What about for audio description? To do that well it's not just about cramming voice into a short space of time, but you have to be strategic and actually under

What about for audio description? To do that well it's not just about cramming voice into a short space of time, but you have to be strategic and actually understand the context. Not a trivial task. Could AI do this equally well in say 2-3 years? I seriously doubt it.