just as a point of curiosity… How does Hawking supply input to his voice synthesizer? Is it some variation of typing, and does it accommodate some shorthands for common words? Also, whatever means of input, how accurate is it? And having never met him in real life, does the device enable him to have conversations in the fast, breezy sense most of us take for granted? In short, how many wpm does this gadget do?
I’ve heard, also, that he has recently switched synths. Any change to the input protocol? And who if anyone else relies on such a system to talk?
These aren’t very detailed, but this says he can move two fingers on his right hand to input text on his new system, but his old one was operated only by a cheek pad (that’s a face cheek, not the other type).
The old one did have “shorthands” where he could type a couple letters and select the word to complete it, but he couldn’t really hold a conversation as it took several minutes to tap out a sentence. Hopefully his new one is a bit better.
His ability is the limiting factor, these systems can probably handle any text you can throw at them in real time. I have text to speech set up on my computer just for the cool factor and can send a file to it and it starts reading instantly.
I bet he macros some preset sentences, medical needs, common stuff like food and drink and comfort stuff, and I bet he also has about 30 questions that people always ask him in interviews that he macros the answers to as well =)
On a fairly recent TV show about Hawking (forget channel, but Science or Nat GEo maybe?) the guy who made the speech thing was saying the speed at which Prof Hawking can choose words/phrases is “amazingly fast”. I was also stunned seeing him in action. On a few seconds of video that caught the screen as he was ‘typing’ something out, the little bar/cursor that flashes through the matrixes of letter/word/phrase choices was just flying around the screen and the screens were changing confusingly fast (imho). The man can actually haul ass with that device of his. Far faster than what you’d assume from his appearance, to be honest. I am not sure if the device allows actual conversation, though.
On a side note, the maker of the device was initially asked to make the thing for an ‘someone wishing to remain anonymous’ that had certain specs to build to. He figured out who it was for early on, gave the device ‘extra-special’ attention to being as fast as possible input-wise and what ended up was something that increased Prof Hawkings output tremendously. Didn’t charge a penny for anything either, iirc.
I hope I remembered the above fairly factually, but nothing is far off, no doubt.
Arun Mehta contributed chapter 30, devoted to Dr. Hawking’s device.
The author discusses the pressing need to upgrade Dr. Hawking’s system since the older one was based on a hardware implementation that is no longer available. The author goes on to describe in detail the process of coming up with a good user interface and the various pitfalls.
A very interesting read, and it addresses this question well.
Have they given him a more natural-sounding voice?
I recall reading an article by Roger Ebert a year or so ago, after he permanently lost the ability to speak, that someone had approached him about a speech-replication device that would use Ebert’s actual voice, since there are years and years of recordings of him speaking on his TV shows. I always thought that would be pretty cool. Not feasible for Hawking, I’m sure, since that library of his speaking voice surely doesn’t exist.
I saw a video a few months ago of him trying out a rough version of this for the first time . . . in the presence of his wife. They both thought it was pretty cool.
Don’t have a cite for it, but I do recall reading years ago that Hawking’s chief complaint about his voice synthesizer was that it gave him an American accent, and he really wished they’d make one with a British accent.
I seem to recall reading at one point that he didn’t want the new vocal pattern to be inserted into his machinery, as he became accustomed to the sound of “his voice,” as it were.
Ah, something I know about ! Hawkin’s uses EZKeys. It’s made by WordsPlus a California company. All his responses are either macroed or pre-typed. There is no way he can “speak” at more than 10 words a minute using this system unless you pre-type answers.
I used this system for about 3 years and it’s EXTREMELY slow.
I rather doubt this, given that, at some point, the person (or persons) who made it would have to sit down with the customer to teach him how to use it.
(aside) I was at a party with Hawking once, but enough other folks were paying enough attention, and asking to pose for pictures with him and such, that I wasn’t inclined to butt in myself and ask him anything. And he had a flock of handlers, too.
Oh, cool, I was hoping you were lurking around and would hop into this thread! Being, y’know, an expert an’ all. And you showed right up! Right on time, too.
I remember pretty clearly the guy saying the initial contact(s?) with the ‘purchaser’ (or whatever he called them - “his people”?) called for keeping it a secret (maybe total non-disclosure, whatever) and he told them if it was for Hawking he’d do it all at no cost (in so many words). The guy was kinda chuckling at the idea of keeping the new owner secret from the maker/designer, fwiw. I don’t know the exact nature, but the ‘secretive’ aspect was one of the points of that ~10-15 second clip with the guy.
I wish I remembered the name of the ‘documentary’, and it has been on a couple times over the last 6 months (at least, iirc). The topic may have been about Hawking getting everything together for an academic presentation (?? maybe his concession on black hole information destruction) including stories of how he came to that important day with quick background(s) of his disease progression (and tech adaptation for him).
While the main content of Hawking’s public lectures are always pre-loaded, he does often take questions at the end. The questioner asks something, then the host fills for several minutes while Hawking types up a very short answer (something between one word and two sentences), and then the answer is voiced. Since the host has no way of conveniently knowing when Hawking is ready with his answer, the answer usually interrupts whatever is being said at the time.
Now I use something called SofType which I can control with head movement which I didn’t have for the first couple of years.
On a side note, Hawkin’s system goes for about $7000. It is covered by Medicaid, Medicare and private insurance. There are several systems as good or better that are available for next to nothing online.
I wonder why Hawking wouldn’t use the state of the art system. It seems nuts that he would be using something that is considerably slower than other options available. Or is what he uses the fastest available for his particular physical limitations?
I have spent what seems like several hours looking for that particular clip I mentioned earlier (no luck, of course, sorry). I will say, though, that one show called Stephen Hawking - Master of The Universe does show him ‘in action’ with his device. A fair amount of over-his-shoulder views while he scrolls/flashes around on the screen making ‘speech’ is shown at various points. It also shows pretty clearly how the switch rests against his cheek and his ‘blinking’ causes the switch to do its thing, etc. One of his students gives a demo of how the words/phrases are chosen from various menus - helps answer OP pretty clearly, imho.
Bootis, Hawkin’s uses that type of system because of his physical limitations, but he uses that particular system because he is a paid endorser of that brand
