is DNA stupid?

Guy Macon <http://www.guymacon.com wrote:


See what I mean!
In any case, I asked each person the same question.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
e7 wrote:

The problem is that you made two statements:
1. That DNA has only 20,000 bits of information (as in 1s nad 0s).
2. Chomsky's so-called statement
I have never insulted you even though you have done this to me. I am from the US.
You refuse to explain either point, even after many people have asked. You have never stated a reference for the so-called Chomsky statement.
In other words, you are attempting a discussion based on a lack of evidence. Every logical attempt to refute you has been met with stonewalling. I think it is time to either put up or shut up. -- D. Jay Newman
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
e7 wrote:

If you'd care to take some time off from cutting and pasting, perhaps you'd be willing to supply us with the following information:
1) How -- exactly -- do you believe Shannon's theory applies here?
2) Where -- exactly -- do you get the 95% figure from Chomsky?
--
(Replies: cleanse my address of the Mark of the Beast!)

Teleoperate a roving mobile robot from the web:
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Please try to avoid leaving in "Guy Macon wrote:" when you have snipped out everything written by Guy Macon. :)
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Guy Macon wrote (really, this time):

Ooops -- my bad. No more paint thinner for me today!
--
(Replies: cleanse my address of the Mark of the Beast!)

Teleoperate a roving mobile robot from the web:
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
The Artist Formerly Known as Kap'n Salty wrote:

If you work in AI, you know that information has to be programmed in, otherwise a desirable behaviour is never exhibited. If you study Shannon as electronics engineer, you would also be told that new information cannot be created out of thin air as you get led into information and compression and all the things revolving around information. Shannon in combination with Chomsky's result is putting a limit on what a human could have learned by year 2. There just isn't enough information that can be programmed into a person in that short period. Thus from Shannon's perspective, it must be preprogrammed. Even if aspects of it isn't, the mechanisms for learning must all be pre-programmed with concepts without which learning itself will fail. Sure enough, in experiments we can reproduce this result. If a child isn't taught language by age 14 for example, that facility is dropped by the human brain. In other words, the ability to learn a language is present in the brain at time of birth, but that one specific computer program is erased at age 14. (Its not erased if you learn a language - and you can continue to learn more languages.) If that specific program exists in the brain at time of birth, then it must be programmed in before birth and that information must be coded in the DNA. Unfortunately, comparing with chimp, there doesn't seem to be enough information for it to have been put into the DNA. May be it is there, but it is too difficult to see where its been stored. So if you follow the Shannon argument, it is either in the DNA and we can't see it, or it is being downloaded and programmed into the brain by some method that is still undiscovered - basically because the information content is too high to be transferred on the DNA.

This is a reasonable figure banded about for Chomsky. Whethers its 94% or 99% thats down to personal beliefs as no experiment has yet been devised to narrow the margin of error.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

If you work in AI, you know about emergent behavior.

If you study Shannon as electronics engineer, you would know that the system you are studying has a one megabit per second source of input information feeding it.

Count the seconds in two years of time. Now multiply by a million bits per second. Oh... That's right... I forgot that you are incapable of doing simple arithmatic, as evidenced by you continued insistance that a 3 billion bit genome has only 20,000 bits. Never mind. You are too stupid to ever understand where your logic went off the track.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Guy Macon <http://www.guymacon.com wrote:

And if read any/all the works of AI, you will know emergent behaviour doesn't exist in any software written by anyone. Each behaviour is a function of the program. Emergent behaviour means behaviour outside of programming - which sadly has never been reproduced.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Poppycock. Emergent behaviour in software systems goes as far back as Conway's Game of Life.
No, don't bother writing a reply. I have seen you in action when I busted you over your claim that the human genome has 20,000 bits when it actually has 3,000,000,000 bits. No need to repeat that performance with your latest error; we already know that yiou are ineducable.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Guy Macon <http://www.guymacon.com wrote in message says...

Ok, yes, DNA has more than 20,000 bits of info. We're on the same page there. But how can you say DNA has 3e9 "bits"? A bit is generally defines as a 1 or 0 in binary, but as I mentioned it before in the other post, DNA is base-4 (4 amino acid possibilities). So using "bits" as a unit in measuring informational storage by DNA isn't accurate unless you're converting the informational storage capacity from base-4 into base-2 "bits".
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

I am merely quoting the experts: http://www.google.com/search?q=%22human+genome%22+%22billion+bits%22
In an earlier post I discussed the strong possibility that the Human Genome Project is using a different definition of "bit" and that the real number might be 6 billion instead of 3 billion, but for the purposes of debunking someone who insuists that the number is twenty thousand, either figure will do.
--
Guy Macon, Electronics Engineer & Project Manager for hire.
Remember Doc Brown from the _Back to the Future_ movies? Do you
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
e7 wrote:

I'd suggest you re-read your so-called "works of AI". Emergent behaviour does NOT mean "behavior outside of programming". A system that displays "emergence" displays seemingly novel properties not predictable (or, more accurately, not easily predictable) from its programming.
Such systems often start as a collection of relatively simple initial conditions and a finite set of rules, which are applied iteravely over time. See John Conway's Game of Life, Stephan Wolfram's Big Thick Book, or any of the zillions of software examples out there.
Really -- this is all very basic stuff. Are you sure you're an "information technologist"?
--
(Replies: cleanse my address of the Mark of the Beast!)

Teleoperate a roving mobile robot from the web:
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
The Artist Formerly Known as Kap'n Salty wrote:

That may be what is taught in USA by those willing and capable of stretching definitions to get a little funding. We don't do that in the UK. Emergent behavior means behaviour outside of programming - specifically behavior 'emerging' outside of mechanical, predictable or computationally predictable behavior.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
e7 wrote:

I'm sure the nice folks at Manchester University, Cambridge, the University of Birmingham, the University of Wales -- to name a few institutions off the top of my head hosting ongoing work in emergent software systems-- would be most interested to know that they are using a definition of 'emergence' that makes their own work impossible.
You appear to be forging ever deeper into the realm of the truly idiotic, despite all best efforts to hold you back.
I give up -- have at it.
--
(Replies: cleanse my address of the Mark of the Beast!)

Teleoperate a roving mobile robot from the web:
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
e7 wrote:

Strange. I thought the main point of AI was to have a program that didn't need to be programmed, rather it would learn from experience. Like a human learns.

Hmmm. What about imagination? Perhaps information cannot be created out of thin air, but new combinations of existing information could be made.

Has Shannon said that? If so, would you please send us the reference.

Strangely enough few people here have disagreed with this statement.

Yes. In other words, if a child doesn't *learn* language early then he can't. I fail to see how this agrees with your main assertions.

What I've read and experienced (I've studied over a dozen languages, three of them to fluency and only in English can I speak like a native is that the language aquisition mechanism goes away. Yes, you can learn new languages after 14, but it is more difficult and you can't learn to native fluency.

Yes, the brain does seem to be programmed to absorb languages during certain years. After that, this mechanism is disabled; I suppose to free up that part of the brain for something else.

How do you know how much information it takes the brain to learn language? If it is like an ANN then only a slight shift in priorities allows it to learn different things.
Perhaps this insight could be used in a robot. Essentially starting with a focus on learning about the environment and then switch the shape of the network for more normal use. I'll have to think on this.

People learn. Young people learn extremely quickly.
Yes, it is probable that our brains are set up to learn languages during a certain age, and that there is some basic simularity between languages because of this.
However, DNA doesn't set up a grammar. If it did everybody would be speaking the same language.
If the learning came from a non-physical source, then why woulc children learn the languages that are used around them?

So you claim. Nobody seems to be able to find a reference.
Please provide one. -- D. Jay Newman
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

I thought this has generally been agreed upon for ages? That the brain is known to go through distinct phases, first of rapid learning during childhood, then to curb the learning ability in order that the information loss that is bundled with learning (to rapidly forget learned information that is incorrect) is also shut down so that what has been learned in childhood can be applied in adulthood and never forgotten.
The idea that this is a programmed high-level mechanism is quite frankly absurd, however, compared to the simpler notion that this change in learning ability could be caused by chemical factors released into the brain during transition to adulthood that inhibit or change the associative processes of each individual brain cell, rather than acting specifically on higher structures. Can it really true that a specific structure, a partially-programmed language system,exists and is lost as the associative capacity of the neurons is lost, or is it in fact the case that it is merely one manifestation of a much more general, global loss in learning capacity?
I dunno if this is common accepted theory or not, I'm really just thinking aloud here, but I've been seriously thinking about conducting experiments on genetically-hardwired GLOBAL changes in brain behaviour in different stages of life - perhaps to construct two AIs, one programmed to learn behaviour and apply it constantly, the other to learn and apply, then slow down learning (not stop altogether, it should be stressed - as mentioned earlier, an adult can still learn, but it's much, much more difficult), and see how they compare in responding to identical situations, or even pitted against each other. There we go, a useful idea from this mostly useless thread, at last! Who said troll-bashing was a waste of time? Heh.
Tom
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Tom McEwan wrote:

I know. I was *trying* to get this discussion back to robotics.

It doesn't matter from a linguist point of view *how* this happens.
It is known that certain things are learned at certain ages and that some of them are linked.
I believe that this is a fairly simple process of the brain. I never said anything else.

This would be an interesting experiment also.
I was thinking of making a net of neural networks, and at some point reducing the feedback to slow down the learning process to simulate human learning. -- D. Jay Newman
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Tom McEwan wrote:

Why is it absurd? How do you propose to make a machine that can learn without having put in the program for it to learn?
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

You are deliberately confusing two entirely different levels of software. In computers, we can easily write software that learns and can write other software. In organic brains, the structures of the neurons' interconnects alter as they learn and the "software" is in the wiring and how it changes. Don't mix the "control and acquisition" program (in hardware) with the "learned behaviors" program (in "software").
Cheers!
Sir Charles W. Shults III, K. B. B. Xenotech Research 321-206-1840
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Sir Charles W. Shults III wrote:

At least you can see that someone has to write some software. Similar identical software exists in the brain at time of birth. The whole thread's debate is whether DNA stupid and doesn't have this software or whether that software is stored in the DNA and we can't see it.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.