The Future of Technology

[Darrell Hougen]

Steve Omohundro, my former master's thesis advisor, discusses the future of technology. I love his optimism and enthusiasm and he always seems to have his finger on the pulse of new technology. The discussion does raise a lot of interesting questions about where the technology is going and where we want it to go. I don't necessarily agree with all of his political views --- he tends to be a bit left of center --- but the talk is interesting nonetheless. It also touches on other topics that have been discussed recently such as the nature of consciousness.

Darrell

[BaalChatzaf]

Unfortunately the progress in real intelligence has not progressed all that well.

Here is the problem. The brain is NOT a computer. It works on a completely different set of principle.

AI started with the famous McColloch and Pitts paper on "neurons" which were not neurons but a linear threshold logic switch. There have been some interesting developments which have produced some clever programming technique, but alas none of this has made our grasp of organic intelligence any better.

Lets see AI has been around since 1950 which means 64 years and it hasn't come a step closer to revealing what real live intelligence is. It is somewhat like controlled nuclear fusion --- forever 30 years in the future.

Ba'al Chatzaf

[Darrell Hougen]

Unfortunately the progress in real intelligence has not progressed all that well.

Here is the problem. The brain is NOT a computer. It works on a completely different set of principle.

AI started with the famous McColloch and Pitts paper on "neurons" which were not neurons but a linear threshold logic switch. There have been some interesting developments which have produced some clever programming technique, but alas none of this has made our grasp of organic intelligence any better.

Lets see AI has been around since 1950 which means 64 years and it hasn't come a step closer to revealing what real live intelligence is. It is somewhat like controlled nuclear fusion --- forever 30 years in the future.

Ba'al Chatzaf

Hi Robert,

I'm not sure what you mean when you say, "The brain is NOT a computer. It works on a completely different set of principle." In the following paragraph, you seem to be implying that neurons are different from transistors. That much is true. But, it doesn't follow from that observation that computers cannot be made to act like brains. In fact, there are good reasons to believe that computers can be made to either emulate brains or, more to point, act intelligently.

A brain is a physical system and computers can be programmed to emulate physical systems. Any continuous valued function can be simulated to arbitrary accuracy using a discrete system. So, if all else fails neurons could be simulated using computers. Granted, you would need a lot of computing power to simulate all the neurons in a human brain, but I don't see any "principle" that would make such simulation impossible.

The fact that people have been working on AI for 64 years doesn't particularly bother me. The field of chemistry was around for at least 200 years before Rutherford and Bohr came up with the nuclear model of the atom, thereby revolutionizing the field. The fact of the matter is that the problems of AI are difficult. However, that doesn't imply that they are unsolvable and I would be surprised if it takes that long to solve them.

Darrell

[BaalChatzaf]

Unfortunately the progress in real intelligence has not progressed all that well.

Here is the problem. The brain is NOT a computer. It works on a completely different set of principle.

AI started with the famous McColloch and Pitts paper on "neurons" which were not neurons but a linear threshold logic switch. There have been some interesting developments which have produced some clever programming technique, but alas none of this has made our grasp of organic intelligence any better.

Lets see AI has been around since 1950 which means 64 years and it hasn't come a step closer to revealing what real live intelligence is. It is somewhat like controlled nuclear fusion --- forever 30 years in the future.

Ba'al Chatzaf

Hi Robert,

I'm not sure what you mean when you say, "The brain is NOT a computer. It works on a completely different set of principle." In the following paragraph, you seem to be implying that neurons are different from transistors. That much is true. But, it doesn't follow from that observation that computers cannot be made to act like brains. In fact, there are good reasons to believe that computers can be made to either emulate brains or, more to point, act intelligently.

A brain is a physical system and computers can be programmed to emulate physical systems. Any continuous valued function can be simulated to arbitrary accuracy using a discrete system. So, if all else fails neurons could be simulated using computers. Granted, you would need a lot of computing power to simulate all the neurons in a human brain, but I don't see any "principle" that would make such simulation impossible.

The fact that people have been working on AI for 64 years doesn't particularly bother me. The field of chemistry was around for at least 200 years before Rutherford and Bohr came up with the nuclear model of the atom, thereby revolutionizing the field. The fact of the matter is that the problems of AI are difficult. However, that doesn't imply that they are unsolvable and I would be surprised if it takes that long to solve them.

Darrell

Computers can be programmed to emulate cartoons of physical things. Does anyone need a real umberella when a model hurricane is being executed on a computer? I doubt it.

Ba'al Chatzaf

[deanwins]

Hello Derrell,

Here are some of my responses to the video:

Can AI make new ideas: practically random generator creates information. Then a harder question is: can a machine make information that is both true (consistent with reality) and useful? Yes... but this is a hard problem/process that is very expensive (the scientific method)

Will machines make themselves better and get out of control / control the world: yes.

Can machines select their own goals: yes... via practically random generators and natural selection

Can machines have consciousness? yes... What is consciousness: having live/current information about the current state of reality

Cheers,

Dean

[dldelancey]

You gentlemen may find the new TV show Almost Human interesting. In the opening sequence a narrator says something like, "In the future, technology advances faster than it can be regulated." Regulated? Anyway, it's probably only mediocre as crime drama goes, but the tech, and more importantly, the ethics that come with the tech, is both fascinating and disturbing.

Then again, I'm kind of a techie-geek, so I may be overstating the appeal.

[Selene]

I second Deanna's endorsement of the show. I also like the "human" actor from his role in RED - Retired Extremely Dangerous - Willis - Helen Mirro really good movie.

Dodging those snow flakes in Louisiana?

A...

[BaalChatzaf]

You gentlemen may find the new TV show Almost Human interesting. In the opening sequence a narrator says something like, "In the future, technology advances faster than it can be regulated." Regulated? Anyway, it's probably only mediocre as crime drama goes, but the tech, and more importantly, the ethics that come with the tech, is both fascinating and disturbing.

Then again, I'm kind of a techie-geek, so I may be overstating the appeal.

I like that show. Some of the lines are very cleaver. Especially from the Android with blue splotches on his face.

Ba'al Chatzaf

[Michael Stuart Kelly]

Some of the lines are very cleaver. Especially from the Android with blue splotches on his face.

Bob,

Would you like to meat him?

Michael

[Jules Troy]

Good show!

I also like the show "Intelligence".

[BaalChatzaf]

Some of the lines are very cleaver. Especially from the Android with blue splotches on his face.

Bob,

Would you like to meat him?

Michael

Meat him? That is a good one. Dorian is my kind of people. Just like Spock and Data are my kind of people. Some of my best friends are Androids or people who can pass for Androids.

Ba'al Chatzaf

[Michael Stuart Kelly]

Bob,

This one cuts too good.

I gotta do it one more time.

It's always nice to meat cleaver people.

Michae;

[BaalChatzaf]

Bob,

This one cuts too good.

I gotta do it one more time.

It's always nice to meat cleaver people.

Michae;

That is very sharp of you. That is what gives you the edge you have.

Ba'al Chatzaf

[tmj]

Or a hack with a keen wit.

[Jules Troy]

Axe scent on the wrong sill able?