Re: [NTLK] Pen computing is dead... long live the iNewt

From: Lord Groundhog <>
Date: Sun Aug 16 2009 - 16:29:18 EDT

~~~ On 2009/08/16 19:14, RAParker at wrote ~~~

> You still haven't convinced me that it is not what it claims to be.

I'm really interested in how you maintain this position, so please indulge
me a little.

Do you include the information at the link I posted earlier, or does it mean
that you know something that counters the data posted there? Here in the UK
(and on the web) it was addressed by the BBC with coverage that included the
admission from Spinvox that they resorted to "conversion experts" to make up
for the shortcomings of "D2", although OTOH they try to deny this.

For example:
 = <>

Note the quotation from a company source: "A source at the company has told
the BBC that the vast majority of messages are in fact converted into text
by staff at call centres." See also comments by people like Kareem Lucilius
who worked at the call centres.

Then I note this follow-up report from the BBC:
 = <>
 including this:
"Late yesterday, someone pointed me to an advert on a site called, where Spinvox appears to be seeking tenders for new
call centre operations. It says that Spinvox "is currently in need of some
significant support with our voice-to-text transcription services."

It outlines the nature of the work and then concludes:

'We would initially require you to provide us with c.50 agent workstations
24/7 for a 3 month trial, which if successful would lead to a 2/3 year long
term commercial deal with significant ramp-up of agent resource numbers.'"

And there's this, including both Spinvox's response to the original
allegations (Spinvox's "spin"?) and the BBC journalist's further evidence:
 = <>

I won't bother quoting from this; it's too long and there's too much. But
there's enough to make me doubt seriously -- very seriously -- that Spinvox
really does work without fairly extensive human intervention, more extensive
than will make it a practical solution *at this time*.

And for sake of completeness:

From my earlier post:
~~~ On 2009/08/16 15:29, Lord Groundhog at wrote ~~~

> Errr, hang on a minute. Spinvox? Are we talking about this:
> <>
> i.e., <> if that link is broken in transit.

Please understand: I'm not writing this because I hate Spinvox as such,
although I dislike it if they are misrepresenting what they do and how they
do it. I think this kind of voice-to-text operation may be posssible one
day and I might be happy to use it when it is. But on the evidence so far I
don't think Spinvox is "it". And even though, as I've posted previously, I
don't believe voice recognition is a complete replacement for HWR or even
for typing as a means of entering data, my objection is not to VR itself and
is a different issue. My only concern is that any argument based on the
'amazing reliability' or whatever of Spinvox apart from human call centres
seems at this time to be an argument built on sand.

In other words, I'm only addressing the original premise that Spinvox "...
very well may be the Holy Grail of text entry on the iPhone. ... No
stylus... No keyboard... Just speak to type", and that it really is
legitimate machine recognition and not dependent upon a human call centre to
make it work.

Thanks in advance.


~~~ ~~~ ~~~

łAny sufficiently advanced magic is indistinguishable from a Newton.˛
            -- what Arthur C. Clarke meant
(With thanks to Chod Lang)

~~~ ~~~ ~~~
Fight Spam. Join EuroCAUCE:
Get MUGged and love it:
Join today:

The NewtonTalk Mailing List -
The Official Newton FAQ -
The Newton Glossary -
WikiWikiNewt -
Received on Sun Aug 16 16:29:32 2009

This archive was generated by hypermail 2.1.8 : Sun Aug 16 2009 - 17:30:00 EDT