report errors(forelton

[Guest loonyl]

Computer people play with their bits,

when they can't get a nibble

 

Yeah i know ...sad !

 

I thought the in thing of the past decade

was Transputers !! They were SEXY !!!!

 

The rest is for Geeks only !!!

.

.

.

.

.

.

.

.

.

.

.

.

.

.

Last chance to look away

.

.

.

.

.

.

.

.

.

.

.

.

To increase processing speed the tracks get nearer and nearer, but then tunnelling occurs ! I heard they already take into account probablility functions etc when designing new chips.

 

Why don't they use a different base to 2?

that would increase their power a lot.

I heard the initial Eniac or whatever was base 10?

 

In 20 years, according to Tommorows World , robots will be competing against

humans in the olympics.

 

And in 50 years they will over take us.

 

If you look at evolution of animals, they take millions of years, computers have been here a few years, and they already have Virus, Worms etc.

 

In fact bugs or illegal operands may be the *spark* that gives machines intelligence !

 

And did you know that most ducks have very very nasty viruses (virii) in them, and if they jumped the species barrier, we'd be in ..., well worried !

 

(I'm not sure about that last bit, so i just made it up for effect )

 

Cheers

Liam

I'm going to lie down for a bit now,

me brain hurts !!

 

Bugs in the system is coz ants etc are drawn to the static , electo-magnetic fields

produced by electrical devices.

[Guest Ian Cresswell]

Originally posted by loonyl:

To increase processing speed the tracks get nearer and nearer, but then tunnelling occurs ! I heard they already take into account probablility functions etc when designing new chips.

 

Nope ... nothing like that scale yet. That's the reason why device surface area continues to increase. That's not going to be a problem at the scale that these guys are working at.

 

Why don't they use a different base to 2?

that would increase their power a lot.

 

It wouldn't. All maths is the same regardless of base. That would make no difference. The whole thing is based on an on/off logic system anyway. Well, apart from neural nets etc ... they're a little different.

 

In 20 years, according to Tommorows World , robots will be competing against

humans in the olympics.

 

And in 50 years they will over take us.

 

Sounds like Kevin Warwicks predictions validated by his best media coverage yet. Read his books ... be convinced. Serious academic (Reading Uni) ...

 

If you look at evolution of animals, they take millions of years, computers have been here a few years, and they already have Virus, Worms etc.

 

Those are human copies of evolved things not acts of evolution itself ... however the idea that such things may evolve isn't as far fetched as it may sound.

 

In fact bugs or illegal operands may be the *spark* that gives machines intelligence !

 

See 'Genetic Algorithms' (the introductory work by Goldberg is a good place to start here). Evolutionary computation covers this and is actively being worked on. There's plenty of this stuff on the web.

 

Some would argue that 'Intelligence' is a difficult word. Before applying it try to define it. The Turing test just doesn't cut it.

 

Others would argue that regardless of the definition that there is nothing 'Artificial' about 'Intelligence'.

 

Hint: When thinking about this don't judge things by human standards.

 

Cheers,

 

Ian.

[Guest Newt]

Originally posted by Ian Cresswell:

Some would argue that 'Intelligence' is a difficult word. Before applying it try to define it. The Turing test just doesn't cut it.[/b]

 

Ian - as I remember the Turing Test - it basically stated that if you were talking (via voice, kbd, ouija board, whatever) to a computer and could not tell it was a computer, then it had Intelligence. This always seemed a reasonable test to me. How is it not? OTOH, I've spoken to people who convinced me after a few words that they were a cyborg using an 8086 processor with 64Kb of memory - i.e., they were pretty stupid.

[Guest loonyl]

To increase processing speed the tracks get nearer and nearer, but then tunnelling occurs ! I heard they already take into account probablility functions etc when designing new chips.

 

Nope ... nothing like that scale yet. That's the reason why device surface area continues to increase. That's not going to be a problem at the scale that these guys are working at.

 

Ohhhh yes,

they alredy take into account 20% probability error

 

Why don't they use a different base to 2?

that would increase their power a lot.

 

It wouldn't. All maths is the same regardless of base. That would make no difference. The whole thing is based on an on/off logic system anyway. Well, apart from neural nets etc ... they're a little different.

 

Nope again

2 to the power of 2 = 4

is less than 2 to the power of 10 = 100

 

In 20 years, according to Tommorows World , robots will be competing against

humans in the olympics.

 

And in 50 years they will over take us.

 

Sounds like Kevin Warwicks predictions validated by his best media coverage yet. Read his books ... be convinced. Serious academic (Reading Uni) ...

 

}

Nope, just look at statistics,

and the % increase over a given decade.

 

If you look at evolution of animals, they take millions of years, computers have been here a few years, and they already have Virus, Worms etc.

 

Those are human copies of evolved things not acts of evolution itself ... however the idea that such things may evolve isn't as far fetched as it may sound.

 

Nope,

A missing link thats all,

between human and super intelligence.

Not far fetched, Reality.

Why do you think the current intelligent life form is so addictive to the future life form? a mere stepping stone.

 

In fact bugs or illegal operands may be the *spark* that gives machines intelligence !

 

See 'Genetic Algorithms' (the introductory work by Goldberg is a good place to start here). Evolutionary computation covers this and is actively being worked on. There's plenty of this stuff on the web.

 

As soon as quotes come from people of this paradigm, like in years gone by "like the earth is flat" etc , you limit yourself for being within the model.

 

Some would argue that 'Intelligence' is a difficult word. Before applying it try to define it. The Turing test just doesn't cut it.

 

Nope just like Alchamy or astrology doesnt cut it.

It just a guestimate on the model created by clever people of their time.

Intelligence isn't quoting from what's been before, or quoting *geniuses*,

progress only comes from

"standing on the shoulders of giants"

or jumping off if they are about to sink.

 

Others would argue that regardless of the definition that there is nothing 'Artificial' about 'Intelligence'.

 

Nope, its a bye product, something that us animals like to think we have the monopoly on, till of course something more intelligent comes along.

 

Hint: When thinking about this don't judge things by human standards.

 

Hint: when thinking about intelligences greater than we can imagine, don't expect to find it from reading a book or two.

 

Cheers,

 

Ian.[/b]

 

Cheers

Liam

[Guest loonyl]

Whoops Sorry Ian,

 

A mental note to myself

 

DONT post when you're out of your tree !

 

My head hurts today, must be something I ate

 

Cheers

Liam

[Guest Ian Cresswell]

Originally posted by Newt:

This always seemed a reasonable test to me.  How is it not?

 

Hi Newt,

 

That's close. The proof of intelligence is based on a good sample of humans. For success the idea is that you get a 50/50 split of those testers that got it right or wrong.

 

The problem with the test is that it is restricted to comparing humans against machines.

 

In other words, there are fundamental assumptions being made about the nature of intelligence and the relationship between human intelligence and other intelligences. The 'i' word is the problem.

 

I've got some pals on the academic staff at CMU and I had the good fortune to talk to some of the real AI gods when I visited there a couple of years ago. Amongst these was Jay McClelland (leading computational neuroscientist) and asked him for his definition of intelligence ... after a pause and a chuckle he replied:

 

'intelligence is as intelligence does'

 

A very clever twist on Forrest Gump.

 

Cheers,

 

Ian.