Main Area and Open Discussion > Living Room

The Story of Merlin: The 1978 Electronic Game

<< < (2/4) > >>

TaoPhoenix:
Starting a *whole* new topic, let's look at this quote:

"The notion of the mind as a computer, or of a computer intelligence, has fascinated me my whole life,” Doyle says.

I *So* believe in this. I truly believe that people are incredibly like computers, involving "hardware problems" and "software problems", and when faced with five misc struggles by people, the minute you ask that question, parts of it all snaps into place.

I absolutely believe that the related but sometimes maligned Loebner prize is important, with a chunk of "defensive code" that people do automatically. (Background. It's a variant-TuringTest competition, but it rewards too highly people who ask nonsense questions. Put a bit of push-back in there, and things might change a bit.)

And my secret weapon in AI is the much maligned but maybe surprisingly important P1 of "math error" fame. Because that's what "intelligence" does. Makes errors. Then just build in some meta-code that expresses doubt about its results and then you're on a new path.

Going a bit darker, I (from my dangerously arrogant flippant layman's chair) think that "AI" is just on the verge of totally owning a whole class of service workers. I still say I welcome PM's / other going into AI topics. (With no way to prove anything, I still think I have useful algorithms on this in my head that no one has considered.)

The Cyc project struggled a little in the early days, but we're just turning those corners. Somewhere between Watson that beat Jeopardy, plus X custom code, the "McDonalds Algorithm" isn't that hard, if you skip the robotics and abstract it a bit.

Heh - I chatter to much. More some other day. :  )

40hz:
Based on what I know about human minds and computers, I think they're completely different mechanisms that operate (to date) on entirely different principles. Beyond a certain surface similarity of outputs and responses (which you can find looking most things - if you look hard enough and aren't too fussy) I can't think of anything further apart. And there's still Turing's 'halting problem' for a machine to get around.

But while I'm not too hopeful for genuine AI in the foreseeable future, I certainly am for a-life research. Self-organizing autonomous virtual machines (i.e. bots, agents, virus-like cyber-mechanisms, adaptive sensing devices, etc.) are an entirely different proposition. That's where I think we'll see the real breakthroughs in the next twenty or so years.

In fact...OMG! Look at the time! Yeah, let's leave this for some other day ... ;D

TaoPhoenix:
Based on what I know about human minds and computers, I think they're completely different mechanisms that operate (to date) on entirely different principles. Beyond a certain surface similarity of outputs and responses (which you can find looking most things - if you look hard enough and aren't too fussy) I can't think of anything further apart. And there's still Turing's 'halting problem' for a machine to get around.

But while I'm not too hopeful for genuine AI in the foreseeable future, I certainly am for ... research.

In fact...OMG! Look at the time! Yeah, let's leave this for some other day ... ;D
-40hz (March 04, 2014, 09:53 PM)
--- End quote ---

Heh joke at the end, ends too much research!

Sadly, human minds are much overrated, and we're risking "No True Scotsman" calling "Human Minds" to be the best 12%! (Tax season has shown me this!)

So I still think we're all falling back to my semi-paranoic theory that we're afraid of super-AI. Armchair wise, I like to say that with 10K, I could overturn some of these theories. But then, I'm just a ranter and I don't matter. But I think it's true. 170k lines of code could change the world.

I think we might have to Agree to Disagree.

Work into Siri is along the lines I have been saying, with a bit to go. But why is "Siri" different? Maybe for once not a cute little nonprofit, but a Big Biz entity working on it?

40hz:
^I don't think I'm being paranoid or giving undue emphasis to what passes for human intelligence. I just don't think the current state of AI is that great as to be overly accommodating towards the assertions being made for it. Because from my own personal perspective (admittedly as little more than intelligent and interested bystander) it ain't nowhere near there yet.

FWIW I have a purely personal belief that 'intelligence' (or something close enough it may as well be) will ultimately be shown to inevitably emerge from from any system once it reaches sufficient power, capacity, and complexity. But that's just as much my accepting something on faith as somebody else insisting human intelligence is unique and non-reproduceable by engineering or technology. Truth is...we just don't know.

Now I'm not saying it can't be done. Just that it hasn't - and it doesn't look like there's even been a significant breakthrough in the last 20 years or so to speed the advent. Most of the approaches being taken seem pretty brute force from the readings I've done.

But that's me. I have a bias for elegant solutions.
 8)

Innuendo:
Back to the OP, I had one of those! Bought it with my allowance and I loved it! Played it every single day for years.

Navigation

[0] Message Index

[#] Next page

[*] Previous page