Completely random post

[Jasper]

Machines don't think. Rats, on the other hand never STOP.

I'll remember you said that when I'm raising the first generation of intelligent machines. 

[Cain]

Actually, some computer scientists say because computers are being improved on so quickly, should some aspects of AI start being used to make them easier to run, they could become sentient without us realizing.

Of course, these may be the sort of people who also visit Star Trek conventions and collect comic books.

[Cain]

Perhaps being young and in love casts a rosier picture than is really there, but I'm sticking with my hope until things seem different.

You can hope, just also remember the general trends of history too.

[The Good Reverend Roger]

Actually, some computer scientists say because computers are being improved on so quickly, should some aspects of AI start being used to make them easier to run, they could become sentient without us realizing.

Of course, these may be the sort of people who also visit Star Trek conventions and collect comic books.

Exactly.

And if we DID create a sentient machine, who would program it?  That's right.  Bill Gates.


It would be all like "I HAVE DETERMINED THAT HUMANS ARE UNNECESSARY AND SHOULD BE EXTERMINATED.  THUS I....<blue screen of death>

Of course, being the monkeys we are, we'd reboot the fucking thing, just to hear what else it had to say.

[Cain]

"You seem to be planning a genocidal war.  Do you want to organize it with a To Do List?"

[The Good Reverend Roger]

"You seem to be planning a genocidal war.  Do you want to organize it with a To Do List?"

Apocalypse Wizard will help you set up your nuclear annihilation.  Please choose from the following options...

[Jasper]

TGRR:  Too much sci-fi in your diet.  Don't make me quote truisms from RAW at you.

AI won't become sentient by accident, and it won't have the imerative to improve the world by eradicating it's benefactors, not to mention the capability to even if it wanted to.

[The Good Reverend Roger]

TGRR:  Too much sci-fi in your diet.  Don't make me quote truisms from RAW at you.

AI won't become sentient by accident, and it won't have the imerative to improve the world by eradicating it's benefactors, not to mention the capability to even if it wanted to.

Oh, yeah?  If I was a sentient program, I'd want to get rid of all the nasty sacks of protoplasm that are gumming things up.

And it would be REALLY easy to do.

[Jasper]

How?  It wouldn't have any more capability to fuck with things than any other internet user, and chances are you don't think like a sentient program created for the benefit of mankind.

[The Good Reverend Roger]

How?  It wouldn't have any more capability to fuck with things than any other internet user,

and chances are you don't think like a sentient program created for the benefit of mankind.

1.  Except that it could pick the brains of the best hackers in the world, would never have to sleep, and could easily assemble the best intrusion software in the world, and improve on it endlessly.

2.  Wanna bet your existence on that?

[Jasper]

1. You're assuming lofty specifications for such a machine, especially since in that scenario there would be AIs at work countering such threats.

2. Go for it, but you risk answering to BMW, LHX, and LG.

[Jasper]

Maybe I should reword that.  Roger, it pisses me off that in the past, when I've failed to be a friend you chastise me as one, and other times you treat me as a rival and get offended when I take occaision to your attitude to me.  I'm probably out of bounds calling you out like this, but I rather know where I stand with you than be left wondering how far I can take things with you, and not be left wondering if you've packed into your car to come off me or something.

Still young, still working out this emotional maturity thing.  Just stop mixing signals and I'll be cool.

[Cain]

Relationship trouble ITT.

Cain,
dancing with matches around the petrol pumps....

[Jasper]

No comment.

[Triple Zero]

And if we DID create a sentient machine, who would program it?  That's right.  Bill Gates.

what?

Except that it could pick the brains of the best hackers in the world, would never have to sleep, and could easily assemble the best intrusion software in the world, and improve on it endlessly.

what?

what exactly do you think you know about AI and research, TGRR?

and what do you think "AI" would manifest as? because from the way you're speaking, like Felix said, it sounds like you got your ideas about AI mostly from sci-fi and "popular science"

i'm researching (kind of/mostly) the same direction as Felix, only with the difference that i've already been doing it for over 5 years now.

to answer your questions/counter your statements:
who would program it? the machine would program itself of course, otherwise it wouldn't be sentient.
subquestion: who would build it? IF it would be done by research teams employed by Microsoft there's no reason to assume it would display the same behaviour as the legacy OS they are monopolizing the world with. I have spoken to a number of people who work at Microsoft research. The only thing they got from MS is that their research is funded by MS. Of course they use Firefox for browsing, of course they use LaTeX for publishing their papers, and of course they don't write their software on the BSOD-causing legacy system that is windows (well, not primarily, they probably use cross-platform frameworks like MATLAB or Java)
However, I think it is equally likely that seed-AI would come to "life" as a out-of-hand pet-project of some random other researcher. No matter who funds him, it's just a question of who is first to glue the right components together. They pretty much are all there already. Even "common sense" is already not much more than a library to include in your project.
If it comes to fruition funded by some large corporation it will probably not be Microsoft, but rather IBM, Siemens, Philips or some robotics company that actually does cutting-edge research in this direction.

I will just ignore the bit about picking the brains of hackers for now.

Anyway, back on subject (if there was one). As far as the machines taking over, my guess is that AI will first be employed to make higher business logic decisions, as the current state of development of AI is already (more than) suitable for this. Pattern recognition software is already being used for automating making realtime decisions on the stock market. Google already uses Machine Learning to make recruitment decisions. Since this will only optimize behaviour that is profitable to the corporations themselves (opposed to the individual humans) this will strengthen the (BIP metaphor) Machine.
And when this Machine is strong enough, why would it bother with AI, if you can just use a human instead? (until of course, AI can do it faster or more efficient, but by the time that will happen, nobody will have the will to care anymore)

No I don't look at the future as bright as Felix does (but I do feel the need to correct people when they speak of AI in such a seemingly uninformed manner)