I think the 1-2 years estimate is a bit short. Right now, a computer can create a droning audiobook out of a written book. Within the next few years, generati

I think the 1-2 years estimate is a bit short. Right now, a computer can create a droning audiobook out of a written book. Within the next few years, generative models will create cheap mediocre audiobooks. But the technology is changing very rapidly. I would expect that by 2030 generative models will have the ability to produce some very good audio. Possibly on par with professional human voice actors.