Open Source ARC Platform

I browse this forum to watch and learn, but this post caught my eye. I am curious why some people demand access to the source code of products. Is the motivation curiosity to see how they built it? Because asking for someone's "secret recipe" gives you access to copy it and make it yourself, right? What is actually limiting you from building a robot by not having access to ARC source code?

Or are you asking for Synthiam to give you their source code so you don’t have to pay for their hard work?

I’m just asking. Don’t take offense to the questions. I just don’t understand how you having source code helps you make a robot?

Edit: I did notice Synthiam has a GitHub and several robot skills and products have published source code. Could you not just make a robot skill? Isn’t that the same idea as making a program for Windows? You don’t have windoes source code but thats not stopping anyone from making programs for it.

Hi RoFLow234, I really like your question. There are certainly many technical solutions that don’t require the technology itself to be open source. However, when the goal is to create a common platform that can evolve rapidly and be shaped by a broad developer community, I believe open source is often the best path forward. I see open-source software as an unparalleled catalyst for the growth of a robotics platform. Wikipedia is a great example. It started as Nupedia, a closed, expert-driven project where content creation was heavily controlled. That model created a significant bottleneck. When Wikipedia introduced an open, collaborative approach, the platform went from thousands of articles to millions. The community drove its growth, contributed knowledge, and collectively improved and validated the content. It demonstrated that large-scale evolution is difficult to achieve when development is confined to a small group. I believe ARC is facing a similar bottleneck today. Yes, developers can create extensions, but there are still significant limitations on how deeply they can interact with and improve the core platform. We are already seeing constraints around how many extensions can run simultaneously, and I suspect there is considerable opportunity to optimize the underlying software if a much broader developer community had access to the core code. The current model feels a little like telling people they can build or buy any accessory they want for a phone, but they aren't allowed to actually own or modify the phone itself. You can create an impressive ecosystem around the platform, but eventually you run into limitations imposed by the underlying architecture. If the software were free and open source, developers would have a reason to become deeply invested in the platform's long-term success. History has repeatedly shown the power of this model. The internet itself was built on open standards. Linux became foundational to modern computing, and Android became one of the world's dominant mobile operating systems. Opening the underlying platform allows innovation to happen in parallel rather than forcing everything through a single development organization. I sincerely believe in the direction this software is heading. But having the right idea isn't necessarily enough to win. I've worked for a technology company that had some genuinely incredible ideas, but they weren't implemented or adopted quickly enough. Eventually, competitors caught up-and once they did, the original innovators lost the advantage they had created. I think robotics is heading toward a similar inflection point. The technology that ultimately wins may not necessarily be the one with the best initial idea. It may be the platform that gives the largest community of developers the freedom to build, experiment, improve, and evolve it. That's why I believe opening up the core software could be one of the most important decisions ARC makes for its long-term success.

I'm not all that knowledgeable about the pros and cons of open source software versus closed source. However from my limited experience I can understand why the Synthiam wants to keep ARC under their control. I won't go into my opinions here but I do have an observation and an idea of how Synthiam has opened up development to the community without losing control of what ARC is and what it can do.

A few years ago @DJ and his team created the ability for anyone to create and publish what they call "Skills". Again, I don't know a lot about making one of these skills but I've used many of them that are written by both developers and hobbyist not associated with Synthiam. From what I can see, by making or/and using one of these skills a robot builder can make ARC do almost anything.

Isn't this what open source actually is meant to do?

Dave's correct, the ability to create your own skills in ARC is still there! See this link.

Synthiam's mentality has always been to be open-concept. Open enough to be inviting but closed enough to still have a solid monetization strategy. Even a lot of the electronics hardware from EZ-Robot was made source-available on GitHub. Open enough to be able to fix or modify the circuits yourself, but closed enough to make money with partner manufacturing licenses.

You see, without a great monetization strategy, most open-source businesses fail. There's not enough money coming in from altruistic donors to keep the lights on. A recent example is the robot company K-Scale. There aren't many investors who will invest in intellectual property that can't be privatized (exclusive) in some way; that would be a huge risk for them. That's what deep-sixed K-Scale; they couldn't find anyone to invest in an open-source humanoid robot.

I've always been proud of how open Synthiam's ARC and EZ-Robot are. Even all the STLs are available for EZ-Robot plastic parts. It's always been about having an open attitude without sacrificing too much on the business side.