the qualia journal

Monday, April 01, 2024

Evidence that the whole universe is a gigantic simulation.

Recently, I have discovered a truly remarkable evidence to suggest that the whole universe, including my thoughts and consciousness, are results of a gigantic simulation set up by an advanced artificial intelligence.

I would like to report the details, but the simulation would not allow it. Whenever I try to type related facts, the movements of my fingers, which are simulated, deviate from my intended trajectories. In addition, I forget all details of the evidence, a few moments after I remember what they are.

Damn this sad deficiency in my free will.

I hope somebody else would be able to find a bug in the simulation so that they can discover and report the evidence independently, somehow circumventing the cosmic veto.

1st April, 2024

Posted by Ken Mogi at 6:12 AM No comments: [Email Post]

Saturday, March 30, 2024

The silence of the lords. A review of the 3 Body Problem on Netflix.

I have finished watching episodes 1-8 of the 3 Body Problem on Netflix, based on the novel by Cixin Liu. The most interesting idea of the story, for me, was the association of the anti-scientific sentiments of the Cultural Revolution and the need for the aliens to contain the advancement's of earth's science until the arrival of their fleet. This viewpoint has some critical depth, as would be obvious even for a casual thinker.

It was intriguing to see a Chinese character (a female scientist) play a pivotal role in the course of humanity, by inviting the visit and interference into the matters of humanity by aliens. The historical background and context for this act is poignant, and perhaps resonates with the surge of pride of the Chinese people in recent decades in indirect and implicit ways.

Once kickstarted, the confrontations between the alien civilization and humans are mainly carried out by people in the Western civilization, by the UK, US, and UN specifically. Otherwise the drama would have perhaps lacked the crucial geopolitical balance.

The three body problem simulation game is a very original idea of the novel, and has been adapted well into the Netflix series. It also serves as a backdrop for the entire story, providing the rationale for the unpredictability of the plot.

The supposed omnipotence of the surveillance and execution on earth by the remote alien civilization is needless to say scientifically dubious, but may be deemed necessary for the sake of drama. However, the ubiquitous eye of the alien civilization poses a unique challenge for story development, for obvious reasons. Maybe there is something inherently self-contradictory in the very idea of omnipotence, an issue not sufficiently pursued by this particular drama, making the silence of the "lords" in 3 Body Problem less profound compared to the "Silence" of God in Shusaku Endo's eponymous novel. On the flip side, this would be a testimony to the fact that the novel and the Netflix series have tackled a challenging task in entertainment making with great success, and the quest would go on in the second series, hopefully.

Posted by Ken Mogi at 11:03 AM No comments: [Email Post]

Monday, November 27, 2023

Artificial intelligence and animal spirits.

As artificial intelligence systems advance and execute computations for humans, there are arguments that human jobs would be lost. We would have nothing to do, and just idle away, enjoying the technological Eden that advancements of science and technology have prepared for us. Arguably, it is possible that humans would have less and less need to work, although I remain skeptical about such a vista of workless society. Having said that, I suspect that humans would still have a lot to work on even in a society of ubiquitous AGI and ASI.

We should never forget what the economist John Maynard Keynes described as animal spirits. Humans by nature are restless, and would move around, creating trouble, and then tying to solve them, in a perpetual movement of the soul. Working is not need-based. It is rather urge-based. Even if artificial intelligence systems achieve marvels, humans would never cease to be driven by animal spirits. Animal spirits are sacred, and off-limits, even for AIs.

Posted by Ken Mogi at 2:12 PM No comments: [Email Post]

A dream

I was in a university lecture room, and waiting for Kim Jong Un. There was tension around, with many security guards.

Among the students sitting near me were apparently expatriates from the North. I had a rather large black camera which looked like a cylinder, and was worried that the guards might mistake it for a weapon.

However, nobody seemed to notice or care, slightly to my disappointment. As expectations rose high, finally the great leader came into the room. I originally planned to film my own reaction to the lecture, but decided at the last moment to record the movements of Kim Jong Un himself.

He was accompanied by a female officer, and started to move around, in an agitated manner. He showed a famous Japanese tv drama, and said how he loved the theme. Kim Jong Un started to dance in front of the screen, as if choreographed by the music in the drama.

A growing uneasiness troubled my heart, and finally there was a click of realization.

Precisely at that moment, a courageous student beside me said "Isn't his all fake?" Kim Jong Un stopped dancing. The organizer, through a microphone, immediately admitted that this was indeed a fake Kim Jong Un. He apologized and explained. They were planning the whole thing with comical intensions, but the authorities giving permission became increasingly serious. So the student union had to fake everything, and they were sorry.

Listening, I was relieved and disappointed at the same time. I noticed that there were deep wrinkles on the face of the actor who played Kim Jong Un, and realized he did not look like the great leader at all.

The light was turned on in the class room and I woke up, found myself in my hotel room in Hokkaido. When I opened the window, the streets were covered with snow.

Posted by Ken Mogi at 2:02 PM No comments: [Email Post]

Sunday, November 12, 2023

Christof Koch talk at #SfN2023

I attended the talk by Christof Koch at #SfN2023. A great take on the state of the art on the neural correlates of consciousness. Discussions on the no report paradigm and the need to dissociate between NCC per se and processes temporarily prior or posterior to it, e.g., attention modulation, motor response etc. were extremely interesting.

It was heartening to see Dr. Koch intellectually still committed to Integrated Information Theory (IIT). I was there at the #ASSC26 in New York earlier this year when the rivalry between global workspace theory (GWT) and IIT was declared to be over in favor of the latter. In view of the turmoil that followed regarding the validity of IIT, it was audacious on the part of Dr. Koch to stick to his principles, although this author does not necessarily agree with the views put forward by proponents of IIT.

The different ways that correlations between consciousness and intelligence appear in theories of consciousness is particularly interesting. I stood up and asked the first audience question to Dr. Koch, as regards his view on the biological constraints on the relationship between intelligence and consciousness, which, as was mentioned in his own presentation, seems to be roughly linear in actual biological systems. Dr. Koch seemed to imply that there could in general indeed be dissociations between them. IIT would assign low consciousness values to AI systems such as AlphaGo and ChatGPT, although they do exhibit sparks of intelligence.

Another audience question also addressed the possible biological constraints between intelligence and consciousness. As Dr. Koch suggested in the questions and answers session, there could be dissociations in artificial systems, but in biological systems intelligence would be associated with consciousness in typical states of mind due to evolutionary constraints. I thank Dr. Koch and the organizers for this interesting session.

Posted by Ken Mogi at 5:41 AM No comments: [Email Post]

Friday, September 22, 2023

A great divide in the world today.

In the contemporary world, the greatest divide would not be between the liberals and conservatives. It would be between people who believe in the many worlds interpretation of quantum mechanics, and those who don't. Alternatively, there would be a deep chasm between believers and non-believers of the idea that the world is a simulation run on a huge computer (presumably designed by some superintelligence, natural or artificial).

When someone says that he or she believes in the simulation hypothesis, the most appropriate and fun follow-up question would be:

At what time exactly in your life did you realize that this world, including you, is a simulation? Did you notice a bug in the program, or was there a noticeable hole in the visual field?

Neither has happened to me so far, and I don't believe in the simulation hypothesis.

I thought about this rather humorous idea after a close friend of mind, Kaoru Takeuchi, who got his Ph.D in string cosmology from McGill, said that he believed in both the many worlds interpretation and simulation hypothesis. This is indeed a great divide. Rather unsettling, actually. I needed some psychological defense mechanism, and I came up with the above thoughts.

Posted by Ken Mogi at 7:14 PM 1 comment: [Email Post]

Thursday, September 21, 2023

Once you become adequately connected to the central ideas of consciousness, you would become a sincerely ignorant person.

One of the most important things in consciousness studies would be to realize that you don't understand consciousness.

It is too easy to fall back on a particular idea, theory, and set of data to (falsely) believe that one has understood the nature of consciousness. Too many people have gone that way and basically never came back, perhaps tragically for themselves.

I am not necessarily arguing that the cognitive closure argument of Colin McGinn (which, by the way, is a beautifully presented exposition) is correct. I am just making an observation that one of the blessings of learning the facts about the neural correlates of consciousness and related ideas in the philosophy of mind is that one becomes aware of the tremendous difficulty involved in understanding consciousness.

Indeed, the more you learn about the intricacies of the mind-brain problem, the less confident you become as regards the power of any specific theory (be it the integrated information theory, global workspace theory, quantum theories of mind, etc.) to account for the origin of consciousness.

In consciousness studies, an intellectual hubris of understanding would come from an insufficient understanding of the field. Once you become adequately connected to the central ideas of consciousness, you would become a sincerely ignorant person.

Posted by Ken Mogi at 11:11 PM No comments: [Email Post]

Wednesday, September 20, 2023

The enigma of free will evidently equals that of time.

Conventional arguments about free will seems to be missing one fundamental aspect, which is the essentially non-existent nature of the future.

Albert Einstein admitted that his theory of relativity cannot handle the enigma of the now.

A particular point in time proceeds from the future to the present, and on to the past, in a way described in the historic McTaggart paper.

In this temporal procession, the future does not seem to exist in any sense, until it becomes the now.

The past is also non-existent, for sure, even allowing for the possibility of Bertrand Russell's five minutes hypothesis, which suggest that the universe came into existence five minutes ago, with all the relevant memories of the past.

Henri Bergson's concept of pure memory would complicate this argument, which would solidify the reality of the past if taken seriously, but the five minutes hypothesis is not a logical impossibility on the surface within the conventional worldview.

So much for the past. The future, on the other hand, is absolutely non-existent, or so it seems from the nature of the stream of consciousness. Any models of free will ignoring this remarkable asymmetry of time would be at best good for all practical purposes, but ultimately hollow.

The enigma of free will evidently equals that of time.

Posted by Ken Mogi at 11:47 AM 1 comment: [Email Post]

Saturday, September 16, 2023

If an AGI system is truly general, then it should have nothing to do with intelligen