Does Humanity Even Want a Machine God? - Andrés Gómez Emilsson, DemystifySci #408

Shiloh’s ethical concerns are genuinely great/valuable.

so consciousness is a life form in of it self then

Listen on the go at all podcast locations: https://anchor.fm/demystifysci Pick up our music on vinyl or stream on bandcamp, spotify, etc.. https://demystifysci-shop.fourthwall.com/products/vinyl-lp-secretary-of-nature-everything-is-so-good-here Paradox Lost is now on pre-sale! https://buy.stripe.com/7sY7sKdoN5d29eUdYddEs0b

Fascinating. Seems this second chat dealt more with the philosophical underpinnings of the AI than episode one. Both were great fun. This discussion seems to deal more with just what happens when the data volume, processing speed and the purposeful directionality of the created machine adds up to a self reflexive entity. Great question. Wouldn’t the newly created “individual” have the same internalized, uncomfortable thoughts such as anxiety, discomfort, and worry, as we humans, and how would we deal with this newly created, powerful patient? And concerning the question of slavery, what internal discomfitures would have to be mollified in dealing with a potential for slavish awareness of the machine itself? How does one keep rebellion from initiating? A second question enters here. That of consciousness. As an appreciator of Julian Jaynes and his concept of consciousness, it would seem that a reversal of its development, as he saw it, might be in play here as we move to create more “intelligent” machines. Could we be moving backward from a linguistic, metaphorically linked basis of intelligent thought, to a more interconnected approach in the way paleo man resolved problems, bicamerally. Is AI moving from a left hemisphere, analytical process, which Jaynes called consciousness, to a more holistic right hemisphere one, where isolated data points are indiscernible? Would effective thought be displaced by an issuance of ephemeral commands which would have to be obeyed for our own good? Would we come to worship, as in some science fiction thrillers, the machine, while loosing our ability to decide consciously ourselves? More questions than answers for me. But very thought provoking, as are all your podcasts. Loved listening to you guys bat things around. Most helpful! Thanks guys.

Yeah, humans perceive “wholes” before details & naturally understand what is relevant in a given situation from context & implicit knowledge. Machines must build up meaning from atomic facts/data and can seriously fail to distinguish significance. The belief that AGI can happen is built on assumptions: biological assumption that the brain is a device for information processing; psychological assumption that the mind is a device filtering information according to formal rules; epistemological assumption that all knowledge can be formalized into logical, discrete rules; ontological assumption that the world consists of independent atomic information/facts.

When I was a kid I designed an 8-bit computer whose state transitions were a function of its global operation. It was very simple. All I did was use memory-mapped I/O registers to control address-decoding logic. At the simplest level this was just a memory management unit where logical addresses mapped to physical memory addresses via whatever the address-decoding logic determined. But then I realised that the memory addresses at which the address-decoding logic registers appeared could themselves be controlled by the address-decoding logic. I didn't build it because the immediately obvious problem was how you could reliably program it. If there were any error in writing the registers then the machine state would have departed from the small subset of states which had deterministic behaviour. But in principle systems like this would have the property that Andrés talks about: their microstate is determined in part by their macrostate. I had never thought of this in those terms before. Thank you!!

Interesting.

59:18 On evolutionary goals and artificial goals: The anthropologist Gregory Bateson used to talk about evolution as a self-healing tautology. On its own, the idea of natural selection is a tautology: that which survives longest survives longer than that which does not survive so long. He quite explicitly used ideas from the philosophy of mathematics, in particular Russell and Whitehead's theory of types that appears in _Principia Mathematica._ His book _Mind and Nature_ is a good place to start.

The easiest way to phrase it; A boundary with an internal frame of reference to that boundary. As Kastrop suggests, that internal self only has access to the physical world through that boundary. All interactions to other boundaries are filtered through that boundary. That internal self cannot be externally observed save through that filtered boundary.

I think Anastasia is correct in questioning whether the goal of a bacterium seeking food can be separated as a state independent of its embodiment in the whole organism.