The open source models are showcasing a lot of cool research ideas that will undoubtedly end up in future frontier models from OpenAI and others. We're at a point where the big proprietary AI firms barely do any foundational research anymore, they're purely about incremental improvement. All the proven substantial progress recently has been coming from the open models side.
I agree that DeepSeek and others have been doing most of the published research for future frontier models.
But I’m not sure we can say they the big US companies aren’t doing foundational research anymore. I suspect a lot of it is just kept unpublished as a result of the intense competitive pressure, which is disappointing.