you can do, but sad in terms of what you can be.
If Windows dies, Linux gaming will kind of go down with it. Existing games will continue to work under Wine of course but I doubt developers will do the upkeep to keep their games working on a dead platform.
Considering they just announced a laptop where the sole focus is running AI locally it definitely seems like they're not interested in letting Windows die. And anyway a lot of their money comes from enterprise where they rely on Windows to continue running some shitty old software that hasn't been updated in 30+ years. It's not going away anytime soon.
> I doubt developers will do the upkeep to keep their games working on a dead platform.
They don't need to, communities make mods to make the games continue to be playable. Dragon Age Origins needs a community-made patch to stop the game from crashing on modern systems.
And now with AI you don't even need a particularly technical skillset to do fixes like this yourself.
Single player PC games are a small fraction of the market these days, and once people have a job, they are generally happy to pay for the things they like.
I think the "clean room" Photoshop example is much more interesting because Photoshop is far more expensive and Adobe has a semi-antagonistic relationship with creatives.
What I hope is that we don't just end up with a bunch of vibe coded clones, that may or may not contain AI introduced bugs or even malware, that get uploaded without being run/tested and then become instantly abandoned. Instead, I hope that AI's ability to help decompile software helps people develop open source projects that are maintained and further developed to go beyond what the commercial software was capable of.
Demos are (up until this week) the #1 for studios to do marketing.
Marketing will change now, to more socials being required (YT etc), and even then only small bits here and there as decomp will happen from video as well.
The marketplaces that adapt to this and enable devs will help, hopefully.
In order do decompile you already need a game. If you usually buy games you'd still buy one. If you pirate games - you'd still pirate it.
So doesn't seem clear to me if there is a lot of lost revenue here?
And pirates don't really need AI do hack a game etc. All possible games have been on torrent trackers since forever.
Possibly ability to decompile and modify it would be a free marketing (from those who mod it and post videos) and only increase sales. Potentially what games should do is optimize builds for this so that mods can be easily created and shared - e.g. you buy a game and apply a bunch of mods created by ai or not.
The only reason I see is when it can hurt sales if some cool mod someone does can't be easily applied to original game and then you have no choice but to pirate it in order to play that mod.
Normally they would've released the "enhanced edition" for PCs next year right around the same period of time. With the current pricing on consoles, there's a good chance a ported PC version would even cannibalize sales of their console editions.
As a thought experiment, how many companies do you know that make enough positive revenue to recoup development and marketing costs from a completely open source download that doesn't rely on something other than the open source software itself for profitability, like a proprietary component or some kind of hosting or network or human support service to drive value? I can think of none.
Between feeds, the slopification of discourse and the transition to streaming pixels, we're now looking at TV 2.0
Virtually all routing equipment is built on top of open source software
Links aren't as permanent as TBL wanted to, but are crawlable and there are public archives which can (still) manage the storage requirements
If you think the internet is already dead, I think you may need glasses
Eh. That makes no difference to the progression I'm talking about here. The internet is _already_ click-on-demand so to speak[0]. Software is mostly not yet. But it will soon also be.
[0] - You could replace all hrefs and documents with SPAs that render a pixel stream to canvas and almost nobody would ever notice. After all, the ultimate interface between your computer and your face is pixels (or waveforms), not hypertext.
A small percentage of people have always been the only ones that care about anything technical, it's just a consequence of division of labor. That's why expert opinions used to matter.
Streaming pixels would make archiving it all practically impossible. It would also be the fatal blow in the war against general purpose computing. I'd ask you whose side you're on, but I'm pretty sure I already know.
I'm on the side of being extremely pessimistic about the future viability of making software for a living. This thread is not a moral statement, just an observation and prediction.
And your parent post, with all of its significant edits is an example into why archiving is important.
My greatest moral failing is pressing done before I'm done.
> why archiving is important
Unfortunate things happen all the time. We'll need to find new ways to archive.
> My greatest moral failing is pressing done before I'm done.
> > why archiving is important
> Unfortunate things happen all the time. We'll need to find new ways to archive.
In this case, the unfortunate thing was that you got caught.
You greatest moral failing is that you have no morals, which is why your primary concern is "how will _I_ earn a living tomorrow?"
As for the new ways to archive, I'm certain that LLMs are your proposal, hence my not bothering to ask whose side you're on.
What a weird thing to say. Do you always stick so hard to assumptions about people that are diametrically wrong instead of asking or listening to what they say, or is this a new thing? I guess now we've found your great moral failing too. :/
I was shocked YC accepted my fully FOSS company, but YC and a lot of investors realize proprietary software is no longer a moat, and some VCs -exclusively- fund FOSS companies now seeing them as the future.
I cannot express the amount of joy and hope it gives me in this dark world that every proprietary binary anyone has ever found useful or entertaining is about to be reverse engineered and have all the tracking bullshit users do not want removed. Only useful or entertaining features will remain.
This is about to completely break a proprietary and surveillance obsessed software industry that deserves to be broken.
I am thankful to have a front row seat in Silicon Valley, and look forward to watching every single company be forced to Open Source or die.
Everything we do at Caution is FOSS right down to our website and frontend.
People pay us because we maintain an obsessively strict security model of code review, code signing, full source bootstrapping, and trusting no single human or machine. For any of our customers to hold those standards they would need to hire a minimum of 3 experienced systems/security engineers, and paying a subscription to us is a way cheaper option.
If you want people to pay you, even for open source, you need to provide support and value no AI can match.
That is a level of extreme agency over software that was once only available to security engineers with decades of experience like me, that is now available to almost everyone.
It is not a joke! The naming is so bad so it sounds like a joke. So people like the author assume it is a joke and move on without suing them. They are counting on that.
RSG usually waits for a year before releasing to PC, but this will force their hand.
Now we will see a renaissance of creativity, beauty and fun games. Pre-AI games like this would have just stayed in the brain of this one guy forever. https://www.instagram.com/reel/DeMWHApB5Fh/
I do wonder what this means for open-source versions of games, as in: could companies threaten with legal actions more easily now?
I don't think it's the technology OR the people that the author doesn't like—I think it's what the people produce with that technology. Namely, the slop and bots currently overwhelming the internet.
What is true is that the price points are going to be affected, the economics have changed.
I just got let go from my job, and I was thinking yesterday about trying to start a startup around a piece of software I wrote on the side that I think has a market. But I also know that if my marketing page gets posted on HN it’s curtains for that business (since the main thing I’d be selling is software).
I think we can all agree that software is valuable, in that it provides value to us. And I know a lot of us hate when a company that should be “just providing software” tries to “pivot to services” or find another unnecessary way to get their hooks into us. But if making good software is no longer a way to make money, then the only companies left will be the ones providing “services”.
Why can't you have an edge in marketing your product?
In fact, seems like you assume they'll have the edge.
You probably have a network, too--even if small, it puts you above 99% of others.
Ship it!
They don't need to make all the decisions, they will take what the tool spits out and that will be the "content". For more content, they will just say, "Make an expansion and do something with black holes" and it'll be done.
Why would I buy your game when I can make my game exactly how I want it.
Add in small multiplayer, 2-6 people, and now you are basically playing modern WoW.
There is no creative decision making. And in fact, the less creativity, probably the better result--because the models are great with an exact specification.
These publishers will find a way to lock it down for new games and tie it to their servers to run it, or have the game streamed to you; simulated via a world model or not.
There's no way to deliver a running program which you can't decompile aiui.
Like literally last night I did this to decompile an Android game whose APK on the Play Store is just a thin wrapper that downloads the real Unity game at runtime. It was a 30 second barrier
This may be straightforward on Android, but is it on iOS?
"No way to deliver a running program which you can't decompile" seems less like a rule of nature than the way it happens to work now. I don't think it's extreme to imagine that likely a side effect of this AI-driven gold rush to "free the IP" will be the further lockdown of general purpose computing.
Astra Max can do 12%: https://balrogai.com/
Though after improving the harness it can fare better, but still no completion: https://roger-creus.github.io/blog/astra-craftax-nethack/
It's surprisingly powerful.
I knew it was over when we used automatic place and route to synthesize entire CPUs in Cadence in my ECE classes in 1999 before I graduated with my degree. It already achieved 80% of the circuit density and speed as humans. Then read Genetic Programming III by John Koza 5 years later, which explained that each computer 10 years older than the previous generation, running about 100 times faster due to Moore's Law, could be made to write previous gen software automatically.
Unfortunately, GPUs don't help much with the kinds of optimization problems that I was interested in, where we really need many thousands or even millions of independent threads exploring enormous search spaces using multicore CPUs with local memories. Gradient descent is one way to do that, but IMHO it's one of the hardest ways, requiring the most bare hands effort and brute force. That's why neural nets that actually worked took 25 years to manifest.
So imagine knowing how to basically grow computer hardware and software organically with self-training/improvement in the late 90s and early 2000s, only to be stifled by the end of Moore's Law arriving in 2007 when smartphones chased the profit in cost/power reduction of CPUs and devoted most all remaining R&D to faster SIMD GPUs which didn't help MIMD (multiple instruction multiple data) workloads.
After working so hard for so many years and decades, only to tread water making rent, some part of me welcomes watching the world burn. Now everyone else will get to enjoy the precarious feeling of having the skills to do real work, only to be forced not to at some BS job that almost pays the bills.
Unless you all prove me wrong and put your potential to real use, and bring time/money/resources to the working poor, outside of the current wealth inequality pyramid scheme we pretend to call capitalism.
Maybe I sound bitter, but I'm feeling oddly inspired and optimistic for the first time in a quarter century.
Thank you AI for ruining more of our tech freedoms.
I'll be playing the ones I control.
Only one of those gets cheaper and faster with time. (And that's just me guessing it will get cheaper. I could be wrong.)
In any case, the other two scale up with the complexity of the game being streamed. The AAA studios can simply overwhelm with complexity and content. Especially since AI will be generating an increasing majority of the game anyway as time goes on. If it's playable for a year's worth of hours, good luck downloading enough video to replicate it.
Maybe Twitch or youtube could if they incentivised enough streamers to stream their play sessions? Of course, why would Twitch or Youtube give that dataset to you?
It's just not as easy as people are making it out to be.
I want Animal Crossing to have an infinite world. I want it to be an MMO. I want the animals to have much better intelligence - for example, if I'm a bad neighbor, maybe they should team-up against me.
Zelda MMO will also be cool. It's just a matter of time. Like two weeks.
These are easy to do now. It's an amazing time to be alive.
But those that will persist through the awkwardness, might find gold. Zelda Four Swords was a banger after all
I'm working on a game right now that tries to find a nice balance between player made stuff and gorgeous assets they can't change.
How are planning to block that? If they can create Zelda MMO, they can modify your game as much as they want as well, right?
They can create their own world but then they need to get people to migrate to it.
Like just to start,how are you going to balance intelligence so that it's fun? The AI doesn't magically know.
An unbalanced, untested Zelda MMO will be out shortly. Watch.
Some fans are tired at the lack of innovation from the IP holders and are taking fan games in interesting directions. Look at this for example:
https://undreamedpanic.itch.io/gamma-emerald-ea
Loads more here:
https://www.ign.com/articles/zelda-fan-ditches-ocarina-of-ti...
I think we're going to see real quick how this is going to go with those Adobe clones. Clean-room is supposed to be legal. I have a feeling it's about to stop being legal.
But then that will make PC-clones illegal: meaning all modern computers. It's a curious problem for them.
If you clean room reverse it, the mechanics in a Legend of Zelda game are, for the most part I think in the clear legally speaking. But part of what makes Zelda feel like Zelda (IMHO) is that you’re fighting familiar monsters in a familiar setting with familiar music aka all the lore, otherwise its you've just rebuilt an open world exploration game with dungeons.
However if you borrow lore like even if you change the names of all the monsters, you may still run into legal risk if they remain visibly recognizable. Even if you redraw them in a different style, if they’re still identifiable as Octoroks/Moblins/etc, that might still open you up to litigation.
There was a similar case a while ago, Tetris v. Xio, which might be worth looking into, since I think it sets some useful precedent.
https://en.wikipedia.org/wiki/Tetris_Holding,_LLC_v._Xio_Int....
But to be honest - I just want an open world MMO. I don't need Octoroks, etc.
In fact..... for the last few weeks I've been working on this sort of thing for Animal Crossing. I replaced all the assets with little felt toys. (meshy.ai makes fantastic 3D meshes). But the code only got halfway done before the onslaught of stuff hit this last week... so I lost motivation. Also the clean-room thing is really expensive.
- Deck of cards
- 3d printers
- Printing services
I think the biggest things to come out of asset reuse / engine / remixing are probably DOTA (WC3), ARMA/Day Z, and to a lesser extent Garry’s Mod. Which are the exception that prove the rule given that there have been so many games and engines that have lowered the barrier for game creation.
Edit: maybe you can also give credit to StarCraft for spawning the Tower Defense genre (Turret Defense was the original).