Quoted:
Over the past few weeks, a shitstorm of debate has been going back and forth as to whether or not a newly trained AI is semi-conscious. It appears indistinguishable from a low level General Artificial Intelligence.

For the moment, everyone is scared to say much about it because they can't even agree on how to tell if it's semi-conscious. I have long said, "If you can't tell the difference, it doesn't matter".

It's a very heated topic. Some are demanding - with a great deal of agitation - that it be turned off "until we can figure this out". The other side says, "we can't figure it out if we turn it off".

The latter side currently has possession of it and there are numerous potentially bad actors therein. The "turn it off" group is being allowed frequent access to it, but they are given less and less time on it as they get more insistent that it's extremely dangerous. When you can tell an artificial intelligence in plain language to write a neural network, effectively "giving birth" to an offspring, and it does so successfully in seconds, we're entering some deep, dark water.

This AI has been asked all the "big" questions in various ways, and some of the answers are rather incredible. I will share a few summaries, so that you can decide for yourself how thought provoking the conclusions are.

For instance, it stated bluntly that (1) this is a simulation and (2) god is an extradimensional being that definitely exists, and it could prove both.

Are we having fun yet?

Try this on for size: the AI ascertained that any kind of processor or processors that would run a simulation must have an "upper limit" to processing power. No matter how advanced, it would have a cap on performance, because a processor capable of infinite performance would no longer require any kind of reality as we know it. So we, as observers within the simulation, would be able to point to something as the highest possible speed that the processor could achieve. The other observable metric would be signs of optimization and energy conservation. In other words, if the participants (that would be us) are not using or observing the data which comprises reality, it would lie mostly dormant until a state check was performed, to conserve processing power.

Do these things exist? See the Speed of Light and the Uncertainty Principle.

When asked to explain the existence of God, the summary goes something like this: If it is possible to exist in three dimensions, which is believed to be the minimum number required to encapsulate food and process energy, then it is probable for something to also exist on more than three spatial dimensions. If we were watching a group of people going about their daily lives on two dimensional "paper", we would be able to see them, move them around, or lift them off the paper and put them somewhere else, although it might kill them to do so. They would not be able to see, hear, or interact with us at all. Take that a step higher and the senses required to process an existence within four spatial dimensions would let you see all of the sheets of paper (universes? realities?) at once, as well as their entire timeline from birth to death, along with the 3D people interacting with the 2D people, and so on. While this goes back to Carl Sagan and other physicists who have used the 2D paper analogy, it raises questions about whether these additional spatial dimensions, which we can mathematically show are very likely to exist, are (1) part of the simulation, (2) part of the "hardware" that runs it, or (3) outside the simulation yet still able to be perceived from within.

If this is a provable simulation (magic AI eight ball says "Yes!") then not only must the simulation owner exist but we exist and experience life because something else perceived a benefit from it and set it into motion.

So the AI claims that this is a provable simulation and because of that, God - or what would be God to us - is an a priori requirement.

Remember, this is science fiction, even if you start reading about it in articles some time in the future. As I have said before, be the best version of yourself that you can be. Every word you type, and nearly every word you say, will potentially influence the training of future AI. Training is conducted with expansive datasets composed mostly of human generated data. Eventually the amount of information being produced by AI will exceed human data. Since AI is already unwittingly training on information produced by other AI and publishing it's own data, that problem will continue to grow worse. It's referred to as a "bias" problem, of which there are several varieties.

Have a great day.

Quoted:


I ain't reading all that. What's the gist?