> there are exciting applications which will end up underfunded after the current AI bubble bursts

Could you provide examples? I am genuinely interested.