I think this view is a bit narrow in terms of what "AI" advancements may depend on. I think it's very easy to argue that large scale AI adoption will require orders of magnitude higher bandwidth than what we currently have. It's not clear that long term electronics will win in all applications, especially with the strong resurgence in interest of photonic computing. Fundamentally, photonic platforms have much higher potential bandwidth (at the cost of power and size currently) than electronics.
GaAs (and other III-V) would likely be an essential material for some kind of photonic or hybrid compute system.
The response below addressed the quantum sensors, but I would be careful of calling "everything" quantum such as image sensors. Sure they rely on the photoelectric effect which is quantum, but not really in the sense of what we would consider a 'quantum sensor' today.
I suspect what could be more relevant are III-V based SQUID Qubits. These are highly sensitive systems that multiple nations are exploring for submarine detection. More near term, quantum communication via quantum light sources also can leverage a III-V platform.
sure, it's totally possible that the advantages of photonics or optoelectronics could win out, and iii–v semiconductors are pretty important for optoelectronics, though not for pure photonic systems like second-harmonic generation. sometimes people even use gaas for that, especially historically
what are iii–v based squid qubits? google scholar is not helpful except for finding https://journals.aps.org/prresearch/pdf/10.1103/PhysRevResea.... i thought a squid was a josephson junction device made out of superconductors and insulators, not semiconductors. gaas isn't a superconductor, is it?
this doesn't sound like a quantum communication and squid research lab though. it sounds like a 50-year-old radar chip fab that's being put on life support as a pork barrel project
brain fart on my end, you're definitely correct that SQUIDS are not something demonstrated quite yet, I should have said Josephson junction, but even that seems more niche than I had thought when I wrote the comment.
your comment oscillates between incorrect and incoherent. squids have been demonstrated for decades (i didn't assert they hadn't been) and are made of josephson junctions, whose nicheness is not at issue in this discussion. i hope you get better because you clearly were not well when you wrote this
GaAs (and other III-V) would likely be an essential material for some kind of photonic or hybrid compute system.
The response below addressed the quantum sensors, but I would be careful of calling "everything" quantum such as image sensors. Sure they rely on the photoelectric effect which is quantum, but not really in the sense of what we would consider a 'quantum sensor' today.
I suspect what could be more relevant are III-V based SQUID Qubits. These are highly sensitive systems that multiple nations are exploring for submarine detection. More near term, quantum communication via quantum light sources also can leverage a III-V platform.