[LINK] We'll just send your speech to our servers ...

[Fernando Cassia]

On Wed, Oct 12, 2011 at 19:53, Pilcher, Fred <Fred.Pilcher at act.gov.au> wrote:
> Now taking bets as to how long it takes Apple to patent speech recognition.
>
> Fred

For some of us who got introduced to Voice dictation with IBM´s OS/2
Warp 4.0, the sad state of affairs is that IBM, once a leader in the
field, sat on its laurels and gave away the market first to Nuance,
and then Nuance in turn engulfed IBM´s product (to sit over it and do
nothing), while in turn Microsoft introduced Voice Recognition in
Vista, rendering Nuance´s market obsolete (ie "why pay for voice
recognition if the OS already includes such an engine").

IBM, in turn, had a Linux version of its Viavoice engine, and it did
an IBM classic: release a half-baked version, watch as sales were not
what what they hoped, underfund the project, see it stagnate, and then
kill it.

So Big Blue lost a big opportunity to give desktop Linux an edge (and
currently that Linux code could have moved to moble platforms), by
giving away its engine, much like Sun did with StarOffice when it
created OpenOffice.org and leveled the playing field of Office Suites
(remember what was the landscape for productivity apps on Linux before
OO).

If OS/2 Warp managed to do Voice Dictation on a 100 Mhz Pentium I, I
guess mobile devices with 1Ghz CPUs could do it locally, too. But of
course, that can´t be sold as "software as a service" and artificial
barriers of entry cannot be created and thus there´s no money making
scheme.

That´s why I call cloud computing and SAAS "the fine art of separating
people from their software".

FC