As we move into the second half of the 2020s and what still feels like AI mania, we seem to be on the inevitable path of relitigating the voice computing debate from the late 2010s. The thinking is that the proliferation of LLMs has in some way addressed friction points that previously hamstringed digital voice assistants. This not only is a misreading of what occurred in the 2010s, but also lacks visionary…