I ponder a couple of things in regard to creating artificial replicas of the brain.

1) There are pattern (or object) recognition studies where the brain waves of responses are characterized first during a "training" process. It then becomes possible to know what the response of a person is just by looking at the brain wave information. Such studies have reported finding that the brain wave pattern will manifest milliseconds prior to the appearance of the image to the subject's visual cortex. IOW, human consciousness, as it interfaces with its avatar matter/energy brain, can appear to violate causality.

In working toward the goal of simulating the human brain, how is it proposed to be able to simulate this characteristic? (We really don't fully understand overall human consciousness interaction with the matter/energy brain to know in detail where causality gets mucked with. It may be a subtle, but constant aspect of our overall cognition/self-awareness.)

2) Will an artificial intelligence need to be designed with a capability to dream? It seems that lifeforms with higher order brain complexity (mamals) dream. This would tend to indicate some sort of biological necessity. How would dreaming even be simulated? For biological intelligence it appears the DMT molecule inhibits some sort of perception barrier to where our consciousness interacts perhaps with what is its "higher self".

I've had dream that was not of the variety of just working out some sort of anxiety related to waking life, but instead had a narrative and the use of symbolism on a repeated basis such that a very clear message was being communicated. Once I wrote down the details of the dream right after waking up (it was during the daytime when I was at home ill with a cold/sinus infection), I was then able to go back through and with a few passes construct the information that was being conveyed. Over the space of a few days I was able to keep going back and decode pretty much every detail of the dream into a consistent and coherent overall message.

There were some things that were kind of unique about this dream in that usually names of dream characters are not explicit. In this case the principal character (other than myself) had a very specific name and in one situation of the dream I literally spelled out that person's name. So when I woke up I remembered that name and its spelling (was slightly quirky compared to usual spelling).

At any rate, I tend to suspect that an artificial intelligence will never dream like a biological intelligence does - what higher self of consciousness would it interface with?

Of course that raises yet another question...could a soul consciousness incarnate into an artificial intelligence? In the end, silicon circuits or carbon based molecules are all just matter/energy. Could the silicon circuit neurons make for a suitable avatar for soul consciousness? (I haven't gone into DNA interactions with consciousness, which may have some bearing on the suitability of avatar host. Am alluding to interesting studies of DNA exhibiting quantum entanglement like phenomena, etc.)

Well, it's certainly safe to say that these various teams that look to be modeling the human mind purely from the basis of its matter/energy composition of neurons and chemistry, are not going to acknowledge studies that indicate characteristics of brain/mind operation such as violations of causality. The problem of willful skepticism is that it doesn't really banish the anomalies to non-existence, but it does insure said skeptics continue operation on a premise of a fallacious model.

It's not that there won't be value coming out of improved silicon neurons - could conceivably indeed improve machine vision and hence image recognition systems, etc. Is just that the transhumanist crowd, such as Ray Kurzweil are pinning their hopes of achieving immortality on eventually downloading their consciousness into an artificial brain. If the operative matter/energy model of the mind is all wrong, though, then the download may end up being not actually, say, self-aware. Plus these studies indicate that there are some day to day things our neural control system does in respect to operating the body that are seemingly dependent upon the causality violation. So the question there is if this was a design necessity.

Dr. Pim Van Lommel has a book where he's written about a study he's conducted into NDE. His is one of the better studies because it ran over several years and filtered to cases that were documented in hospitals that participated in the study (i.e., the circumstances of death and resuscitation were documented in hospital records). Dr. Van Lommel in his book explains why, on physiological grounds, these experiences were not the product of any neurological activity.

Interviews with Pim Van Lommel

Obviously the notion raised in the study he documents in his book is that self-aware consciousness persist even when there is no operable brain any more. That's pretty much a working definition of the concept of a soul.

EKUMA1981, you might be interested to also check out Pim Van Lommel's book "Consciousness Beyond Life: The Science of the Near-Death Experience." 20 years of rigorous scientific research that leads to the conclusion that consciousness is separate from the brain.

OBE research also shows that the consciousness can travel, which indicates that it either can't be strictly a function of the brain, or that there is an unknown communication mechanism with an Akashic Record-like database.

Past life regression research (see Brian Weiss, for example) also seems to point to the conclusion that consciousness is separate from processes in the brain.

In any case, I see nothing wrong with the idea of a consciousness occupying a sufficiently advanced AI, but I don't think it is an emergent property of the AI (based on the above research). The definition of "sufficiently advanced" is certainly TBD. And how might it make the jump?