That Paris clip above -if this is real-time audio processing that HAs can do, not playing with the microphones- is extremely impressive. Now where I live, in the quiet suburbs, there is not much traffic. Would be nice to have other examples with other types of noise, like someone sitting in a busy bar, background music. Sound clips are far more informative than those sleek marketing videos. But why do we see (hear) this on LinkedIn and Fakebook rather than on the company website?
This conversation suggests the noise reduction concerns speech vs non-speech noise (only):
https://www.reddit.com/r/HearingAids/comments/1ehoqr3/new_phonak_audeo_sphere/
I am still not sure if this is truly achieved by “IA on a chip”. IA sounds a bit like the “fuzzy logic” hype a couple of decades ago… While its application to produce far superior chess and go software is crystal-clear, whether all electronics from toasters onward will now improve is unclear to me.