Comment by geremiiah
5 months ago
This is no different than what happened in Germany. Despite having tons of funding, tens if not hundreds of AI research institutes, an abundance of talent and enough cluster time to go around for everyone, they were also blindsided by LLMs and are still to this day completely out of the race. Here in Germany, they tended to bet on "AI for science", because it sounds "sciency" and that somehow appears more substantive, even though a lot of these "AI for science" or ML4Science projects are utter bullshit. They also invested substantially in Quantum Computing, again because it sounds "sciency".
Ime it's not just being blindsided but also simply preferring to do research in other areas.
TabPFN comes out of the German University of Freiburg: https://github.com/PriorLabs/TabPFN
Almost every organization (private or public) was blindsided by LLM, right? That’s… just the nature of a big new idea.
> Almost every organization (private or public) was blindsided by LLM, right? That’s… just the nature of a big new idea.
... or of a hype bubble. :-)
Stable diffusion was a productised version of work done at LMU. Not sure Germany is the best example of how AI funding goes wrong.