Neal Stephenson on Artificial Intelligence (1997)

From the 1997 documentary Understanding : computers

Category:

Science & Technology

Tags:

• Neal
• Stephenson

[oldthinkertube]

There's a chance he was being at least partially facetious. He's also operating on the same concept line that he used in The Diamond Age, whereby he surmised that, if you could force human brains to devote most of their processing power to a single problem AND to "cluster" with a large number of others in the same way we can cluster servers or teamed microprocessors, you have the most powerful computer ever made.

[Tiddlefish]

do we need to understand the universe this deeply???

[AClarke2007]

Let`s take it one step at a time; teach the brain to masturbate first.

[FractalBolt]

I'm just going to sit back and enjoy the show of these youtube idiots taking everything personal and arguing about what Stephenson said more than ten years ago in some corny TV show hosted by Gilian Anderson on the Discovery Channel... fun times.

[Zarrov]

@mrhnm indeed your right

[mrhnm]

@Zarrov Text is hardly the medium for which to transmit sarcasm.

[Zarrov]

@mrhnm yeah, your comment just confirms what I said before. If you are one of those guys working on AI->its no wonder you have uch poor record, since your own inteligence is that of a plant, since you are unable to grasp concept of a joke. Have fun with building AI as useless as you.

[mrhnm]

@rivethead1982 He has a point to an extent and a limited one at that. Building AI will be a watershed moment for all of science and AI is more powerful than any human being could ever be probably more powerful than we can imagine.

An AI with enough processing power can reduce the time it takes to research anything.

[mrhnm]

@Zarrov You are a dumb ass. The point of AI is to build something with the processing power of a computer the memory capacity of a computer and the intelligence of a human. AI has infinite possibilities and infinite power.

[PraxisAxis]

I think he's really just talking about the practicality of it, and not so much being very philosophical. It's like the flying car: interesting sci fi device, plausible, but not a practical option for various reasons. Normal cars and planes already do the job.

Human-like AI is the same as the flying car: People's brains and computers (perhaps better, but much as they are today function-wise) will cover most problems.

[tyrbolo]

LOL so if we want an AI for a war machine we just grab a child and wire it into

the machine. Seems highly inefficient, because the life support would decrease

the ammunition storage capacity. The response time is also not optimal.