Well,
one part of it could actually work?
Type text into Poser, and let Poser do the talking? ? 
Make talking animations.

At first only the mouth movements. But...why stop there?

First speech would be letter by letter, but afterwards, why not learn poser "words", like an automatic spell checker works..., Put the words on a timeframe, together with the animation?
There are automatic reading progs for the blind already. And some are past the letter by letter stage.

First year; automatic lip-mouth movement with typed text, over a controllable time frame in an animation.
The Pro version could actually produce sound. (let us call it sound , OK?) Ok, noice is fine too..
And, OK, why not a market for sound packs, males, females etc. . . 

Then word recognition?

There is a new market here.....

Next year the normal Poser would get the same.

And the pro version would go a step further, and combine speach with complete facial expressions, where you could indeed use angry, and smile dials to influence the expression, AND the tone. . . .

Never stop dreaming.....

yeah-yeah, by poser 2020.

Poser 1, 2, 3, 4, 5, 7, P8 and PPro2010, P9 and PP2012, P10 and PP2014 Game Dev
"Do not drive faster then your angel can fly"!