According to their website, IPsoft seems
to have improved the technology. Remember that Eliza was written in the
60’s when computers were tiny and memory was expensive. There was
also less software infrastructure.
The article mentions that ten IPsoft robots
were sold for $100,000 ($10K each) whereas the original Eliza was (and still
is) free.
So I don’t think these Elizae are
quite as ignorant as the first one. But its likely they are still disappointingly
dumb. But then how much real help do you usually get from humans in call
centers?
For more from IPsoft:
http://www.ipsoft.com/
-Rich
Sincerely,
Rich Cooper
EnglishLogicKernel.com
Rich AT EnglishLogicKernel DOT com
9 4 9 \ 5 2 5 - 5 7 1 2
From:
ontolog-forum-bounces@xxxxxxxxxxxxxxxx
[mailto:ontolog-forum-bounces@xxxxxxxxxxxxxxxx] On Behalf Of Obrst, Leo J.
Sent: Wednesday, February 27, 2013
2:40 PM
To: [ontolog-forum]
Subject: Re: [ontolog-forum]
Improved Elizae being used in Call Centers
_____________________________________________
From: Obrst, Leo J.
Sent: Wednesday, February 27, 2013
5:37 PM
To: '[ontolog-forum] '
Subject: RE: [ontolog-forum]
Improved Elizae being used in Call Centers
Actually,
it’s “ontologists”, to keep it more distinct from
“astrologers” and “necromancers” and the like. Anyone
who uses Eliza-based NLP templates, I say: pity the fools.
The Economist has a special report on
"outsourcing and offshoring" with an embedded article titled
"rise of the software machines". In it, they claim that
descendants of Eliza have been installed in call centers to replace Indian
English speakers with automated conversationalists. Here is the link to
the overview of that article:
Having just discussed Sapir-Whorf topics,
ontolgers might also like this material. Here is an excerpt from the
article:
IPsoft’s Eliza, a “virtual
service-desk employee” that learns on the job and can reply to e-mail,
answer phone calls and hold conversations, is being tested by several
multinationals. At one American media giant she is answering 62,000 calls a
month from the firm’s information-technology staff. She is able to solve
two out of three of the problems without human help. At IPsoft’s
media-industry customer Eliza has replaced India’s Tata Consulting
Services.
CIO magazine has a reference on the same
topic where IPsoft’s product is mentioned in more detail:
Automated natural language assistants seem
to be finally becoming commercially viable.
Rich AT EnglishLogicKernel DOT com
Those are indeed fascinating anecdotes
that make the abstract
discussions about connections in the brain
more concrete and vivid.
It's especially poignant to hear the
stories from someone we've
known and loved (and sometimes get into
fights with) over the years.
> She ... told me draw a sketch map of
the route to our home, about
> 5 miles away. It was a surprisingly
difficult task, and my final map,
> which I know was inadequate but was
powerless to improve, showed
> a diagrammatic equivalent of
Wernicke's aphasia: it had well formed
> pieces but they were not connected
into a coherent whole. It was
> this experience, in fact, which
started my interest in thinking
> of maps as a symbolic language rather
than simply a kind of pictorial
The parietal lobes, which are immediately
above Wernicke's area, are
considered the region for storing and
processing patterns that are
called frames, schemata, or cognitive
maps.
Peter MacNeilage about the information
flow during the production
of speech. It shows how the
*inferior parietal cortex* which is
adjacent to Wernicke's area is central to
that flow.
Slide 37 has a diagram of connections by
the linguist Sydney Lamb,
who spent half a century collaborating
with neuroscientists.
Lamb (and many others) believe that what
we call "concepts" can be
correlated with nodes in the parietal
lobes. Note the plural '-s'
-- the corpus callosum has direct
connections from each area in
each hemisphere to its mirror image in the
other.
Those nodes have much richer and more
complex connections to both
hemispheres than those shown in slide
37. When Gabby Giffords
got a bullet through her left hemisphere,
the connections in
her right hemisphere were sufficiently
rich to compensate to
Antonio Damasio, who wrote an impressive
book on consciousness,
said the following (quoted in slide 40):
> The distinctive feature of brains
such as the one we own is their
> uncanny ability to create maps... But
when brains make maps, they
> are also creating images, the main
currency of our minds. Ultimately
> consciousness allows us to experience
maps as images, to
> manipulate those images, and to apply
reasoning to them.
Note that Damasio calls maps the
"main currency of our minds."
> I believe that, for me at least,
listening to music uses many
> of the same brain areas.
There's a huge literature on the
psychology and neuroscience involved
in music. The following book
summarizes the developments up to 2006:
Mithen, Steven (2006) The Singing
Neanderthals: The Origin of Music,
Language, Mind, and Body, Harvard University
Press, Cambridge, MA.
Mithen claimed that the Neanderthals had
music, but not language.
That idea was dubious in 2006, and
evidence since then suggests that
they had some kind of language. But
he has intriguing examples:
> In 1953, at the age of 51, [the
Russian composer] Shebalin suffered
> a mild stroke in his left temporal lobe,
which impaired his right
> hand, the right side of his face, and
disturbed his speech. Shebalin
> recovered from these symptoms within
a few weeks... Then on 9 Oct 1959,
> Shebalin suffered a second and more
severe stroke... After experiencing
> two epileptic fits, he died from a
third stroke on 29 May 1963.
During that period from 1959 to 1963,
Shebalin exhibited symptoms of
Wernicke's aphasia, which made it very
difficult for him to understand
and speak his native Russian.
Following is a sample Shebalin's
speech, as quoted by the neurologist
Alexander Luria:
S quoted by AL quoted by M:
> The words... do I really hear them?
But I am sure... not so clear...
> I can't grasp them... Sometimes -
yes... But I can't grasp the
> meaning. I don't know what it is.
But amazingly, Shebalin continued to
compose music and to teach music
at the Moscow conservatory. He published 11
major works between 1959
and 1963. A few months before his
death he completed his fifth
symphony. Dmitri Shostakovich described it
as "a brilliant creative
work, filled with the highest emotions,
optimistic and full of life."
> I met a PhD Mathematician and another
PhD Physicist, both of whom
> claimed to NOT have a visual memory
in which to sketch their thoughts.
Bertrand Russell also said that he
couldn't imagine thought without
language. But Whitehead, who I
believe was a much better mathematician,
said that when he was writing in English
he had the feeling that he was
translating from a foreign language.
> despite western governments,
including the Obama administration,
> apparent enthusiasm for the subject,
"mapping the brain" tells us
> nothing much of consequence
The research summarized above was based on
mapping dead brains by
dissecting and mapping live brains by
scanning. Those results are
essential for both diagnosis and
treatment. Without them, surgeons
would still be doing lobotomies.
See the web site for the Human Connectome
Project:
One of their goals is to diagnose
Alzheimer's disease prior to doing
an autopsy. If they can detect it,
they can compare the effectiveness
of different drugs and therapy.
_________________________________________________________________