@Avant Brown Clearly, you do not know what a scientific theory is. If you are indeed getting a PHD, I highly recommend you get a refund.

scientific theory
noun
a coherent group of propositions formulated to explain a group of facts or phenomena in the natural world and repeatedly confirmed through experiment or observation:

Notice that it is repeatedly confirmed. Scientific theories are not fiction because they have been proven possible and have immense amounts of evidence supporting them.

Any other idea/model is a hypothesis, which may fall within the realm of fiction if they have not been proven possible.

Source: https://www.dictionary.com/browse/scientific-theory

I urge that you carfully read the response below, especially point 4, as I have carefully read your responses thus far.

++++++++++++
1. Aren't you aware of Artificial General Intelligence Research? Are you saying that Artificial General Intelligence, is not a valid scientific theory/model/area of active research?

You would be demonstrably false, because:

-----------------
2. You still continue to completely ignore that AGI is a quite real model in scientific endeavour, and not fiction.

++++++++++++
3. Why do you feel that the very real scientific model, namely AGI, is fiction? I.e. Why do you consistently presume scientific theory to be fiction?

-----------------
4. Crucially, the human brain, as far as science goes [Wikipedia/neuroscience], generates something called "general inteligence" in neuroscience, and this is observed in science to be substrate independent, i.e. not limited to flesh. [See Lessons for Artificial Human Intelligence and Substrate-Independent Minds.]

However, if one subscribes to some sort of religion, that stipulates the mind to be some special otherworldly, unscientific product, then it would be explainable why one would refuse to detect evidence regarding replicating general intelligence, a very real scientific quantity, in inorganic form, namely artificial general intelligence.

-----------------
++++++++++++
5. Max Tegmark is not a neuroscientist, but a phycist instead, who recently began to work on Machine learning. This is no surprise; machine learning benefits from applying physics in the learning equations.

You may want to check out Max Tegmark's simple explanation regarding what "substrate-independence" concerns:

Consciousness is a mathematical pattern: Max Tegmark at TEDxCambridge

First I want to say this topic is fascinating to me and I like to read and learn all I can on it. I am by no means an expert at this stuff and I could very easily get something wrong.

1. That is cool, did not know they were up to 2048. They must of made some major advances to stabilizing the qubit to make 2048 qubit possible, or perhaps they went more of the "brute force" route, where many, perhaps most of them spit out bad data, but the numbers are high enough to error correct. I unfortunately did not have time to read the article in its entirety I hope to do so tomorrow. If 2048 qubit every went mass scale, we would all have to update our encryption methods away from prime integers method. Which could end up requiring a lot more overhead for your non quantum computers to remain secure. As I do work in cyber security now, that is a big deal in my line of work.

2. Agreed, but if we are just changing definitions and setting new bars we are not really talking about moore's law anymore. My base point that actual targeted useful improvement by reducing transistor size to realize gains in space, heat issues etc is pretty much at an end. We can however modify, add to, and run stuff in parallel, like what GPU's are good at.

Take Nvidia's latest 20XX series of graphics cards released last summer. The 2080 (retails for at best 750 bucks (and this is months after launch!) but is can only play the latest games maybe 10-20 percent faster than the 1080 that was launched 2.5 years ago for ~$550. The price to performance ratio for all games has actually gone down! The only real gains in power is for games that utilize ray tracing to a great degree. Basically they just added on more hardware, more transistors to do new things, not a whole lot of "improving upon" what it could do before w/o throwing more memory at it. Now at least part of it is Nvidia cashing in on the cryptocurrency craze which allows them to fatten their profit margins, but overall even for nvidia the "inventor of Hyper moore's law" is having real trouble seeing advancement at a practical usable level. It is a bit like a car manufacture saying: This car is much faster and better then our last model, we added turbo instead of doubling horsepower, and that is fine, but the car also cost a signficant amount more, weighs more etc with the inclusion of turbo.

3. Fully agree with you here. It is pretty cool reading about how they taught the computer how to play GO well enough to beat the top GO players in the world, by having it play many thousands of games and learning based on the outcome of them. Instead of trying to brute force calculate every possible move, as the developers were well aware that the possibilities quickly ballooned to unmanageable numbers even with with a computer that had performance measured in tera flops.

3C. Disagree with you here. That is quite the unevidenced leap, there are many many other possibilities here other then "human intuition going on." We humans still do not even fully understand human intuition.

Most of your arguments demonstrably occur on ignorance.

I never claimed to be an expert and stated these were my own opinions I even learned some new stuff, like the dwave simulation that happened in late august from your post. But how sure are you that there is no gaps in your knowledge? Enough to be certain "most of my arguments demonstrably occur on ignorance?"

It is nice that you learned something, we should probably all aim to learn something, that's a part of how Science advances.

1. When you mention that NVIDIA "just added more hardware", be careful to not underestimate the miles of research that may be required to produce incremental changes. [I am a gamer as well, I use gtx GPU cards to do things other than machine learning. If only I could afford the more efficient TPUs.]

2. I didn't say human intuition. I said human-like intuition. Alpha-go is not brute forcing all possibilities as you seemed to insist before when you referred to "blazing" speeds, for the reason I priorly stated, and crucially, we already know that ai is loosely inspired by biological brain models. This is quite literally why I used the phrase human-like, rather than "human".

3. When I said most of your arguments are demonstrably based on ignorance, it was primarily because they were based on ignorance; and as you admitted, you didn't know about some of the items I had mentioned. Now, this doesn't mean I am without ignorance; notably I am very ignorant when it comes to many topics outside of my immediate field (of ai), although the things I work with mostly (artificial neural networks), with more advancements , will reasonably contribute to universal problem solvers, that can approximate most if not all cognitive tasks or fields of human endeavour.