Is artificial intelligence theoretically possible?

I just had a philosopsychology lecture today that made me think – is it actually possible to create a machine that actually THINKS (not just pretends to) and has actual consciousness. In other terms, is it possible to create a Frankenstein that is capable of everything that another human being does?

My thoughts against this were that it had to be something naturally produced – created through the evolutionary cycle – that could actually possess conciousness. But now I look at it, my view of consciousness is just another structure in the human brain that sits next to attention and memory retrieval. If this is the case, surely anything that can run self-sufficiently by balancing attention, memory, choices etc (all human behaviour) MUST display consciousness as that is in essence what I have defined.

What are your views? Is it possible to create an artificial intelligence? I am of course ignoring all technical issues which deem it very extremely unlikely to actually create AI any way.