is DNA stupid?

I just told you that there are a a few billion bits of information in the human genome, and yet you repeated your "20,000" claim. Please provide evidence that you are educable.

Reply to
Guy Macon
Loading thread data ...

The more interesting case (especially to robot builders) is the insects. They have brains that are hardwired and have far fewer connections, yet they have an astonishing range of behaviors.

Reply to
Guy Macon

NO - its already highly patterned.

That is a resonable enough argument - hower it has serious problems. DNA is not variable enough to pass on human diversity.

99% is the same as a chimp. The synapses must already be weighted in advance (i.e. programmed in advance), otherwise, you start with a vegetable. According to Chomsky, 95% of the human functionality is already programmed in before it is born. By year 2, there is no way a child can do 95% of all the things it does by learning.

You go to your friends house and you have not been there before. You can ask a 2 year old to go into the kitchen and help with washing up. That means picking up all the plates, putting it in the water, cleaning it, and then drying it, and putting it on the shelf.

Computer program to do an equal task - yeah right! You would be lucky if you can locate this new kitchen defining and avoiding all the obstacles on the way. Then incredibly lucky if you could define and separate dirty plates from clean plates. Even more difficult is to clean, visually separating out patterns on the plate with dirt on the plate. Then you have to locate the sink and the tap. Do the washing up, dry the plates, locate the shelves, and put them into the shelf in a stacking order that doesn't defy gravity.

Who the hell had the time to write software for all this? There are millions of microscopic and very diverse computed rules that must be obeyed every minute of every day before you have reasonable human behaviour. According to Chomsky, 95% is already programmed in. Once we start identifying all the rules, it becomes quickly apparent that (a) nobody taught the child those rules and (b) there isn't enough time to put in 95% of the rules.

Human DNA is 99% same with chimp. There is only 20,000 odd genes active in the brain. OK each gene is complex, but its the same between humans and chimps. There is nowhere to fit human software if 99% is the same as a chimp! That 1% doesn't have enough terrabytes to store human rules of computing. But according to Chomsky, 95% had to be programmed in. So where is all the software stored if it can't be learned in the time allocated, and may not be stored in the DNA?

The entire genome I have on one CD. If I were to count up AGCT and then work out the number of bits, I would have may be a few hundred megabytes of information. I just don't see constructing a human out of that with 95% pre-programmed human behaviour.

J
Reply to
e7

A quick look at his web page shows that he has already managed to pick a bunch of fights with a bunch of folks.

Let's do a quick test to see whether he can admit that he is wr e7: "The trouble is, as an Guy: "The human genome has billions of bits of information, Guy: not 20,000. You are factually incorrect"

Could Guy possibly have remembered incorrectly? Let's check:

formatting link
Nope! The figure is 23 pairs of chromosomes, roughly a hundred thousand genes, 2.9 billion base pairs, 5.8 billion bits of information.

(Took me a few minutes to figure out that the biologists are calling a base pair a "bit of information" while we computer types know that information theory shows that CACCGTGACACGTATGCCAGTAAGCAAG has the same information as

01000101101110000100011011001110010100101100001001000010 or 45B846CE52C242.)

Now for the test. Will e7 admit that he was wrong?

Enquiring minds want to know!

Reply to
Guy Macon

I think the problem here is, our open-minded colleague e7 is treating the brain something like a Von Neumann computer, which must hence have an entire, complex intelligence-defining program loaded into it as you would load an AI program into your PC, on top of which actual sentience emerges. What he fails to realise is that the DNA probably (I say probably because my own knowledge of this entire topic is sketchy at best, please feel free to correct it or just flat-out tell me I'm wrong) only needs to carry enough data to define just a few types of brain cell, how many are needed in the brain, and their basic behaviour (and it is very basic - as far as I know, all they do is make or break connections with each other, influenced by the various chemical factors in that region of the brain, acting basically like discrete logic). All higher behaviour patterns, learning, etc are built up by the basic making and breaking of connections of those cells (one could say it's sort of like the evolution of society - countless individual, basic nodes all constantly interacting & associating to form more and more complex groupings and structures with ever more complex behaviours of those groups, on their own agenda and completely without consciousness of the higher structures they form).

I think we're fighting a losing battle here, though - it's not a good sign when someone asks a question then promptly insists the answer must be wrong.

OK, flame away.

Reply to
Tom McEwan

There is a big art of you problem right there. Chomsky is an idiot. You should be reading Marvin Minsky instead.

formatting link
BTW, I am still waiting for you to admit that you were wrong about how many bits there are in the human genome. Are you capable of admitting error?

Reply to
Guy Macon

If this bit were even mildly true, a chimpanzee which is only 1% different from a human will be 1% deficient from a human.

A chimp would also be able to do this. But unfortunately it cannot because it is not pre-loaded with human software. Where is the human software stored? It can't be in the 1% difference between human and chimp.

J
Reply to
e7

MIME-Version: 1.0 Content-Type: Text/Plain; charset=US-ASCII

formatting link

Reply to
Guy Macon

You are factually incorrect again.

You are getting closer, though. In your previous error (which you have yet to own up to) you underestimated the number of bits in the human genome by a factor of 290,000!

Reply to
Guy Macon

Ok, genes are not made of single DNA sites or locations, and it can take thousands or more DNA sequences to have a gene. Saying DNA has

20k genes involved in brain functions is not close to saying there are only 20k DNA sites that comprise those genes. It's millions, if not billions.

Also, just for clarification, a given bit has 2 possible values 0 or

  1. DNA has 4 per site, depending on which of the four amino acids is present.

Total information storable by 20k bits: 1.99 e6020 Total information storable by 20k DNA "sites": 3.96 e12040, or

1.99e6020 times greater than your original argument. And that's not taking into account that 20k genes is millions, if not billions of times greater than 20k DNA "sites".

Dave

Reply to
David Harper

You're probably assuming all parts of the genetic code have uniform significance, and any difference in data corresponds to a proportional effect. I doubt this is true, in fact I think they showed that only a single bit change can mean the difference between a hair and a feather, quite a big change. (this is a remembered fact from some documentary I heard ages ago - probably incorrect)

And besides, as an analogy, 1% difference in the code of the average computer program, depending on where is is, can cause anything from 0% to

100% failure of the program.
Reply to
Tom McEwan

A nucleotide is not a bit either. But it's closer to a bit than a byte.

A typical gene might have

3,000,000,000/40,000 = 75,000 "bigbits" nucleotides

So one gene could encode (approximately) 3^75,000 bits.

20,000 genes could encode (3^75,000)^20,000 bits.

That should suffice to make (at least) a schizophrenic.

Or a moron.

Reply to
Karen J

Looks like we have our answer. He cannot admit that he was wrong.

How do people *learn* without admitting mistakes and learning from them? I get things wrong all of the time, and I *welcome* it whan someone more knowledgable corrects me. Nobody is an expert at everything

Reply to
Guy Macon

The thing that everybody seems to have missed is that DNA does not encode the data that is a personality. It is the data (blueprint) for the computer, storage system, and mobility system that contain and move the personality about (the mind and body). Personality is 'stored' in the brain. Thinking (processor power) is also in the brain, while motor control is partly in the brain and partly in the 'short path' nervious system (reflexes).

The homosapien mobile computing system is delivered (literally) with only a basic boot strap program designed to carry on the most basic of self maintainence, sensor, and motor operations. This boot strap program is however being run on a massively parallel nural network learning computer with mutiple sensor subprocessors whose only purpose for it's first 157680 hours of operation is survival and learning.

Trying to compare this to current computing models is rediculess. If we are to ever build a human equivalent robot we are going to have to up processing power considerably AND develope an operating system that can distribute processor load over multiple parallel processors (on demend) while maintaining data integraty and interprocess data exchange. That system will also have to be self programing AND self debuging and the processor load for those function will also have to be accomodated.

It is currently believed that a great deal of this programing and debuging, along with data compilation and the biological equivalents of degragmenting, deleting of extranious data, and compressing are done while we sleep.

The biological equivalent of a bit is not a DNA pair. It is one nerve juntion on one nerve cell in the brain or central nervious system. A biological byte might be all the connections to a single nerve cell or it might be one (relative) position in the connections of many nerve cells. As the brain is self organizing in both memory storage and processor allocations it has proven to be very hard to translate that information into other media.

DNA is closer to the CAD/CAM file that is used to assemble a machine (in this case a person) than it is to the human personality, which would be more analigus to the output of mutiple data processing programs. People judge other people more by the perceved behavior of the entire system, data comunications (speach), movement, ect. than by any one factor. So perceved personally is more like the output of the system than any one part of that system.

Charles L.

45B846CE52C242.)
Reply to
Charles Prevatte

Read up other threads - nobody missed it. According to Chomsky, 95% of the personality is built in - there isn't enough input from anywhere and the time to learn all of it by the time a child reaches age 2. If you disagree with that, you would have to knock down Chomsky result first.

It would all be so simple if Chomsky's result is negated.

That is a picture, but consider from a robotic engineer's point of view. You are given a long tape with AGCT as the only symbols written on it. From that, you must construct a human being. The first things we have to look for is the machine construction itself. That seems to be OK. We know how to read DNA and make proteins that then make structures, that then make the complete organism through concentration gradients and so on. However, to recap the story of the human genome project, it came as a great shock that the human genome doesn't seem to contain the blueprints for the software. With 100 billion neurons and 10,000 connections per neuron to program up, and Chomsky's result that 95% is already programmed up at birth, you can see the shock in the human genome camp.

The question now is whether DNA is stupid and doesn't contain this information, or that it is so clever in encryption, that we haven't found where personality is stored on the genome. Further more, the genes themselves cannot be altered to pass on information. Even a single change could result in death. More difficulty is the genes themselves. You can't read out bits of it - the complete gene must be read out and turned into a protein. This results in great information loss. Effectively, you only have 20,000 proteins to make the brain. To confound the problem even further, 99% of these proteins are the same as in a chimp.

Even to say the boot strap program is pre-loaded is making the assumption that you know where that software is stored in the DNA. But to discourage you persuing that simple avenue is Chomsky again saying that a whopping 95% is already pre-programmed. The software contains nearly the complete program

- not just a bootstrap program. Each protein may be responsible for some information storage of personality - but that is a giant leap of faith considering they are randomly interacting molecules. Thus they end up being modelled as bit and you have 20,000 bits of information to play with to make a human being.

I downloaded the entire genome onto one CD. I've downloaded rasmol if I need it to view it. But where to next?

Reply to
e7

Which "Chomsky result" do you mean? Citation please. He's made a lot of wild claims in addition to his more solid work. I'm pretty sure that the result you're talking about _has_ been disproved already, but without knowing which one you're referring to it's impossible to say.

For an introduction to a few of the alternatives to and debunkings of Chomsky's ideas, I recommend the works of George Lakoff. For example:

"Women, Fire, and Dangerous Things"

formatting link
"Philosophy in the Flesh"
formatting link
There are other theories than the (untenable) one that all experience must be hard-wired at birth.

Reply to
Carl Burke

Talk about a broken record. Are you even reading the responses to your posts? All of your questions have been answered.

Although, this sure makes a good drinking game.

Every time you mention Chomsky, take a shot. Every time you say "95% of the personality is built in", take a shot. Every time you mention the "human genome project", take a shot. Every time you mention the 1% difference between the human and chimp genome, take a shot. Every time you say "20,000 bits of information", take a shot.

Reply to
Chris S.

Reply to
Guy Macon

WARNING: the above procedure will result in alcohol poisoning!

"e7" is ineducable.

Reply to
Guy Macon

formatting link

Reply to
e7

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.