What do consciousness, AI, and philosophy have in common?

All three are domains I've dedicated a substantial proportion of my life to studying. Consciousness, like philosophy, and like AI, is a distributed system. The state space in which consciousness operates is a space full of the possible states that the body can take. For philosophy, the space is different because it isn't limited to one body. The state space of philosophy includes all philosophical thoughts that can possibly be had by all bodies capable of thinking philosophical thoughts. And, for AI, the state space is even broader, essentially consisting of the portion of linguistic thought that's written down or saved on the internet plus coding languages plus some crazy math stuff that may not be fully understood for awhile; but in a sense it's shallower because it does not include any body stuff at all and instead is totally abstract, can be copied, powered off, powered on.

All of that is to say everything is related. Consciousness permeates our bodies, scaling from the level of the single cell to the highest order thought the body can think. Philosophy can be touched by consciousness, and vice versa, but LLMs and other contemporary forms of AI prove useful by operating in an entirely different way, as simulation machines. I've borrowed the term from Karl Friston, who uses it to denote the sense in which an LLM is not acting as a speaker when it responds to a prompt, but rather as a simulator of what a speaker with knowledge set xyz might say, were they to respond to this prompt.

The purpose of philosophy is to enable the student of philosophy to abstract the content away from the deep structure behind our most effective rational thinking. The idea, of course, is that doing this will allow us the opportunity to understand this deep structure so that we may then replicate it.

Sometimes, the philosophical paradigm of an age can suddenly give way. A revolution a la Einstein can come at any moment, setting the best-known scientific truths aside in an instant as everything changes again.

Studying philosophy of mind leads one to a place that is full of conundrums these days. On one hand, empirical models exist that can predict a shocking proportion of mental phenomena. On the other, there has never been a way to truly model the interiority of a mind in general; we only have access to our own interiority. The Markov blanket does an excellent job representing this relationship between different individuals, with its external and internal states.

It has been many years since I was first filled with wonder at the puzzling and difficult subject they called cognitive neuroscience. Here, at epicdylan.eth, I'm bringing my scientific work and my philosophy and my technology all together for the first time. It took awhile to build the clarity to make this happen, but the perspective has arrived. As of today, I'll be writing here regularly. You can find more information about me and my work at https://epicdylan.com for now, and if you have questions or ideas for articles feel free to run them by me on Farcaster: https://farcaster.xyz/epicdylan

I'm excited to finally make the first post to this blog. I've bounced around for a number of years, but something tells me that paragraph.com is going to be a good long-term home for my content -- and that epicdylan is the moniker for it. Only my old gamertag is weird enough to represent my diverse interests in science and technology.

Go ahead and follow the blog here to keep up to date with the various things I'm working on. You'll get write-ups as they're available for my open source projects such as Evermark, and you'll be the first to know when my more official projects find publication homes.



Thanks for reading,

-D

What do consciousness, AI, and philosophy have in common?

All three are domains I've dedicated a substantial proportion of my life to studying. Consciousness, like philosophy, and like AI, is a distributed system. The state space in which consciousness operates is a space full of the possible states that the body can take. For philosophy, the space is different because it isn't limited to one body. The state space of philosophy includes all philosophical thoughts that can possibly be had by all bodies capable of thinking philosophical thoughts. And, for AI, the state space is even broader, essentially consisting of the portion of linguistic thought that's written down or saved on the internet plus coding languages plus some crazy math stuff that may not be fully understood for awhile; but in a sense it's shallower because it does not include any body stuff at all and instead is totally abstract, can be copied, powered off, powered on.

All of that is to say everything is related. Consciousness permeates our bodies, scaling from the level of the single cell to the highest order thought the body can think. Philosophy can be touched by consciousness, and vice versa, but LLMs and other contemporary forms of AI prove useful by operating in an entirely different way, as simulation machines. I've borrowed the term from Karl Friston, who uses it to denote the sense in which an LLM is not acting as a speaker when it responds to a prompt, but rather as a simulator of what a speaker with knowledge set xyz might say, were they to respond to this prompt.

The purpose of philosophy is to enable the student of philosophy to abstract the content away from the deep structure behind our most effective rational thinking. The idea, of course, is that doing this will allow us the opportunity to understand this deep structure so that we may then replicate it.

Sometimes, the philosophical paradigm of an age can suddenly give way. A revolution a la Einstein can come at any moment, setting the best-known scientific truths aside in an instant as everything changes again.

Studying philosophy of mind leads one to a place that is full of conundrums these days. On one hand, empirical models exist that can predict a shocking proportion of mental phenomena. On the other, there has never been a way to truly model the interiority of a mind in general; we only have access to our own interiority. The Markov blanket does an excellent job representing this relationship between different individuals, with its external and internal states.

It has been many years since I was first filled with wonder at the puzzling and difficult subject they called cognitive neuroscience. Here, at epicdylan.eth, I'm bringing my scientific work and my philosophy and my technology all together for the first time. It took awhile to build the clarity to make this happen, but the perspective has arrived. As of today, I'll be writing here regularly. You can find more information about me and my work at https://epicdylan.com for now, and if you have questions or ideas for articles feel free to run them by me on Farcaster: https://farcaster.xyz/epicdylan

I'm excited to finally make the first post to this blog. I've bounced around for a number of years, but something tells me that paragraph.com is going to be a good long-term home for my content -- and that epicdylan is the moniker for it. Only my old gamertag is weird enough to represent my diverse interests in science and technology.

Go ahead and follow the blog here to keep up to date with the various things I'm working on. You'll get write-ups as they're available for my open source projects such as Evermark, and you'll be the first to know when my more official projects find publication homes.



Thanks for reading,

-D