It's true that Tsetlin Machines are currently a fringe area of ML research, especially compared to the focus on deep learning advancements coming out of SF and China. It's early days, but the energy efficiency potential is insane. I believe further investment could yield significant results. Having been supervised by the creator, I'm admittedly biased, but the underlying foundation in Tsetlin's learning automata gives it a solid theoretical grounding. Dedicated funding is definitely needed to explore its full potential.