Received: by 10.66.88.231 with SMTP id bj7mr4407251pab.45.1349733192424;
        Mon, 08 Oct 2012 14:53:12 -0700 (PDT)
MIME-Version: 1.0
Received: by 10.68.135.103 with SMTP id pr7mr4841769pbb.7.1349733192407; Mon,
 08 Oct 2012 14:53:12 -0700 (PDT)
Path: g9ni23679pbh.1!nntp.google.com!kr7no10220772pbb.0!postnews.google.com!pz10g2000pbb.googlegroups.com!not-for-mail
Newsgroups: sci.lang,comp.ai.philosophy,comp.ai.nat-lang
Date: Mon, 8 Oct 2012 14:53:12 -0700 (PDT)
Complaints-To: groups-abuse@google.com
Injection-Info: pz10g2000pbb.googlegroups.com; posting-host=101.172.213.65; posting-account=vGh9tAoAAAANs3L3QhlQnZhTxK72bUPf
NNTP-Posting-Host: 101.172.213.65
References: <X6SdnYKmS4V5ld3NnZ2dnUVZ_gednZ2d@giganews.com>
 <20121005123056.481$lT@newsreader.com> <6e5d4126-d3b4-4798-892a-68c3c6c6cccf@v9g2000pbi.googlegroups.com>
 <20121005173111.195$v3@newsreader.com> <2LmdnSbIVr1Ptu3NnZ2dnUVZ5hCdnZ2d@giganews.com>
 <766d30c1-f75b-4fe4-a1a3-09e097279340@kg10g2000pbc.googlegroups.com>
 <5pSdnQy-qsYZo-3NnZ2dnUVZ5owAAAAA@giganews.com> <097d2562-da33-4742-bb45-b83099f9d214@wz4g2000pbc.googlegroups.com>
 <JeOdnQA-68AmFO3NnZ2dnUVZ_qcAAAAA@giganews.com> <10fd5135-1bc6-4ab4-a578-a2ed5a27d8bc@j2g2000pbg.googlegroups.com>
 <fqudnUbvBsjIPu3NnZ2dnUVZ_vGdnZ2d@giganews.com> <6c691c24-09f4-47b5-ae6e-af2326f3ad0c@ro10g2000pbc.googlegroups.com>
 <H_GdnUZGyNu4Ju3NnZ2dnUVZ_hqdnZ2d@giganews.com> <2fd87f36-3078-432b-9f9e-dda431cbdbfc@q7g2000pbj.googlegroups.com>
 <m4OdnVRZBpn5bO_NnZ2dnUVZ_rednZ2d@giganews.com>
User-Agent: G2/1.0
X-HTTP-UserAgent: Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 6.1; WOW64;
 Trident/4.0; SLCC2; .NET CLR 2.0.50727; .NET CLR 3.5.30729; .NET CLR
 3.0.30729; Media Center PC 6.0; MAAU),gzip(gfe)
Message-ID: <cf1ae072-cd98-478d-9b91-2cb2546ff6a4@pz10g2000pbb.googlegroups.com>
Subject: Re: What is Needed to Teach a Computer to Read?
From: casey <jgkjca...@yahoo.com.au>
Injection-Date: Mon, 08 Oct 2012 21:53:12 +0000
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

On Oct 8, 8:29=A0am, Peter Olcott <OCR4Screen> wrote:
> On 10/7/2012 6:36 PM, casey wrote:
> >> Curt of course believes all GOFAI programs will be
>> replaced by a single uniform blank slate reinforcement
>> learning network so your efforts will be a waste of
>> time as you will be able to train this network to to
>> do anything we do (and more) without having to, or
>> being able to, understand in detail how it does it.
>
>
> If it is based on an artificial neural network, he
> is incorrect. If the network that is referred to is
> the same as the ontology (and set of taxonomies)
> that I am referring to, then an outside observer
> will be able to see how this process works because
> the knowledge base will have the same structure as
> the human knowledge base.

What ontology and set of taxonomies you are referring
to is unclear for me I don't know how clear it is to
those on sci.lang and comp.ai.nat-lang

However it seems to me that the brain doesn't work
like CYC or any other GOFAI program so far written.
A program that can learn the way we do will develop
the same way our brains develop. It will not start
and end as the Watson program it will develop over
time just as our brains developed over time.

When we learn by reading we don't store some formalized
version of sentences and we don't remember the data
that way. We get the gist of the text and integrate
new knowledge from that text into an existing knowledge
base the structure of which I suspect is unknown.

We try to emulate the behavior of the brain and given
the same problem may come up with similiar solutions
or at least get ot understand the problems the brain
has solved but the behavior of programs I have read
about so far don't seem to emulate the learning from
reading ability of real brains.

jc