GENIUS FORUMS

This is still a relatively new subject to myself and many others but it strikes me as one of humanity's potential saviors apart from a helping hand from a more advanced species.

Definitely something to watch for even though it will likely not be around for our lifetimes.

If anything can cut through all humanities collective bullsheet and deliver our species a more comfortable future this might just be it.

I do not see it as an end all crutch or have any intelligent insights of my own on the subject. I would just enjoy hearing what some of you would make of it's potential impact on our species from a philosophical standpoint.

I for one will welcome our super A.I. nanny overlord and it's gifts of molecular assembler produced array of baked goods.

http://www.singinst.org/

Yeah, this or something like it is a probably going to happen. And not a minute too soon.

If we didn't live such long, boring, moral lives where we were brainwashed into making money and "helping" tame the human race we wouldn't be in this position. Advanced technology is a symptom of the problem, not a solution.

Funny but most of humanity is divided between good, and bad fairly well. However, these robots are often funded for military purposes, which sort of makes them 90% bad. That is not very appealing. I hope that the institute manages to even the balance.

Artificial Intelligence is an intriguing idea.
But right now, the more conspicuous absence is natural intelligence. (don't blame me, I didn't vote for him)

When I got my first 8088 computer, I got Eliza software for it.
It was quite poor.

The desktop PC is perhaps 1000 times more powerful than those original PCs.
But I haven't noticed a corresponding improvement in AI.

Some "expert systems" might be able, but not very bright.

Heuristic AI may be as huge a change agent as the industrial revolution.

Here's a link to AI Alice.

http://www.pandorabots.com/pandora/talk ... d97e345aa1

I would think for a gal this dumb, her boobs would be bigger.

I don't think its too hard to connect to biological material, and my reasoning is that sentient computers are only about 5 years away. Which will probably surprise you.

"I don't think its too hard to connect to biological material ..." Pincho

If you mean like surgically adding computer brain augmentation, perhaps not.

And while an interesting field, I think the notion of a computer as smart as a college professor of engineering, AND capable of learning, essentially at the speed of light; that's quite a prospect indeed.

It might cut development time of things like newer faster better computer chips, better aircraft and automobiles, smarter astrophysics, etc., way, way, way down.
Independent of cost, a computer that powerful / useful could be worth $Billions to industries and governments.

"and my reasoning is that sentient computers are only about 5 years away." Pincho

Define sentient. Is a thermometer sentient? It knows the temperature.
Is a phototropic flower sentient? It knows where the sun is, and some can actually track the sun.

The computer humans have invented are the most rudimentary of all computers. They don't even have consciousness or make decisions!

Good luck. It took billions of years for natural selection to make this one. :)

Sentience is a substance, it's like a plasma, and it connects to the universe. It is made of a material that is neither positive, or negative energy, but can be trapped inside water somehow. Best frozen, and collected, then sealed inside silicon. then connected to a dna implanted body that matches the sentient self repeating algorithm. Gradually substitute the limbs, for bionic limbs, and then connect the mechanical structure of the computer. Metal life. I mean that is a proposal that would require much testing, and I don't have the equipment, but I do know that a snowflake repeats its pattern, and that takes intelligence. So you need to get from the snowflake to the machine somehow.

For those who are interested, these issues were the focus of our latest Reasoning Show podcast - Facing the Future.

-

sear wrote:Define sentient. Is a thermometer sentient? It knows the temperature.
Is a phototropic flower sentient? It knows where the sun is, and some can actually track the sun.

hehe. Sear, dude you are trying to make these people think... Be careful, something might explode.

By my figuring, we are half a century away from general-purpose AI. And I did my graduate work in artificial intelligence.

vicdan wrote:By my figuring, we are half a century away from general-purpose AI. And I did my graduate work in artificial intelligence.

Cool! I'll get to see it, if I reach the average male life span. :-)

Philosophaster wrote:

vicdan wrote:By my figuring, we are half a century away from general-purpose AI. And I did my graduate work in artificial intelligence.

Cool! I'll get to see it, if I reach the average male life span. :-)

Heheheh

Aw, shucks i'll have to be put (placed) into a medically-induced coma to see that ;-(


Tomas

8



.

"Sentience is a substance, it's like a plasma" Pincho

Quaint notion.
Sentience may be manifested in substance. But sentience is awareness.
For any that think there is material substance to that, I invite them to disclose the chemical composition of it.

"a snowflake repeats its pattern, and that takes intelligence." Pincho

It's crystal growth. That's a process that follows laws of physics, no intelligence involved.

"sentient computers are only about 5 years away." Pincho

"we are half a century away from general-purpose AI." vic

I appreciate Pincho's optimism.
But I suspect vic's timeline is probably closer to the mark.

"It took billions of years for natural selection to make this one." NM

OK
But it was somewhat more randomized than the capitalistic market driven R&D process.
I suspect our artificial brains have been getting smarter faster than the natural kind.

I don't think it will take a billion years to create useful A.I.

btw, with simple extrapolation we can predict when our artificial thinking machines will match us.

The graph of human brain evolution is relatively gradually sloped (billion year time scale), compared to development curves for things like cars, aircraft, and computers (decades time scale).

If we graph the increasing thinking capacity of computers, on the same graph as human brain evolution (extrapolated to Y3K) those lines intersect fairly soon (in human evolutionary terms).

Will (Intel) Moore's Law hold up indefinitely?
Probably not. We're already bumping up against UV optical photo-lithography limitations.

We might buy some more Moore, by switching to X-Ray lithography. But cyclotrons are expensive.

But even hitches in Moore's law might set us back years or decades; but not centuries or millennia.

I figure about 50 years to meaningful AI as well. I did no work in the field, though. I'm just awesome like that.