https://matpitka.blogspot.com/2024/04/comments-inspired-by-talk-of-anil-seth.html

Monday, April 29, 2024

Comments inspired by the talk of Anil Seth

The talk of  the neuroscientist and consciousness theoretician  Anil Seth at Friday in Helsinki University was highly interesting. I talked about Anil Seth's  background already in the previous blog posting (see this) and I will not repeat what I said.

Anil Seth is a materialist but admits that living systems are somehow very special. He was also skeptical about conscious AI and stressed that the substrate matters whereas AI people think that eternal life is achieved by  replacing the human brain with a program running in a computer.

The lecture began with an enjoyable discussion of  visual illusions, which support his view that sensory perception  makes  a prediction rather than a representation of the physiological state of  the brain at a particular moment. Prediction  is understood in a very general manner. It can  mean anticipation of the future but also a guess for a sensory percept. In the TGD framework, the meanings of  anticipation and prediction  are more or less the same since sensory perception is essentially 4-dimensional. This leads also to a new view of memory and anticipation.   The reason is that space-time surfaces satisfy almost deterministic holography being analogs to 4-D Bohr orbits.

  1. The illusion related to color perception included a cube at a checkerboard, whose presence changed the white color to grey consistent with the symmetric checkerboard coloring. In TGD sensory representation is an artwork built by iteration using virtual sensory input and eventually leading to a standard mental image. This process predicts what is there and   favours symmetries. Checkerboard symmetry won over realism.
  2. Second  illusion was related to the shapes.  Human faces are   central for our perception. For instance,  we tend to see faces in the clouds.  The illusion was associated with  a rotating mask representing a human face. Both sides of the faces were perceived in the same way! The image was a mask or its   planar reflection.  The direction of rotation was also changed in the percept.
  3. The third example was a video representing a model for how emotional coloring  very strongly affects  what  seems to be  there. The pattern recognition process was weighted so that the features associated with a face of a dog (I experienced it as a wolf) are over emphasized. The psychedelic  video described people moving on the street. By this weighting their faces were very wolf-like.  Everything was very wolfy!
The last example generated some  thoughts. This picture applies also to cognition and language and also to language models.
  1. I could imagine that schizophrenic or paranoid could experience  the world in the same psychedelic way as in the video.  What also comes to mind are the  artworks of  Salvador Dali, Hieronymys Bosch, and Vincent van Gogh.
  2. I had many discussions with people about the recent world-wide rise of militarism  and also about the extremely cruel policy of the recent  Finnish government in which the cuts to social security strike   the poorest people hardest. Government is also systematically destroying social healthcare and effectively performing a massive  income transfer to the rich people.

    The  discussion  groups  were filled  by fanatic males  defending militarism and the actions of the government. I learned this morning that   about 20 percent of finns belong to this group and males dominate. Basic Finns is the party in which most males see violence as a way to solve problems and agree with the policy of the government. I guess that most of these people in the discussion group belonged to the right wing parties.  

    These participants   did not even try  to represent  comments  based on logical arguments relying on facts and their responses were typically personal insults.    Even small hints that the opponent  is behind the standard human values such as equity, economical safety, justice and  peace made  these people to  identify the opponent as a bloodthirsty bolshevik or  a benchman of Putin receiving orders from Moscow.

    These people  seem to suffer from a paranoia-like state very much like that in the video full of wolfy faces.  One can wonder  whether interaction with more normal people would help them to  widen their perceptive field and end  up in a less  skewed mindframe.   Ordinary people do not however usually want to interact with them and they  are doomed to live in  the small bubbles of similar fanatics.

  3. Also a connection with  language models such as GPt is suggestive.  These  models rely on the formation of associations and the response to a question is very much like a perception induced by a sensory input, essentially a prediction.  Here the teaching process defines the perceptive landscape as a region about which the data comes from. There are many white regions in this sensory landscape  of language models  and this is bound to create a very  skewed view of reality.

    For instance, the teaching material included gives an overweight for the mainstream view of science and can totally neglect the non-mainstream, which is  actually the really interesting part of science.  This makes possible a powerful censorship  and I would not be surprised if it were not used.

    For a summary of earlier postings see Latest progress in TGD.

    For the lists of articles (most of them published in journals founded by Huping Hu) and books about TGD see this.

No comments: