Re: I can build an AI machine!!!!
From: Amin (amin_at_afga.com)
Date: 10/11/04
- Next message: Sam Wormley: "Re: underwires, do they work"
- Previous message: Sam Wormley: "Re: Inertial Frames"
- In reply to: Joe: "I can build an AI machine!!!!"
- Next in thread: jimp_at_specsol-spam-sux.com: "Re: I can build an AI machine!!!!"
- Reply: jimp_at_specsol-spam-sux.com: "Re: I can build an AI machine!!!!"
- Reply: Mike Lepore: "Re: I can build an AI machine!!!!"
- Messages sorted by: [ date ] [ thread ]
Date: Mon, 11 Oct 2004 00:27:46 GMT
The AI subject in the college is almost dead. It is not Physic. I haven't
seen any new break through AI theory from the researcher for the last 20
years.
"Joe" <jhelfand@umd.edu> wrote in message
news:e62610ea.0410101234.2a45c7ea@posting.google.com...
> I have always been mildly fascinated and wondered about how to build
> an AI machine, or at least one that looks like it has AI, one that for
> example can fool the experts. This time, I think I found it, it
> should work. Basically I have done some experiments in the lab, and
> analyzed some theory, (I have some equations to back it up, but the
> idea looks sound without them), and I am convinced I have a VERY good
> working plan. Now let me tell you what I think my machine can do. I
> think I'll be able to set you up in a terminal, in a room, and have it
> connected with some other complicated electronic interface (which I
> can build or aquire from another lab) to my machine. And then you
> will be able to chat with it, typing on the computer and looking at
> it's responses, and we'll see if you can determine whether it's a
> "human" or not, once I program it and it teaches itself to act like
> one (like even childhood memories, etc.). Furthermore, I see no
> reason why I can't put you on the phone in one room, and build some
> kind of complicated voice synthesizer, that mimicks the human voice, a
> sound processor with advanced voice recognition capability inside my
> machine, and you can talk to it live! Again a prize if you think it's
> not human. The machine "voice" will even have tone inflection,
> emotional quality, etc. I will be able to do all this once we program
> it to do language. In fact, and this will be even more interesting,
> me and the research team will be able to "anthropomorphisize" the
> machine, give it a human like form, with arms, legs, face, etc. And
> teach it/have it teach itself, to walk around like a human, so that
> from the distance you'd be hard pressed to tell the difference between
> it and an android. In fact, maybe even at close distances this will
> be hard. Like I can have it easily run around and kick a soccer ball,
> exactly like a human, and I wonder if you'll be able to tell the
> difference. And I can again offer prize to the person who thinks it's
> not human. Because there must must be great preception here (and you
> would be offered to explain your conclusions and we could evaluate how
> you came to that conclusion). Finally, we will take this machine to
> the greatest computer scientists, engineers, physicists, and
> mathematicians of the world, or anybody who cares, who can figure out
> how it works, and a prize to anyone who can figure out all the details
> of what we did. (I know this will cause great debates among the
> greats, like is it a Turing machine? Does gravity collapse the
> wavefunctions, etc. All kinds of crazy ideas, because the experts
> will be stumped of course. At least those who are interested.) Now,
> I have done some experiments, and I see that this will work, but I
> still need to get a hold of some additional equipment, for the
> details. I can't quite do it myself. I am looking for people to
> partner with me in this enterprise (admittedly, not a good job), to
> experiment with the equipment and even build an AI machine. Or at
> least a machine that looks like it has AI [see below]. Now funding is
> limitted, in part because I want to keep my idea secret, and also in
> part because I don't want my idea, the fruits of my labour, enslaved
> to any corporate, government lab, etc. (Even though ultimately it
> will, I'd like to stave off that possiblity for as long as possible.)
> So it will take some hard work building, and financing, with our own
> money, the machine. And yet for the love of life, it will be a great
> enterprise!
>
> Note: I know that when we build that machine, and if I claim it has
> AI, there will be a great quibbling over words. FOr example, what is
> AI? For example, just look at the web. However, if we say "behaves
> like a human" this will get closer. But still I am convinced there
> will be quibbling over words. So let me readress it to "look like it
> has AI", this will fulfill even the greatest skeptics, the greatest
> quibblers, and all will have to agree that the machine satisfies this.
> What do you think?
- Next message: Sam Wormley: "Re: underwires, do they work"
- Previous message: Sam Wormley: "Re: Inertial Frames"
- In reply to: Joe: "I can build an AI machine!!!!"
- Next in thread: jimp_at_specsol-spam-sux.com: "Re: I can build an AI machine!!!!"
- Reply: jimp_at_specsol-spam-sux.com: "Re: I can build an AI machine!!!!"
- Reply: Mike Lepore: "Re: I can build an AI machine!!!!"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|