OpenAI x Palantir - What we know.

What links tell about this unusual relationship

When Palantir released the Q2 financial results for 2022, one thing stood out from the rest of the call, and that is the public announcement that OpenAI and Palantir were experimenting inside Foundry. Even if at that time ChatGPT wasn’t the big phenomenon that it is today, the fact that Foundry+GPT was a thing well before anyone heard of it means something.

But how does it work? How can OpenAI benefit from Palantir, or otherwise?

My theory is; Clients could use GPT capabilities inside of Foundry to get insights and even code without knowing how to. Not even low-code, but directly zero knowledge, but same results, faster and with a smaller window for human errors. In exchange, OpenAI can use the software to learn and help make their chat-bot better by organizing and streamlining their data more efficiently, so that the next versions of GPT become more powerful and adaptive to their users.

Also, I strongly believe both companies benefit from working together in the form of having jointly developed technologies they can each transform into their own marketable products. Palantir, for example, inside the Ontology. And that isn’t only me saying this, but links prove it.

I’ve seen a new link, as DataBase users know by now, and it establishes a direct line between GPT and Ontology, proving that they use it. The name is “OntoloGPT”, and they share connections with other codenames that link to OpenAI. For those who may not know, Ontology is a powerful tool that enables users to define and manage complex data relationships. Palantir defines it like this on their website:

The Ontology sits on top of the digital assets integrated into Foundry (datasets and models) and connects them to their real-world counterparts, ranging from physical assets like plants, equipment, and products to concepts like customer orders or financial transactions. In many settings, the Ontology serves as a digital twin of the organization, containing both the semantic elements (objects, properties, links) and kinetic elements (actions, functions, dynamic security) needed to enable use cases of all types.

So what about a possible rivalry between the 2 companies? Didn’t Palantir just announce AIP as a similar tool but for companies and their digital twins to better know and predict what they can and can’t do? I think these are 2 different uses for 2 different softwares, and as I’ve just said above, I’m also convinced that both companies share technologies that they created while experimenting almost 1 year ago.

We will soon find out what AIP will look like, and I’m very excited to see what they have adopted from their work with OpenAI and how they have decided to differentiate themselves. The fact is that Palantir has the tools, expertise, culture and technology to develop a business oriented AI platform that will be the final touch to convince a company that may be considering to use Foundry for their own business. And I’m in the process of detecting how Palantir plans to sell and distribute this new platform, to dig and find what I can.