UNDER funded??? The AI space is horribly OVER funded, with a lot of people trying to get in on the ground floor of the next "big thing". Most jump on the LLM/GanAI bandwagon.
The thing that needs to be focused on are AI use cases that bring value. A framework for developing AI (think of an SDLC for AI models) is something that the AI space desperately needs. Wastsonx.gov helps teams address this issue - but no industry standards have really evolved yet. There has been an evolution of some standard measures and metrics around speech models, classification models, and predictive models. I have yet to see a standard for Gen AI (other than having a second Gen AI model "rank" the first one - which is not really a valid measure). Language translation is fairly good, but still cannot consistently survive the "round trip" trip test.
So I would love to see some investment in the process and methods for AI development. Software in general didn't really take off until we got some basic development methodologies and testing principles agreed upon in the profession. I would LOVE to see something similar done with AI - I think it would help us find the useful and cost effective use cases much more quickly, and have us wasting few resources and less time on things that don't matter.