Video: BigDog quadruped robot

Depends on your definition of consciousness James. conscious in the sense of "he was knocked unconscious" or conscious in the sense of, awareness of self and surroundings. I read this introduction to consciousness by Susan Blackmore than convinced me that most of the time we're not even conscious, we're just existing and operating automatically most of the time, it's not until we stop and ask ourselves a question like "Am I conscious" that we are aware and conscious. I don't think creating memories is enough..

The blogger at Gizmodo pointed out that the natural human response is to feel bad when watching this robot slip on the ice or get kicked by his human controller. We naturally feel empathy for this machine's struggle even though we understand rationally that it has no "feelings". While it is slipping there is a flurry of CPU activity that kicks in to correct the destabilization, but can this be called "panic", or is it merely an automated response. Does it feel "relief" once it is stable and the CPU can slow down and return to normal walking routines? This machine obviously has an open feedback loop between reality, perception, and behavior that allows it to make real-time decisions and choose actions based on dynamic sensory input. Isn't this what consciousness is?

Keep in mind you can't prove that free will exists; creating a machine that thinks it has free will is enough. If you create something that so perfectly imitates our intelligence that you can't tell them apart, then they are effectively the same thing. (are we actually making decisions? Or are we merely glorified calculators?) We've created machines that can learn, true not in as complicated ways as humans do, but why does that mean we can't?

I have no doubt that you can make machines think, feel pain, become sentient, etc. However, why would you want to? An unfeeling automoton makes a better slave.

There's enough people that don't know how to think.

It's an interesting meditative exercise to ponder over anyway - How to make a circuit and program it to be able to make decisions itself, therefore give it free-will. Helps to realise what 'free-will' actually is.

Cause essentially when you make a choice or decision, you're drawing upon all things you know of / have learned (or at least, what you know and can actually access at that point in time - so different types of memory storage and retreival come into play), and also you're 'reading' the situation using very right-brain intuitive faculties,

the kinds that most of science hasn't really started to investigate properly

therefore when they talk about making an AI - it's more of a Dr. Frankenstein idea, that you emulate something that exists already but that you can't exactly explain everything about, and put electricity through it, and it comes alive..

it comes down to creating something whilst still not having answered various fundamental questions about your own species origins and purpose. Whatever you create will know that about you; it'll see all the flaws there and unless the maker is able to give a good explanation for why they seek to create without first having sorted themselves out!, the creation is going to lose a lot of respect for them there and then.
And that, is what the hair-trigger to going on a mad rampage of destruction comes from.

COngratulations to the constructors, but still I don't consider this robot life-like. I don't believe in artificial intelligence, I don't believe people will ever be able to create a machine that "really thinks".

Too bad the first thing DARPA is likely to do is put a machine gun on him...