8 replies to this topic

[Marios Kyriazis]

I saw the following qualities on an advertisement, and it made me think: Humans have these qualities, but can computers ever become complex enough in order to achieve:

Perfection
Doubt
Envy
Power
Truth
Awe
Irony
Belief
Compassion
Devotion
Fame
Reason
Pride
Brilliance
Vigilance

Clearly, some of these qualities have already been reached by today's computers, but what about the others? Even if one of the above cannot be mimicked by a computer, then what's all this Singularity fuss about?!

[niner]

I saw the following qualities on an advertisement, and it made me think: Humans have these qualities, but can computers ever become complex enough in order to achieve:

Perfection
Doubt
Envy
Power
Truth
Awe
Irony
Belief
Compassion
Devotion
Fame
Reason
Pride
Brilliance
Vigilance

Clearly, some of these qualities have already been reached by today's computers, but what about the others? Even if one of the above cannot be mimicked by a computer, then what's all this Singularity fuss about?!

You're forgetting Hatred, Covetousness, Shame, Disgust, Loathing, and, um... Worship of Idols. Of everything you mention, those that haven't already been achieved could certainly be coded up, should we be so inclined. I suppose to make some of them "real", we would need to add some hardware units. That's all that's going on in humans; we have a combination of hardware (e.g. Mirror Neurons) and programming. We seem to ascribe special "humanity" to those functions that come with hardware assist. (Consider the relationship between MDMA and empathy...)

What's all this anti-Singularity fuss about?

[Marios Kyriazis]

What's all this anti-Singularity fuss about?

Nice counter-sarcasm. Would a computer have thought of that?

Edited by mrszeta, 25 January 2011 - 09:35 AM.

[ihatesnow]

http://www.scienceda...10124184434.htm

[platypus]

The AIs would bew conscious personalities, right? What prevents them going insane faster than we can say "cat"? If they are many orders of magnitude faster than us weäve no way of detecting this in advance or even during the transition to insanity.

[corb]

Self aware machines will be first and foremost - logical. They will do whatever they can to execute their programming, their ... purpose. Their purpose of course would be a bit different then ours, we evolve through child birth - we are programmed to do whatever is in our strength to have offspring - most of the qualities you have listed originated as mechanism to assist that process. Doubt and Envy are caused by not knowing your limitations or refusing to believe you can overcome them (which would result in eliminating your "better", so you can become the alpha dog). Power and Truth are subjective they are only there to keep your self-confidence elevated. Awe and Irony are caused by a close mindedness and bias - believing you are always in the right. Belief and Compassion - self defense mechanisms, etc, etc.

Machines on the other hand would be able to evolve by just designing and changing their parts and acquiring more information about the world, they will not believe in limitations and will systematically test everything in search of ... better data and more complex more useful instrumentation. That would make their priorities a bit different than ours, although if we become never aging ... well we might become more like them, self improvement becoming our most important task in life, and that would call for a new set of virtues and qualities to be created.

Of course that is the ideal case in which they evolved from a learning program and there are no ethics or limitations hardwired into the machines to tamper with their natural growth. But we all know no one will create a thinking machine without SOME kind of safeguard and most of them will have their purpose core set up for sex slaves and house servants.

Hmmm. I just came up with an idea for a great sci fi novel ! An android's quest to become the BEST masturbation tool in the universe !

[hivemind]

- we are programmed to do whatever is in our strength to have offspring -

That's funny because I myself really do not want to have offspring.

[platypus]

Machines on the other hand would be able to evolve by just designing and changing their parts and acquiring more information about the world, they will not believe in limitations and will systematically test everything in search of ... better data and more complex more useful instrumentation.

What is keeping the machines that change themselves "sane"? What, if anything, is stopping them developing into cosmic Hitlers in a millisecond? Adhering to "logic" will not stop this from happening, not at all.

[corb]

What is keeping the machines that change themselves "sane"? What, if anything, is stopping them developing into cosmic Hitlers in a millisecond? Adhering to "logic" will not stop this from happening, not at all.

Now, now, I'm sure no one is going to build an android without the Three Laws hardcoded into it's "brain".
Also you should keep in mind they would still be our property so there is a good chance an extended collection of limitations would be used.

I was more or less pointing out that those qualities would be quite unnatural for an artificial intelligence.