What it feels to be a microprocessor?

The following ponderings were inspired by discussions in our Zoom group (Marko Manninen, Ville-Einari Saari, Rode Majakka and Tuomas Sorakivi). The topic of discussion was whether one could assign to language models what might be called symbolic consciousness and whether the avatars creatable using language models are real conscious entities.

Personally, I would consider the language models to be what their builders intended them to be. Context-recognizing association machines, perfect students, which do not perform logical thinking or generate new ideas. Computers might develop consciousness, if they become genuine meso-micro or macroscopic, or even longer-scale quantum systems, for which statistical determinism no longer applies. However, I would guess that the level of consciousness in recent computers is far from that assignable to humans.

This can be considered more concretely. A microprocessor is the operational center for a PC.

1. An interesting observation is that the upper limit for the length of the wafer containing the microprocessor is .5 cm, which is the gravitational Compton wavelength for any system regardless of mass in the classical gravitational field of the Earth and equal to half the Schwartschild radius of the Earth. Is this a coincidence?

I’ve been wondering if quantum gravity in the sense of TGD could make computers conscious beings within a zero-energy ontology. The gravitational magnetic bodies of the Sun and the Earth are natural candidates. For the Sun, the gravitational Compton frequency is 50 Hz, the average frequency of the EEG. This is another thought provoking observation.

Interestingly, for instance in Chinese there are no letters as symbols, only words: a kind of holism and no reduction to the level of letters. Also in the TGD based quantum realizations of the genetic code only codons are realized whereas chemical realization decomposes them to letters.

To conclude, I find it really difficult to see how higher level consciousness could be realized at the processor level. Rather, the output of the language models produce the contents of consciousness in our brain (or its field body) as associations, and a huge amount of information from completely different levels determines the generated mental images. The amount of information generated in us by the text is enormously greater than the text itself, and its amount depends on the recipient. The power of language models is that they manage to generate the important bits managing to generate sensible mental images. Just like shouting the name of the dog generates a lot of sensible activity.

This argument did not consider the possibility that the user and the microprocessor could entangle to form a larger conscious entity and make it possible for the user to affect the behaviour of the computer so that it would not be a deterministic machine anymore. There is an experiment in which a chicken imprinted on a robot seemed to be able to affect the behavior of the robot. If true it might be understood in terms of entanglement with the random number generator determining the behavior of the robot. See Are Conscious Computers Possible in TGD Universe? and Space-time surfaces as numbers, Turing and Gödel, and mathematical consciousness

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.