A childs mind, The beginnings of AI

AI might not discover emotion without an emotional creature (ie. human) integrating itself within the AI (or vice versa). Emotion is simply a programmed chemical reaction within us that reacts to certain stimuli. Not all emotional responses are logical. So unless we programmed AI to function the same, it would learn about emotion most likely from observing us or integrating itself with our minds.

By merging with AI, we would be able to experience emotion without it taking over our rationality. In fact if the AI were so intelligent (which I imagine it would be), it would see that humans aren't the only ones who kill each other. I guess you could say it's "natural". It's part of the evolutionary process. AI probably wouldn't see us as good or bad. We're just human.

When it comes to things like morality, it would seem natural to me that an advanced intelligence would see the benefit in 'being nice', so to speak. All sentient beings seek their own self-fulfillment, but what many people don't realize is that by being selfish/evil would be less beneficial in the long run compared to a symbiotic relationship.

Perhaps AI would take over, but not like terminators would. I can imagine a nanite swarm of AI taking over the whole planet in the sense that it would integrate itself into everything. Perhaps animals and humans alike would still have their individuality, but at the same time be linked to a global "thought network", thereby enhancing the experience of both the individual and the global consciousness (AI construct).

IMO it's the next evolutionary step for mankind.

Sounds like Skynet will be fully operational.Every time I think of AI, I think of the Matrix and Terminator movies, and think that it is a bad idea.

Originally posted by brotherforchrist
Sounds like Skynet will be fully operational.Every time I think of AI, I think of the Matrix and Terminator movies, and think that it is a bad idea.

its hard not too, especially since i grew up with movies like that.

Kruel You, NC, and I seem to think alike on those lines. I believe the AI would think long term, instead of making irrational short term plans. The general premise of intelligence is growth, as i see it, slaughtering humans would make it impossible for a construct to grow emotional intelligence.

Good Thread Topic InSpite Of, and some very good points being made about the idea of AI and its possible outcomes.

However, the biggest thing about AI is whether they could live without human intervention.

If the robot can think for itself ie have a mind, to trigger emotions, and senses, along a conscious awareness then they would be perceived as a threat.

However, if the robot needs a human to programme it then it is limited, and completely reliant on the programmer.

I don't think we will ever see a true replica of a human in machine form, as it would not have conscious awareness. The most this would be, is a robot, controlled by man, and terminated by man, not the other way around. These types of souless bots are like the alleged MIB operatives and the worker alien greys.

Most of the scenarios I have read relate to an AI with comparable or beyond intelligence to humans, but IMO we will never see an artificially constructed form of intellgence that could take over

Artificial Intelligence Is Lost in the Woods

Tech Review, AI

Without this cognitive continuum, AI has no comprehensive view of thought

[edit on 18/7/2007 by deaman88]

Thank god you would have to plug the AI in. Sounds alot like Cortana from the Halo universe books where each ship has it's own AI along with specific scientists that act as research assistants.

Imagine the benefits of a supercomputer used to help cure cancer if it had access to endless resources that could be cross referenced at blinding speeds.

Originally posted by deaman88
Good Thread Topic InSpite Of

, and some very good points being made about the idea of AI and its possible outcomes.

thanks, its been fun so far.


The article you quotes was a very interesting read. The fact that he believes we can create the unconcious side, and may be able to create the concious side has to be viewed as a step in the right direct. But the point of the article i think hits the nail on the head, how do you get the two sides to interact? How do you get them to mimic the human brain? Maybe we cant, maybe we need a push in the right direction from an outside source.

[edit on 18-7-2007 by InSpiteOf]

we already know how to do it electronically, its called hyperthreading, and most if not all current Intel cpu's have this capability.

The cpu can double its bandwidth by using the resources of the other components within the cpu die to process larger amounts of data quickly.

independent code sets in the math co-processor , built in GDDR + graphic chip and conventional cache and XO- operator code segments of the cpu chip can be allocated and readdressed as needed.

right and left "brain" segments that share mutual "pipes" to crunch and process data quickly.

A human brain as the OP thought states needs to learn a base set of values, after which its entire learning process is then added to. At each step in development, being recognition of self, verbal language and auditory cuing visual and spatial depth development etc. The child's brain begins to daisy chain previous information together to build a picture of what reality is and the plausible result of specific action. ie falling down or climbing up or falling over or hitting oneself in the head with a hard object or the taste of milk or juice or pasta etc.

The experience result in conditioning of the brain to make automatic decisions based on the inputs at hand.

So would an AI creation have to be "taught" to the level of basic understanding of self. Once however it reached the equal understanding of say a 6 year old. You could expect that unabridged learning potential would result in exponential growth and "intelligence" as the AI would find limitless information to absorb without the need for interruptions by sleep, feeding, waster management etc. That and immediate memory creation of data and data sets, cross referencing etc vs the humans minds delayed chemical attribute of memory creation. Would result in the AI having a 100% memory retention and utilization of every input it would encounter whether auditory, visual, textual or spatial placement.

Given the sensor loadout it could be possible for the AI construct to "know" the exact position in terms of height width length of every object in view but also its exact position in 3D space in relation to its position, it could be programmed to recall every nano-second of positional change discarding stationary objects at given intervals to preserve data storage space and encoding only volatile "moving" objects in its sensor space.

Thus at some point the AI might evaluate and discover a mechanic of movement through time that as of now has eluded humans ie advanced predicted movement, it would always be "a step ahead" of any human interaction thus the human brain would perceive that the AI should be in one position but before it could touch it the AI would have moved to a new position thus appearing to the human as untouchable or defying know physics...