I would actually like to see a transnational effort to build the next two orders of magnitude in model sizes, something along the lines of the human genome effo

I would actually like to see a transnational effort to build the next two orders of magnitude in model sizes, something along the lines of the human genome efforts and space programs. The efforts at those scales are nation state level efforts, and if we continue to see the linear improvement in model performance, I think we might have something transformative. But even if we discover a plateau, we at least will have ensure large mega models are in public hands not a few megacorps.