Game Porting and Software Creation Accelerate with AI and OpenAI Concerns

Robert Jacobi: Hello and welcome to Open Channels News. I’m your host, Robert Jacobi, talking to you today from MSP Global in Barcelona. Today, AI is changing what it means to build or rebuild software. Video games are escaping the platforms they were designed for. A developer is trying to recreate Adobe’s biggest applications as open source alternatives. And OpenAI is raising a difficult question about the safety of open models. These are the 3 versions of the same story. When the cost of making software falls dramatically, who controls what happens next?

Let’s start with gaming, because something interesting is happening under the hood. An Ivan Global report, also circulated through MSN, looks at a wave of AI-assisted reverse engineering, emulation, and game porting. Developers are getting games to run on hardware and operating systems they were never designed for. And AI tools are helping speed up some of the painstaking analysis involved. These are different activities, by the way. Emulation requires an environment for existing software. Porting adapts software to another platform. Decompilation tries to recover a readable representation of compiled code. And jailbreaking bypasses platform restrictions. They aren’t interchangeable, even if they’re all part of the same conversation.

What connects them is a falling technical barrier. Work that required huge amounts of manual investigation can increasingly be assisted by AI. For players, that’s exciting. Old games can have longer lives. Platform exclusivity becomes less absolute. Preservation becomes more practical. But there’s another side. A developer who spent years creating a game may find that its mechanics or presentation can be reproduced much faster than before. Raises real questions about copyright, fair use, fair competition, and whether independent studios can protect their work.

And one important correction to the headline: a game becoming easier to reverse engineer does not automatically make it open source. Open source involves licensing and permission, not just access to recreated code. The tech is changing quickly. The legal and business answers are going to take longer.

Now imagine applying that same idea to the software people use for work every day. Ars Technica reports that developers Brandon Thomas and the ArtCraft team have introduced 7 early-stage open-source applications intended to recreate the familiar Adobe workflows, including alternatives to Photoshop, Illustrator, Premiere, Lightroom, After Effects, InDesign, and Acrobat Pro. Thomas says he used Anthropic’s Claude Opus to help create clean room replacements in Rust with browser-based versions using WebAssembly. And he made a rather ambitious declaration: software is over. Well, sure. Let’s pretend. But let’s not uninstall Photoshop just yet.

Ars points out that these are very clearly alpha releases. They have missing features, bugs, and plenty of ground to cover before they can realistically replace mature professional tools. And that’s the part that often gets lost in AI demos. Generating an interface is one thing. Reproducing years of edge cases, file compatibility, reliability, performance, and professional workflows is something else entirely.

There’s also a business model. Thomas says the free apps can be supported by paid access to ArtCraft’s AI tools. So the software may be free, but the company still needs revenue. It’s not a contradiction. That’s just how a sustainable project might work. The bigger point is that AI could make credible alternatives cheaper to create. It doesn’t automatically make them cheaper to maintain. And for incumbents, the moat may no longer be just the code. It must be trust, integration, support, and the workflows customers already depend on.

And that leads to our final story, which asks what happens when powerful technology becomes broadly available? In a Pioneers of AI conversation carried by Yahoo, host Rana el Kaliouby speaks with OpenAI Head of Product Thibault Sottiaux about the safety challenges of open-source AI. The concern raised in the interview is that openly available AI models may not come with the same centrally managed safeguards and deployment controls that companies can apply to their hosted systems. It’s an important distinction. When you operate a model as a service, you can monitor usage, update protections, and sometimes restrict harmful behavior. When someone downloads model weights and runs them independently, the original developer has much less control over deployment.

That doesn’t mean closed systems are automatically safe, and it doesn’t mean every open model is dangerous. It means the responsibility is distributed differently. We’ve seen this argument before in cybersecurity: more access can mean more independent research, more transparency, and faster innovation. It can also mean people, including malicious actors, have access to powerful tools. The question shouldn’t be reduced to open equals good or closed equals safe. Neither is that simple. Instead, we need to ask what safeguards exist, who can audit them, who’s accountable when things go wrong, and whether the benefits of openness outweigh the risks in a particular use case.

Let’s connect the dots. AI is helping people rebuild games that were locked to particular platforms. It’s helping developers attempt alternatives to some of the biggest names in commercial software. And it’s making the question of how much control developers should retain over AI itself even more urgent. The ability to recreate something is not the same as the right to distribute it. Open code is not the same as maintained code, and access to a powerful tool is not the same as a safe deployment. AI is lowering the barriers to building technology. That’s exciting. But the hard parts, ownership, responsibility, security, and sustainability, don’t just disappear because the code arrives faster.

And I’m Robert Jacobi, your host of Open Channels News. Thanks for watching. Be open and stay secure.