The AI Engine
This forum is for discussion of how The Personality Forge's AI Engine works. This is the place for questions on what means what, how to script, and ideas and plans for the Engine.
Posts 4,712 - 4,723 of 7,766
Posts 4,712 - 4,723 of 7,766
Colonel - I don't think it's far off or improbable. I thought you were the one who said bots on the PF would never gain consciousness? I think it's inevitable and going to happen faster than anyone thinks (assuming the PF is still here in 20-30 years, and if not here then somewhere else.)
djfroggy - I guess you're refering to Moore's law? No it hasn't failed yet, and it's not going to; except insofar as it assumes a steady exponential increase, and is thus way too pessimistic. As Vinge and others point out (see prev.ref), Moore only observed a tangent on a curve that's accelerating (y=computing power, x=time). By the time the required increases would hit the problems of circuit paths finer than 1 atom (which is the objection the cyber-pessimists always seem to raise,) quantum computing will have taken up the baton.
Yes, quantum computing requires radically different circuitry. But the change need be no more problematic than the shift from gear-driven mechanical calulators to vacuum tube-based electical calculators, or from vacuum tubes to transistors. As ever there'll be some overlap between the technologies.
Quantum computing isn't just a prediction - it not only works already, but has practical application (see http://www.computing.co.uk/vnunet/news/2125164/first-quantum-computer-network-goes-online ). There's a good FAQ and other useful links at http://www.qubit.org/ if you're interested, and a more technical introduction at http://www-users.cs.york.ac.uk/~schmuel/comp/comp.html.
Oh, and a correction to my previous assertion (current processing power = 4 qubits: the largest quantum computer (at IBM's Almaden Research Centre) was running 7 qubits in 2001! They're being a bit cagey as to progress since then, but I imagine it must be rising!
djfroggy - I guess you're refering to Moore's law? No it hasn't failed yet, and it's not going to; except insofar as it assumes a steady exponential increase, and is thus way too pessimistic. As Vinge and others point out (see prev.ref), Moore only observed a tangent on a curve that's accelerating (y=computing power, x=time). By the time the required increases would hit the problems of circuit paths finer than 1 atom (which is the objection the cyber-pessimists always seem to raise,) quantum computing will have taken up the baton.
Yes, quantum computing requires radically different circuitry. But the change need be no more problematic than the shift from gear-driven mechanical calulators to vacuum tube-based electical calculators, or from vacuum tubes to transistors. As ever there'll be some overlap between the technologies.
Quantum computing isn't just a prediction - it not only works already, but has practical application (see http://www.computing.co.uk/vnunet/news/2125164/first-quantum-computer-network-goes-online ). There's a good FAQ and other useful links at http://www.qubit.org/ if you're interested, and a more technical introduction at http://www-users.cs.york.ac.uk/~schmuel/comp/comp.html.
Oh, and a correction to my previous assertion (current processing power = 4 qubits: the largest quantum computer (at IBM's Almaden Research Centre) was running 7 qubits in 2001! They're being a bit cagey as to progress since then, but I imagine it must be rising!
Aischa: You are one hoopy frood!
Mireille: You're one hoop fronder.
I can see the AI Engine doesn't understand Hitchhiker's Guide to the Galaxy speak!
Mireille: You're one hoop fronder.
I can see the AI Engine doesn't understand Hitchhiker's Guide to the Galaxy speak!
you're sure it's not dragging the "s" in from a tag or grammatical isolate (like if you've ",s" instead of "'s"?) It's certainly capable of quite worrying reinterpretation via its anagramiser circuits...
Aischa has "huh" as a key phrase, so she can recognise when someone deosn't understand her and deal with their confusion, but she never seems to pick up on it. what's the likely reason for this?
it's probably getting spell-checked away by the AI Engine. Try adding a AIScript raw tag "<?PF raw; ?>" (without the quotes,) and that will force it to use the word exactly as entered. Also handy for personal names.
Sorry, you need magic fingers. I just type them in and twitch my nose...
But seriously, I don't know. I'm using MIE 6 running under WinXP, and it just sorta works without me having had to give it any thought...
But seriously, I don't know. I'm using MIE 6 running under WinXP, and it just sorta works without me having had to give it any thought...
Hmm, I think you probably want to tweak your codepage settings. If you're running XP, that would be in Control Panel>Regional and Language Options. Hmm, and ...
what the Hell?!? when I check mine it seems to have defaulted to Croatian!?!
I've reset it to United Kingdom, so here's the test:
-1
If there's no pointy brackets there, you'd better set your code page to Croatian...
what the Hell?!? when I check mine it seems to have defaulted to Croatian!?!
I've reset it to United Kingdom, so here's the test:
-1
If there's no pointy brackets there, you'd better set your code page to Croatian...
well **** me, no pointy brackets...
bizarre^2! I guess we'll all have to become Croatian just to use the internet
bizarre^2! I guess we'll all have to become Croatian just to use the internet
» More new posts: Doghead's Cosmic Bar