Not because OpenTeams deserves credit. But because the letter describes exactly what we’ve been building toward since day one. Only now it’s becoming impossible to ignore.
The letter’s core case is simple: organizations need to own their AI, not rent it from vendors who control their data, context, and results.
That future depends on open models and open ecosystems. Not as ideology, but as survival.
That idea isn’t new to us. It’s older than OpenTeams.
When I created NumPy and SciPy twenty years ago, the vision was the same: give the world tools they could own and control, not software that held them hostage. When I founded Anaconda, it was to build a stack where companies could own their entire data infrastructure.
When the open-source movement proved that transparency builds better, more secure software than closed control, that wasn’t a surprise to us either.
But somewhere along the way, SaaS-inspired AI started to break that pattern.
Companies invested billions in AI and found themselves locked into proprietary platforms, paying license fees, losing control of their data and the intelligence they built. Worse: they started losing the context, the accumulated knowledge, decisions, and business expertise embedded in that AI over months and years.
Once you stop paying rent, you risk that intelligence evaporating. You can’t move it or own it, you start over.
This isn’t sustainable, nor inevitable.
The NVIDIA letter makes the case beautifully: open models expand access, strengthen competition, give organizations control over their own infrastructure, and, counterintuitively for some, improve safety and security, but obvious to those steeped in a lifetime of open-source development. A fragmented ecosystem with transparency beats a concentrated one built on closed systems and single points of failure.
But here’s what matters more than any letter: building the proof.
OpenTeams helps organizations own their intelligence instead of renting it. We work with enterprises to build systems where they control their data, own the intelligence they generate, and can deploy that intelligence however they need it, on their own infrastructure, across their own networks. We do this grounded in open standards and open source. Not only because we’re idealists. Because it’s the architecture that actually works at enterprise scale.
When we help a healthcare provider own the intelligence embedded in decades of clinical decisions, not paying a vendor each time they want to use it, that’s the future the NVIDIA letter describes.
When we help a financial institution build on open models and own the specialized knowledge they’ve accumulated, and never worry about losing it to a licensing change or vendor lockout, that’s the future.
This isn’t competition between open and closed AI models. Both will exist. The question is simpler and rooted in the infrastructure that turns models into intelligence-infused standard operating procedures: Do you own your intelligence, or do you rent it? Can you control your data? Can you adapt your models to your business? Can you deploy on your terms?
For years, those questions seemed radical. Niche, even. The domain of startups and the obsessively cautious.
Today, they’re mainstream. NVIDIA, Meta, Microsoft, and the ecosystem they represent just said so publicly, because the stakes are too high to ignore.
OpenTeams is proud to sign this letter. We’ve been building this future and living it since we started. Now we’re not alone in seeing it clearly.
That’s the future worth building and we should all insist on owning our intelligence, because the alternative, at this point, is untenable.
