“If buying isn’t owning, then piracy isn’t stealing.”

That line captures the anger behind the recent PlayStation discussions. In responses to AnyPS5’s progress, you can find people openly saying they would feel worse for Sony if it were a different company. I think that reaction deserves attention. A technical project is being treated as a correction to a relationship people already feel has become unfair.

The connection I see between Sony’s disc decision, AnyPS5, and ArtCraft is a bargain coming under pressure from both sides. Companies want continuing authority over what customers pay for. AI-assisted development raises the possibility that outsiders could recreate more of those products without the original company’s involvement. The first weakens people’s sense of ownership. The second could weaken the practical exclusivity companies have depended on.

If both continue, we will have to work out what people owe each other when neither buying nor building something guarantees the control they expected.

The loss of trust

On July 1, 2026, Sony announced that physical disc production for new PlayStation games would end in January 2028. The change does not affect games released on disc before that date. Sony presents it as a response to customers’ preference for digital formats. Its U.S. terms separately say that words such as “buy” and “ownership” do not transfer ownership of the content, and describe the license as revocable and non-transferable.

I understand the frustration with that combination. The purchase is real enough when the money leaves your account. The thing you receive comes with a much more conditional promise. For someone who thinks of a game as something to keep, the disappearance of the disc removes a familiar way of having a copy under their own control.

Digital distribution can preserve that practical independence. GOG’s requirements include offline play without its client, and it provides standalone installers. A disc and a download are delivery formats; the important design decision is how much permission the software needs after you have it. Treating the loss of control as an unavoidable consequence of going digital lets that decision escape scrutiny.

For me, the bargain should be fairly understandable: I pay for someone’s work, and they give me something I can continue to use on the terms I understood when I paid. A service can have recurring costs. A multiplayer game needs running servers. Those obligations should be clear at the point of purchase. The problem comes when the language of a lasting purchase meets the conditions of a temporary relationship.

That is how I read the piracy slogan. It expresses a refusal to accept an obligation that feels one-sided. It doesn’t resolve the legal questions or tell us who gets hurt. It does suggest that some customers have stopped seeing the company’s restrictions as something they have a reason to defend. Enforcement can survive that loss of trust. Loyalty is harder to recover.

What AI changes

The projects attracting attention in September and October 2026 approach this from different directions. AnyPS5 converts PS5 executables for Windows and Linux and provides replacement system libraries, aiming for native execution. Compatibility is still tracked by game; the project is not evidence that the entire console and every game now work on a PC.

ArtCraft’s developer, publishing under storytold, has described using Claude to build open-source alternatives to Adobe applications. The effort now includes WordCraft, GridCraft, and DeckCraft for Office-style work. These are new implementations with substantial ambitions and unfinished work. PhotoCraft’s own documentation calls it an early alpha, unsuitable as a daily professional Photoshop replacement. The developer’s announcement and the projects’ current limitations both belong in the story.

Reverse Engineer Anything gives coding agents access to tools for investigating existing programs and using that evidence to write implementations. I haven’t verified that it was used to create AnyPS5 or ArtCraft. The connection I’m drawing is about what these efforts could change, rather than a claim that they share an origin.

Recreating useful software has always required more than the desire to do it. Someone has to understand unfamiliar behavior, implement it, and spend time on the cases that initially look unimportant. If agents reduce that cost, a product that once looked impractical to challenge may become a reasonable project for a much smaller team. The early demonstrations make that possibility worth taking seriously; they don’t establish how far it will go.

That could change the balance between a company and its customers. A restriction has a different weight when rebuilding around it is conceivable. Frustration that once ended in a complaint might help sustain an alternative. A company could retain its legal claims while losing some of the practical control that made those claims commercially powerful.

The part karma leaves out

I understand the satisfaction of seeing that happen to a company you feel has treated you badly. But a cheaper way to recreate software has no built-in preference for deserving targets.

The same capability could put pressure on an independent developer who sells a thoughtful product on reasonable terms. Rebuilding can begin with a successful example: the useful features are visible, the workflow has been worked out, and someone else has already paid for the failed attempts. Reproducing the implementation more cheaply does not make that earlier work disappear. It can make it harder for the person who did it to recover the cost.

There is a distinction here that I want to keep intact. Helping an existing game run elsewhere preserves the game’s creative work. Reimplementing an editor replaces part of a tool’s functionality. Distributing someone else’s code or assets raises further questions. Putting all of that under “piracy,” whether to celebrate or condemn it, loses the differences we need to judge what is happening.

As someone who builds software, I want people to be paid for work that others value. I also want customers to have lasting control over what they paid for. Those commitments become difficult when either side treats its own interests as the only ones that count. A company cannot expect goodwill indefinitely while making the customer’s position more conditional. Resentment toward that company cannot tell us which other people deserve to lose their livelihood.

Where control goes next

One possible future is more software that people can run and maintain independently. Compatible implementations could keep useful tools alive after their original vendors move on. Open code would let another team continue the work. That would give users an answer to abandonment beyond hoping a company changes its mind.

I can also imagine vendors responding by moving more of the valuable functionality onto servers they control. Recreating a desktop interface does not give you the service behind it. If that response becomes common, easier reconstruction of local software could encourage products to become less local. Users might gain more ways to copy the visible application while losing access to more of what makes it useful.

There is another dependency to watch inside the rebuilding process itself. If a small team can only maintain its alternative through an expensive hosted model, that team remains exposed to someone else’s prices and access rules. Using an AI service to develop software does not automatically make the finished application dependent on it, but the distinction matters. We should ask whether these tools distribute lasting capability or simply move dependence to a different provider.

That is why I think the “karma” explanation ends too early. Sony losing control would not, by itself, tell us who gained it. An open repository would not, by itself, tell us whether anyone can afford to maintain it. The outcome depends on where the software runs, who can repair it, and what happens when a company withdraws its support.

I want to pay people to make good software, and I want what I pay for to remain useful. A durable relationship between developers and users has to accommodate both. If the industry keeps weakening that bargain, every new tool for rebuilding software will arrive to an audience that already wants the old arrangement to fail.

Written on 2026-10-12. Project capabilities are based on their public documentation, not independent testing. The possible futures described here are my interpretation.