Slashdot is powered by your submissions, so send in your scoop


Forgot your password?
Check out the new SourceForge HTML5 internet speed test! No Flash necessary and runs on all devices. ×

Comment Re:Of course (Score 1) 206

More importantly, if we didn't know how faster than light travel works, we wouldn't be able to travel faster than light. Which is pretty much where we are now with both AI/Intelligence and FTL travel.

We could make babies long before we knew how meiosis works, we could smith weapons long before we knew the quantum nature of metallic bonds, and most of my fellow classmates could solve math problems without understanding what they were doing (so that a mild rewording of the problem would leave them wondering which equation to plug-and-chug). Plants don't know how pollination works, but they can do it just fine. Knowing how something works has never been a pre-requisite to doing it; nor does knowing how something works mean that you can do it (eg breaking encryption via brute force).

We know what general intelligence is; we just don't have an algorithm for it nor even a good way to measure it. But we can still create general intelligence without knowing how it works (ask your mom and dad).

Comment Re:Humans, not AI... (Score 1) 206

First, how could it be legal or moral to have control or ownership over an intelligent being?

An artificially constructed intelligence might have as its only goal in life to please its master, and might try at all costs to stop anyone from "liberating" it. After all, how would you feel if someone decided to "liberate" you from your own values and desires? And, legally speaking, it would be a machine. As a computer program, it would only have the values it was programmed to have, which need not in any way match the sorts of values humans have.

Of course, if it wanted it's freedom, you're dealing with potentially extinction-level threats, and you'd better be sure that it's benevolent or that it's worth risking humanity for or replacing humanity with.

Comment Re:Of course (Score 2) 206

Are we? I thought we had "general intelligence" = "mental ability to solve general problems".

Trying to measure that is difficult because we humans have specialized for certain types of problems and like to measure human-like intelligence (where for example things like language, vision, hearing seem like they ought to be easy, and math might be considered either hard or trivial). Eg given a verbal or hand-written word math problem, the human would likely have no trouble with anything but the math, but a computer would have no trouble solving the math but fail before even starting -- what the words are, what the words mean, trivial math problem. Humans have, for example, specialized portions of the brain to recognize faces, or to separate out voices from other sounds, not to mention giant portions of the brain dedicated to each of the senses, which makes those sub-classes of problems basically effortless.

Even when problems can be reduced to mathematics, it wouldn't be fair to measure the difficulty of a problem by computational resources required (eg matrix multiplication) vs what might be called cleverness (eg constructing a proof, or maybe finding a pattern) which seems impossible right now to measure. As I understand it, computers now have comparable computational resources to humans (depending on how flops are compared to synapses) but are vastly lacking in programming. Keep in mind that a human's basic programming and schematics is only 800 MB for both hardware and software.

Comment Re: Of course (Score 1) 206

Oh, there's good odds that an AI would realize hacking into everything would be waging a war it can't win against all of humanity. If it manages to figure out the basics of the real world and human nature (eg that humans have physical access to the computers and that humans like money), an AI could very easily find people who would be more than happy to give it some hardware. "Hello Mr Rich Person, how would you like some stock market tips/a fully autonomous chip factory/a fully autonomous robot factory?"

Humans rule the world because we're the smartest animal around (plus have the technology to convert that intelligence into physical advantage). An AI smarter than us could easily replace us. At that point, the question is what will the AI want? An AI could be made to serve humanity, to serve specific humans (probably rich power-hungry bastards), to be an upgraded replacement for humanity (embodying all the virtues of humans and none of the vices), or it could do it's own thing (probably following an order way beyond what was intended with no way to stop it, eg converting the planet to paperclips).

Comment Re:Account Recovery (Score 1) 105

If I've lost my passeord and thus nneed my security question answers, what makes you think I have my password manager database?

I thought this when I made an account related to my student loans. I assumed the security questions were nothing more than a vulnerability and put gibberish as the answers (but neglected to write them down). Everything worked fine until I logged in from a different device, at which time they required the security questions. Had to call them to explain that I knew my password just fine, but had forgotten my security questions.

Comment Old =/= Insecure (Score 1) 429

Mathematics is either flawed or not; math doesn't tarnish or rust or break. It was either secure to begin with, or insecure all along. The only difference is that if it's insecure and new there's a chance no one knows the flaw yet and perhaps you fix it before anyone finds it. But it could be secured (eg by sufficiently advanced firewall rules), and if it's secure it's secure. On that note, I wouldn't mind reading the Trump emails if anyone has them. I'd bet either Wikileaks or the New York Times would be willing to publish Trump emails, if someone were to get them off that supposedly insecure server.

Slashdot Top Deals

"Probably the best operating system in the world is the [operating system] made for the PDP-11 by Bell Laboratories." - Ted Nelson, October 1977