Given the likelihood that AI will increasingly permeate the software and systems we rely on, the claim that these AI models should be open source is fair but unrealistic. Vaughan-Nichols puts the blame on “leading AI vendors who are unwilling to open source their programs and datasets” and suggests that “companies hope to tout their programs with the positive connotations of open source, such as transparency, collaboration, and innovation.” Maybe? Or maybe they don’t have the luxury of exposing all their code because that would turn out to be really bad business. I know some like to nonchalantly point to Red Hat as a classic example of business success, but compared to Meta, AWS, etc., it’s a terrible example. As Sasha Luccioni of Hugging Face said at the United Nations OSPOs for Good conference, “You can’t really expect all companies to be 100% open source as defined by the open source license. You can’t expect companies to just give up everything they do to make money and do it in a way that’s comfortable for them.”

Maybe we would like the reality to be different. But after decades of open source and proprietary software coexisting peacefully, why should we expect AI to be any different?

Just like cloud and on-premise software before it, most AI software will not be open source. Today, as then, most developers simply won’t care, because most developers care more about going to their kids’ soccer games after work than about existential open source issues. For years, we’ve focused on the wrong things in open source conversations, and younger developers have largely turned away from it. But young or old, developers care about getting things done. They care about the cost, speed, and performance gains of Mistral’s latest model, not so much about its non-open source license. The same goes for OpenAI, Meta’s Llama, etc.