All I can tell you all is it was a mystery to be and a big one at that.
Professor Stephen Hawking could use complex language shortly before he died. And he was almost completely paralyzed.
Oh, I know he did it with the aid of technology (ironically). But tech hasn’t advanced that much, it seems to me. The device would practically have to read his mind. And I know nothing can do that now.
He would literally sit there with a labored half smile on his face, while his voice synthesizer said all these complex scientific things. How did he do it?
In addition to Telemark’s link, the biggest method was time. If it was a lecture or presentation, pretty much all of it was written ahead of time. Unplanned stuff like answering audience questions could often result in long stretches of silence while he composed a response. The SwiftKey customized system helped with that.
And systems like the one developed for his use are now standard on every smartphone. Most of his words were autocomplete options, just like most of yours are. Though of course, his system was very well-customized for him.
I attended a Hawking lecture in about 2012, and while I was listening to the talk, I couldn’t help wondering whether it was essentially pre-recorded. I figured it must be, because it seemed silly to make Hawking encode a written speech into synthesized speech at each stop on his lecture tour.
But that means he was just sitting up there, listening to his voice synthesizer read his own talk for the umpteenth time, which is also a little absurd. Still, it made the most sense at the time. I guess this confirms my hypothesis.
I assume he had an assistant do any writing on the blackboard, and perhaps also help operate the computer slide projector, which slides were prepared ahead of time? Preparing chunks of text in advance to be read out by the speech synthesizer seems no different from preparing visual materials and reciting your own practiced lecture from memory, incorporating suitable adjustments, for the umpteenth time. He would still need the good text-to-speech interface to answer questions afterwards in a reasonable amount of time.
My understanding is that towards the end of his life much of his public interaction was effectively pre-programmed. Lectures were obviously set in advance, but it was more an issue that the questions he would be asked by members of the public afterwards were so predictable that he had “stock” answers that could just be trotted out. Indeed I think that his final, posthumous, book is basically a compilation of those established answers.
His personal interactions and those with his research colleagues were, no doubt, very different, and presumably very much slower.
On the other hand, communication with his colleagues would be much facilitated by the fact that they’d understand what he was saying, without the need to explain every detail that a member of the general public would have.
I’d just like to emphasize that I have no problem with anyone in Hawking’s position preparing text ahead of time in any medium. I wasn’t objecting to the situation; I just thought it was interesting.
