Unfortunately, this is only going to push Sony/Nintendo/MS more towards cloud gaming. AI enabling these amazing reverse engineering efforts and democratizing access to gaming is amazing, and preventing vendor lock in is always something I'll support. However it's hard not to see them pushing gamers towards needing to play games on cloud only, preventing these kind of things happening.
Steam is larger than any single console now a days, has a DRM which can be trivially bypassed, and still does excellently. The problem with consoles is that they're not economically competitive and rely instead on user lock-in paired with marketing and the antiquated belief that PC gaming is complex or requires a $1000 computer. Both were, more or less, true 20 years ago, but haven't been true for many years.
It's never wise to say never, but I think cloud gaming will never go far with gaming. Physics means there's going to be input latency. So it simply won't work for VR where you really need at least 90FPS with no input latency. Of course the year of VR seems to be coming just about after the year of Linux, so maybe that doesn't matter. But even for normal games, the input latency is major. This [1] article shows input latency of more than 1/4th a second on one game ran at 4k. Then there's reduced visuals and of course recent times have also demonstrated beyond any doubt that the whole argument that the cloud won't go down, or won't lose your data, are both marketing fictions.
This continues to be an absolute no-go for competitive real-time games due to input latency. I honestly think there's a chance that game clients might go the way of the recording industry, and they just might not be worth that much.
As doubtful as I am about the viability of cloud streaming, this IS going to cause a bigger push into live service functionality where maybe you have a light client but the server is authoritative over more and more aspects of the game. Even then, AI can write its own server, so it needs to be full of FOMO if you're not on the official server.
I don't even know if a smaller company like Nintendo can pull something like that off. They have been behind so much in the tech space because they spend modestly and their primary differentiating factor is their IP. Their lawyers are going to be working non stop until they figure something out. I guess have a strategic partner?
Cloud game streaming has gotten to be extremely impressive, at least on a wired connection, to the point where it’s almost hard to tell you’re streaming. It’s been hard for me to consider upgrading my old gaming PC considering how good GeForce Now is for the $20 a month that I spend.
Well Microsoft always seems to be making the wrong decision at the wrong time and this is probably yet another one. Probably due to competing factions hat don't really have a unifying vision for the company. It provides enormous amounts of comedy and I guess we will find out if this was another instance of "whoops we closed down that thing we could really use right now"! :)
Is there any analysis of the profitability on this? They made such a drastic cut to availability it seems like: Microsoft wants out of the cloud streaming entirely, they are burning money, or they are so hardware strapped they would steal form the gaming division to feed the Copilot usage.
Well isn't Roblox kinda this? I play it all the time with my niece and nephew...it sucks, its enshittification to the max. Part of an entertainment property is you are devoting time to enjoying the creative process of someone who put in the effort to produce something polished and with effort. The end result is the culmination of all the work done (possibly years/decades) to get to that point and you are paying for the result of all that effort. Could a Roblox game be like that? Possibly but the incentives dont seem to be as aligned despite the democratizing force it has had.
I am reminded of a scene from the Jobs movie where they were arguing Open vs Closed systems. This feels similar...
I could in principal see something that isn't clearly entirely derivative, like if a person takes their own original thought and uses generative AI to make it into a video instead of, for example, learning animation and animating it or finding (probably amateur) actors to film it with. This of course still has some derivative aspects that may be easier to avoid than in the non-generative-ai case, but it is also clearly distinct from a 100% generative AI case.
For example, ever since roughly the mid 2010s when I finished Star Trek: Enterprise, I've been thinking about someone using some sort of generative AI to complete this series, once its good enough to be entirely convincing.
There is an AI-generated YouTube series/channel called "TK-42 Greg" that follows the life of a hapless Star Wars stormtrooper as he pursues a career of menial jobs, unemployment and drifting about a galaxy far, far away.
It's dumb and derivative and isn't getting a lot of laughs from me, but at the same time, it is freakishly realistic, and absolutely ripping off every character and scene and locale of the Lucas series.
I don't know how long it will last against Disney's DMCA juggernaut, but the "Imperial HR Lady" is way funnier and she just has to throw on a turquoise necklace and grating accent.
Speaking of self-vibe-coding games, I wouldn't see the point. When I was a younger boy, I did write games for myself, and it was an arduous project that was mostly aimed at teaching myself programming rather than gaming per se. And eventually the big didactic moments were hacking someone else's game code in terms of debugging and extending it, and that was more exciting than building my own from scratch.
Also, copyright is an issue! If consumer AI could ignore copyright and just code me, say, "Freecell but it's Harry Potter" or "Put me in the cockpit of a Y-Wing video game!" then it'd be exciting. People pay big bucks and invest big bucks for in-universe games of certain franchises like that. But AI won't provide for the consumer in this way, ever.
Much of the joy of gaming for me, is seeing a polished product that is well-balanced in terms of rules and fairness, and has good sound/graphic assets, and challenges or surprises me in novel ways as I go along. I don't think I would mind if someone else vibe-codes a game for me to play, if it's good enough and engaging, but it seems that designing/coding one for myself would feel empty and unfulfilling. And I wouldn't apply the same principle to utility software, but really I'd rather not invest a lot of time in building an office suite for myself, rather than just grabbing it off the shelf.
I've had AI make little learning games, coloring stories, etc for my kid. I and it absolutely will ignore copyright. So far it's never objected. It'll also happily reverse engineer binaries, extract art assets, whatever.
It'll happen regardless, as of now the public internet infrastructure still isn't fast enough to be comparable to local rendering, cloud gaming needs to deliver >50mbps bandwidth at sub 16ms latency.
How would cloud-only prevent this? Do you think LLMs can't exfil the game from the server? They're incredibly persistent and are willing to try many things. Random credentials, vulnerabilities in the game servers / dev machines, or even social engineering the devs/the deployment.
Unfortunately, I don't know if you've tried cloud gaming. It's just not very good now. Also to force people to use it given the current cost climate for components, it would be extremely expensive.
Also a use case for world models to hide the actual source of the game and have it stream a simulation of the real game; never to be exposed or downloaded.
But nobody will play it? Steam exists and all those games get pirated. You can’t force gamers to buy duds. I have a gigabit connection wired and found a cloud version of Control to be totally unplayable.
It is in the realm of design to make games playable within the context of their technological limits.
In this case, I could see games moving towards being turn-based i.e. online chess.
Revenues suffer because game studios and publishers make awful decisions. See the botched releases of Marathon (Bungie) and the more infamous examples — concord and split gate.
I honestly welcome cloud gaming.
Action RPGs might be dead in the water,
But there were many games suffering from data mining, and having their secrets revealed day 1.
Bringing games to cloud gaming could restore the mystery in games
This is saying you'd be ok with books being only permitted to be read at the library, behind a sealed kiosk screen, because otherwise people might post spoilers online.
If the library is completely AI generated, and only the sealed kiosk screen contains (possibly) human-authored content,
How does the equation change?
I am saying I'd rather enjoy POSSIBLY human content, than AI generated ones. And if cloud gaming cannot deliver on this, I'd rather stop gaming altogether. The idea of consuming AI generated content is completely disgusting to me.
Also if that's what you believe, players who start after day 1 will not have mystery. Even players who start day 1 but can't dedicate all of their time to the game will inevitably spoil themselves when they have to do anything else, meaning someone will progress farther than they and will share that info.
Perhaps, but I doubt it. The immersiveness and suspension of disbelief come partially from the richness and massive bandwidth of the data stream players receive. Streaming degrades it massively. Even with media streaming, e.g. Netflix, there are many folks (including myself) who can't stand the low-ish bitrate they _have_ to supply to customers because of the infrastructure strain / $REASONS. With games it will be even worse. And there's also the latency problem, which is kinda massive for this. Unless of course they invent some kind of protocol where assets are stored and loaded on the console, but the logic of the game is received via a low-bandwidth stream.
Regarding your post: I think the future of RE of server-side bound software will be something like 1. Point your AI robot with eyes and ears to an interface designed for you (the human customer). 2. Make it explore software at massive speeds, map and record every possible behavior, closely construct/approximate the human-computer interaction model. 3. Rewrite it completely in Rust. 4. Profit.
There will be no moat. Popular non-FOSS will be laughed at and rewritten by its users in hours/days. Proletariat will finally own the means of production, and the Stallman dream will finally become real.
Your post makes sense, but most software was moving to the cloud even before AI. Why sell someone a copy of ms office when you can get them to subscribe to O365?
I think the self hosting moment will gather steam and maybe prove itself superior for certain cases
in a few years, won't I have an agent that can play the game via the video stream and then code my own local copy of the game? it might have to iterate through story branching, but that's hardly an obstacle
You think people play Dark Souls because it is "hard" and "dark fantasy"
Or do people play it because the gaming community at large plays it, and people want to join in the conversation and have fun with the community?
High quality games were never the point, look at Roblox
Nobody wants 15 year lead times for sequels that cant innovate because its too risky. Their audience will also have a major backlash towards any generative AI use or known agentic workflows in development
All while indie developers wont be constrained by any of this and can even do a full voice suite thats more immersive and even dynamic, in a smaller payload because there wont be audio files, just the voice model
I think streaming vs downloadable is a red herring, AAA games are on their last breath. The indie developers that break through will be the AAA experiences if the market chooses a title enough times
The Hollywood equivalent to this is to make only AAA blockbuster titles in order to drive theatrical ticket sales. It is a strategy with diminishing returns - just look at the box office trajectory over the last 25 years.
It's always nice to have local git clone backups/mirrors of these projects because they may eventually be taken down through legal threats eg, Yuzu and Ryujinx.
Nice moment to hook the git-maintenance[1] command, which setups automatic pre-fetching of a repository in a scheduled manner. That way you don't need to remember to manually update your mirrors. Since the repo is public, cloning with https is a simpler setup for prefetching, but ssh is also supported with some configuring.
From the amount of activity and how much this all smells like heavy AI usage, I have zero doubt that something will be working shortly after launch. The agents won’t sleep until it boots.
This project is the first hint I’ve seen of the complete destruction of all copy protection schemes. Scary times for any industry that relies on those!
GTA 6 is releasing on PS5 right?
Who is going to pay a single cent for software if everything can be ripped day 1?
Countries with a huge stake in the software industry are going to get crushed?
Most software and games is available pirated already and has been for decades.
You could get pirate games on the original PlayStation two decades ago. Cracking the most recent PlayStation is not going to change things
Piracy is a service problem! If they release a good PC port, I will buy it, as I have done for nearly every other Rockstar game. Hell, if they were to even just automate the port using a similar tool, I wouldn't really mind if the end result were indistinguishable from a decent PC port. I just really don't want to buy a PS5 for a couple exclusives.
I think the replies to my comment are not realistic.
The Goldilocks zone for earning revenue for movies & games were always: Day 1 earnings = Week 1 earnings = Year 1 earnings
The revenue in this case is heavily protected by your IP NOT being copied in a single day, and that people want to experience the product together with their friends.
When people talk about convenience, is there really any inconvenience in downloading a pirated game/movie on release day?
Especially for GTA6. The most amount of money is riding on this. It's still going to be a lot if it's copied on day1, but there is no way it is near their projections.
Do people pay only because there's no pirate alternative? Valve showed this with Steam, people pay for convenience and for good service quality. Piracy is an answer to lack of corporate respect about culture and art that's used to create a commercial product, which manifests in enshitification.
False, myth. This is a common claim pushed by MPAA/RIAA (the copyright police for music and film industry).
Consumers want _access_ to their games/content. For example, netflix made it easy to "rent" or stream content. But as they slowly clawback those features (multi-streaming, password sharing, incorporate ads, increase prices beyond inflation). Then access to content decreases, visibility decreases.
There’s also a big factual error in the submission title because the project readme has this caveat
> System libraries: percentage of the functions known to the project so far (declared in core/libs/prx), not of every PS5 system function. The total grows as more functions are declared.
I always wonder who the folks who build these things are. If they become too popular, it seems like a good way to have the government pay you a visit you will not enjoy. But they also seem like technically interesting problems to solve. Glad someone is out there preserving these games in any case.
They might be technically interesting problems to solve but these days they're just using AI to solve them so they aren't really learning anything about those interesting problems.
A lot of people live in places where the government does not really care to use police to enforce copyright, or where this just isn’t really considered illegal.
The PlayStation platform in general is the easiest console to reverse engineer. There’s a lack of security in depth that makes it easy to reverse engineer.
It’s only recently (earlier this year) that the _last gen_ of xbox was cracked.
Since the PS5 is already x86-64, the main friction point here is likely the proprietary graphics API rather than CPU instruction translation. I wonder if they handle GNM command buffers by pre-compiling them into Vulkan during that static relinking phase or if there is still some runtime interception happening when the game tries to queue those draw calls.
It's never wise to say never, but I think cloud gaming will never go far with gaming. Physics means there's going to be input latency. So it simply won't work for VR where you really need at least 90FPS with no input latency. Of course the year of VR seems to be coming just about after the year of Linux, so maybe that doesn't matter. But even for normal games, the input latency is major. This [1] article shows input latency of more than 1/4th a second on one game ran at 4k. Then there's reduced visuals and of course recent times have also demonstrated beyond any doubt that the whole argument that the cloud won't go down, or won't lose your data, are both marketing fictions.
[1] - https://www.pcgamer.com/heres-how-stadias-input-lag-compares...
As doubtful as I am about the viability of cloud streaming, this IS going to cause a bigger push into live service functionality where maybe you have a light client but the server is authoritative over more and more aspects of the game. Even then, AI can write its own server, so it needs to be full of FOMO if you're not on the official server.
There goes 95% of your market. If not more.
I am reminded of a scene from the Jobs movie where they were arguing Open vs Closed systems. This feels similar...
[1]:https://youtu.be/TYvXOHAJTiU?t=81
Ugh I know the answer but it’s just sad. I don’t want all media to become cotton candy.
For example, ever since roughly the mid 2010s when I finished Star Trek: Enterprise, I've been thinking about someone using some sort of generative AI to complete this series, once its good enough to be entirely convincing.
Two quotes from Star Trek "The Ultimate Computer" - a very prescient episode.
KIRK There are certain things men must do to remain men. Your computer would take that away.
---
BONES Compassion. That's the one thing no machine ever had. Maybe it's the one thing that keeps men ahead of them. Care to debate that, Spock?
SPOCK No, Doctor. I simply maintain that computers are more efficient than human beings, not better.
It's dumb and derivative and isn't getting a lot of laughs from me, but at the same time, it is freakishly realistic, and absolutely ripping off every character and scene and locale of the Lucas series.
I don't know how long it will last against Disney's DMCA juggernaut, but the "Imperial HR Lady" is way funnier and she just has to throw on a turquoise necklace and grating accent.
Speaking of self-vibe-coding games, I wouldn't see the point. When I was a younger boy, I did write games for myself, and it was an arduous project that was mostly aimed at teaching myself programming rather than gaming per se. And eventually the big didactic moments were hacking someone else's game code in terms of debugging and extending it, and that was more exciting than building my own from scratch.
Also, copyright is an issue! If consumer AI could ignore copyright and just code me, say, "Freecell but it's Harry Potter" or "Put me in the cockpit of a Y-Wing video game!" then it'd be exciting. People pay big bucks and invest big bucks for in-universe games of certain franchises like that. But AI won't provide for the consumer in this way, ever.
Much of the joy of gaming for me, is seeing a polished product that is well-balanced in terms of rules and fairness, and has good sound/graphic assets, and challenges or surprises me in novel ways as I go along. I don't think I would mind if someone else vibe-codes a game for me to play, if it's good enough and engaging, but it seems that designing/coding one for myself would feel empty and unfulfilling. And I wouldn't apply the same principle to utility software, but really I'd rather not invest a lot of time in building an office suite for myself, rather than just grabbing it off the shelf.
Considering the economics, the graphics rendering would have to be shifted to the server too and output as video
Sony's greed has been incredible, let's just accelerate tbh
I have only had success with single player games.
How does the equation change?
I am saying I'd rather enjoy POSSIBLY human content, than AI generated ones. And if cloud gaming cannot deliver on this, I'd rather stop gaming altogether. The idea of consuming AI generated content is completely disgusting to me.
Also if that's what you believe, players who start after day 1 will not have mystery. Even players who start day 1 but can't dedicate all of their time to the game will inevitably spoil themselves when they have to do anything else, meaning someone will progress farther than they and will share that info.
Funnily enough I ended a recent blog post on this topic with "You will own nothing and you will be kinda pissed off about it."
https://www.da.vidbuchanan.co.uk/blog/reversing.html
Regarding your post: I think the future of RE of server-side bound software will be something like 1. Point your AI robot with eyes and ears to an interface designed for you (the human customer). 2. Make it explore software at massive speeds, map and record every possible behavior, closely construct/approximate the human-computer interaction model. 3. Rewrite it completely in Rust. 4. Profit.
There will be no moat. Popular non-FOSS will be laughed at and rewritten by its users in hours/days. Proletariat will finally own the means of production, and the Stallman dream will finally become real.
I think the self hosting moment will gather steam and maybe prove itself superior for certain cases
High quality games were never the point, look at Roblox
Nobody wants 15 year lead times for sequels that cant innovate because its too risky. Their audience will also have a major backlash towards any generative AI use or known agentic workflows in development
All while indie developers wont be constrained by any of this and can even do a full voice suite thats more immersive and even dynamic, in a smaller payload because there wont be audio files, just the voice model
I think streaming vs downloadable is a red herring, AAA games are on their last breath. The indie developers that break through will be the AAA experiences if the market chooses a title enough times
Eg. https://torrentfreak.com/nintendo-wipes-out-400-switch-emula...
Edit: wrong link
[1]: https://git-scm.com/docs/git-maintenance
This project is the first hint I’ve seen of the complete destruction of all copy protection schemes. Scary times for any industry that relies on those!
The Goldilocks zone for earning revenue for movies & games were always: Day 1 earnings = Week 1 earnings = Year 1 earnings
The revenue in this case is heavily protected by your IP NOT being copied in a single day, and that people want to experience the product together with their friends.
When people talk about convenience, is there really any inconvenience in downloading a pirated game/movie on release day?
Consumers want _access_ to their games/content. For example, netflix made it easy to "rent" or stream content. But as they slowly clawback those features (multi-streaming, password sharing, incorporate ads, increase prices beyond inflation). Then access to content decreases, visibility decreases.
There’s also a big factual error in the submission title because the project readme has this caveat
> System libraries: percentage of the functions known to the project so far (declared in core/libs/prx), not of every PS5 system function. The total grows as more functions are declared.
It’s only recently (earlier this year) that the _last gen_ of xbox was cracked.
https://en.wikipedia.org/wiki/Portable_Executable