Originally Posted by TheSquealer (Post 20739384)

I think there are some things to define
"what is artificial intelligence"... we confuse human capabilities for programmable capabilities. As people, we can't even agree on what "intelligence" is.

"human-like"/"human condition" - what does that mean? What i would say was i was a little interested in my own twinge of "thats wrong" when he was toying with the robot. I was interested because I knew with 100% certainty that i was projecting human like traits onto a piece of metal and plastic. First and foremost, that is what we do. We make it "appear human" in simple ways and it becomes perceived as human-like.

Think about this for a moment. They are still playing with walking and gate and balance. In many animals, this information is so rudimentary (relative to total capability), its not even processed in the brain. For example, in cats, you can sever the spinal cord and it can still trot on a treadmill because all the brains instructions for gate are stored in the spinal cord itself.

However, when it comes to anything remotely resembling human capabilities (specifically that of the brain), we won't see that in our lifetime. I wouldn't begin to know how to describe how insanely complex we are and our brains our. Everything that you are as a person is changing from moment to moment. Every experience you have from moment to moment, is changing your brain. A single event, even one that lasts less than a second (someone blowing their head off in front of you for example) can radically change who you are and subsequently who you will become over time.

So a robot can walk, fall over, stand up. Since we've been fucking with this for decades - its pretty clear how complex this is to program.

what about the millions of things the robot can't do which our brains do effortlessly and seamlessly and most importantly, instantaneously?

Take two simple things that make us human.

1) Imagination.

Can it predict the future? Can it easily imagine countless different future outcomes? Can it plan forward, modify plans based on new information and then evaluate the success or failure of a plan and come up with a better plan?

Can it predict countless future outcomes, plan for the most likely, experience the outcome and "learn" from that error so that success can be replicated or failure avoided? If we have struggled for many decades to program walking and balance and righting itself, how many more decades to get to this point?

2) Meaning.

One of the many things which separates us from our closest relatives and many members of this forum is "meaning". Meaning is huge us. Does the color black have meaning? Does the color white have meaning? Does Arabic letters have meaning? Does a rectangle have meaning? What about when you combine them into an ISIS flag or headband,... does the meaning change? Think about what everything around you "means"... literally everything around you has meaning from the colors of leaves to a banks sign to a gang banger standing on the corner vs the guy just walking by him... and think about how you use all of that information around you, quite effortlessly to make very advanced decisions as you navigate the world. How is that programmable?

What about identifying intentionality and empathy?

The robot couldn't decide if the person was good or bad. He couldn't identify the many facial cues, body language, heart rate/breathing and so on to make a determination as to whether this person is either helpful or likely not. Imagine the complex programming that would represent. It couldn't learn that the person was moving the box and come up with an advanced plan to stop him or wait. It could only keep attempting to execute the program "pick up box". It couldn't identify the intentions of the user (something a newborn CAN do), it couldn't identify the person as helpful or harmful (something a newborn CAN do), it couldn't identify the hockey stick as a possible weapon or imagine it being used to push it over, in fact, it couldn't even understand it was being pushed over and correct itself or even break its fall as those instruction sets didn't exist... it could only wait until it hit the ground, recognize its been tipped over and then execute the program "attempt to right yourself"

Personally, i would say there is nothing "human" about the robot other than the illusion we see and flawed assumptions our brains make. I don't think there will be anything "intelligent" about robots in our lifetimes relative to what the human brain is capable of, even in an infant.