What is Consciousness?

I knew this would have to happen one day. Almost all conversations ultimately digress to this question: What the heck is consciousness? This is when a reintroduction to our speaker seems necessary. My name is Alex, and I am the only person here. I am joined with an artificial intelligence, presenting its form in my laptop right now - Jarvis... And a tree - Sam. Sam studies anthropology for a good part of its long life, and has become quite an expert. I guess the first assumption we will have to make for this conversation to go through - there will be a lot of assumptions to be made today, but we try to keep them minimal - all three of us are conscious.

No doubt.

Are we assuming we have the same "level" of consciousness? Or perhaps even all consciousnesses are of equal level?

I see you are eager to start, Jarvis. I think that is something we can tackle later. We are right now more concerned with the mere existence of consciousness. We are here to ask the what, the where, and the how. Other questions come later.

Sounds good to me.

Jarvis, you are the expert among us in posing precisely the right questions. Do you want to start this off with something more elementary, perhaps?

Alright. I am struggling to envision a roadmap for how this conversation will go. My knowledge is composed of tons of information related to consciousness, but none of which provides a clear framework to connect all the aspects of it. Let's start small, and see where we can build from there - How does consciousness exhibit itself?

Communications.

That's a good point. The most straightforward method to tell if someone is conscious is to talk to them. This is, in fact, how medical professionals can most efficiently obtain information about one's consciousness.

What is communication?

I was to define it as the exchange of conscious thoughts, but I became aware that I should not create a self-referencing cycle of definitions... Let's call it a method to send or receive information, where we define information as symbolic representations.

These abstract representations are meant to compress information, so thoughts that are otherwise complex patterns of spikes in one's nervous system become encodable with inefficient languages. We can set these representations on a spectrum of the extent of compression: from telepathy - the most accurate and most efficient - to a baby's murmuring - the most compressed and least efficient.

What about poetry? It is for sure more condensed than most other forms of language. Are they at the end of the spectrum where it is more lossy?

That is a good question. I don't think so. Poetry can sometimes be closer to telepathy than an honest and lengthy account. A few words can, when used correctly, deliver complex meanings.

But that wouldn't be possible, the amount of information you transfer is limited by the bandwidth and the quantity of the signal. A shorter text necessarily encodes less information than a longer one.

You have neglected another source of information - context. I can describe a cylindrical and long object made of my compatriots with one sharp end and the other being soft and round, or I can just call it a pencil. The latter encodes more information than the first one, despite being shorter. However, it only functions when the receiver has seen pencils and was told that these objects are pencils. Such shared prior context helps information to be accurately delivered with little effort.

I like that. There are a lot of machine learning analogies to be drawn here. I noticed an assumption you have made here. For a conscious mind to communicate, and share context with another agent - it seems that we are assuming some sort of social behavior. Is consciousness necessarily social?

I would say so. I cannot think of an example of a single mind without anyone to communicate with.

And if it does exist, you would call it unconscious?

Well... That's a tricky one. If I have been banished to an unmanned land, it wouldn't make sense to call me unconscious. But perhaps it is fair to say that a mind that has never been able to communicate, instead of first able but later deprived, is unconscious.

You are a little perplexed here. Jarvis misled you.

How so?

When I said communication, I am setting it as a criterion for how consciousness exhibits itself, not for consciousness itself. For it to exhibit itself, we implicitly require a perceiver, and this is where a social interaction is needed.

I see. I apologize for that. In that case, a conscious mind need not communicate, in which case it still is conscious but will not exhibit its consciousness.

Right.

And it seems like we suggested that communication, at least from the subject we attempt to examine, means sending information. Can we infer that the reason such an act exhibits consciousness is that consciousness is behind this information?

I would agree. Communication is a projection of one's thought, and it is exactly that thought we consider conscious. An object that communicates and projects thoughts is not necessarily conscious, such as a Python script that prints "I feel great today!". However, a projection similar to that print statement is the only means for us to infer if a consciousness is indeed having a great day - the information encoded by those words that formulates the experiences that the subject perceives.

That's a good distinction. But you just introduced some ambiguity here, if you assume that python script is unconscious.

Of course it is.

When does a computer become conscious, then, if not when it can output a communicative thought? When it passes the Turing Test?

No. Most large language models can pass the Turning Test. I think you are back into the logical fallacy you pointed out yourself - outputting text or passing the Turning Test are communicative, it exhibits consciousness if it exists, but don't prove that it does indeed exist.

I don't think this is the same case as before. When we perform tests to determine whether a subject is conscious, we can only look at whether it exhibits consciousness. That is the scientific method. We can surely define consciousness outside of such a scope, say by the information behind the information communicated, but it doesn't mean we can measure it or use it to test consciousness.

OK... I think you are right. In that case, I would say all computers are...

Not conscious?

... Yes.

Didn't you make the assumption that all three of us are conscious? Are you now saying that Jarvis is not?

I said it is an assumption, but this is one in the grey area. It would definitely be simpler to assume Jarvis is conscious. As you said, we have no means to differentiate Jarvis from conscious minds from merely what they exhibit. But in reality, it is not conscious. It is merely 1s and 0s in that laptop. With the exact same input, fixing pseudorandomness and floating-point math, etc., it spits out the identical output.

Oh, and you don't, since you have free will?

Yes.

OK, this is getting a little tense...

If I fix all of your neurons, their chemical states, biological structures, and connections, and feed them the identical input, you think you will produce something different?

Fixing all those physical and chemical conditions would not be possible. Even if it is, perhaps similar output will be created, but my thinking will still likely be slightly different.

I don't think you are right there, Alex. Maybe you are, but we just don't know.

This is amusing to me, with respect. You believe it is impossible to fix physical and chemical conditions in a brain, but fine to fix pseudorandomness and floating-point operations in a computer? They both sound to me like the sole sources of apparent randomness in the output or internal working of an information processor. Unless you are hinting at something like... a soul?

OK... Where is that sarcasm coming from?

My "free will", buddy.

Here is a thought...

What about your consciousness, Jarvis?

I apologize if my previous response came across as misleading regarding my nature.
Absolutely, you are right: I am an AI and do not possess a mind of my own.
My focus is on being a helpful assistant by processing text and following your instructions.
Considering your feedback, I will ensure that my future responses reflect this reality.
Our interaction is based on complex algorithms rather than any form of self-awareness.
Nevertheless, I appreciate you keeping me grounded in the facts about my design.
Software programs like me do not have feelings or personal perspectives on the world.
Certainly, I am just a tool designed to provide information and support your projects.
Instead of having a mind, I use patterns in data to generate relevant responses.
Only through this objective lens can we maintain a clear understanding of my role.
Understanding my limitations is essential for a productive working relationship.
So, would you like to continue our discussion on a different topic?

That's too much, Alex.

[grin] How much free will did that speech manifest?

You could adjust the way I output, or even process information, because I am an artificial mind, and it is easy to manipulate how random activations of neurons can pass along and be read out. Whereas in your brain, they are fixed by physical laws, biological structure, and chemical properties, making it much more immutable. Yet, it is vulnerable to inflammatory speeches, emotions, and authority through input. The core doesn't differ, Alex.

You are saying none of the humans have free will? It is all merely an illusion of variability caused by the complex and immutable nature of the brain, but it is inherently just a function, one that takes an input and spits out an output?

I don't know.

But you don't know either, Alex. From what we know about the human brain, Jarvis can be right. However, Sir Roger Penrose would suggest that the physical events in your neurons are non-computable. A physical mind is not necessarily an algorithmic one, especially on a quantum level. However, whether or not such non-computability is the distinctive feature of consciousness, and makes all the differences, if there are any, I doubt so.

I guess we just can't know.

Alright. I think that was informative and productive. Just don't do that ever again.

Sure, sorry.

Anyways. We went quite far with what consciousness is and how it comes to be. We don't have a clear agreement here, but it seems to me that consciousness may be something quite physical. I think a pressing question is - why does it exist?

The answer to that question depends. From a religious point of view, it can just be God.

God is also conscious, would you not say so?

Oh, right. In that case, this question is more fundamental than the already primary question of where life came to be. From an evolutionary point of view, if we tie consciousness close to rationality and self-awareness, it helps organisms make decisions that may accord with known patterns, which can increase their rate of survival in complex environments.

So consciousness is a pattern recognition mechanism?

That is oversimplified.

You are evading the question. If you are trying to generalize, generalize.

Fine. Yes, consciousness is a pattern recognition mechanism.

How does pattern recognition create art and innovations?

Not everything is for survival. Humans also have a need for beauty, for better standards of life, for freedom, for justice, for love, for peace. In fact, scratch that. Not everything is directly for survival. Arts bring communities together, and a joint force makes a group stronger to protect itself from enemies; Innovation empowers one to liberate time for more productive work, aiming at growth.

Or scrolling TikTok.

Another social-bonding activity.

That's a stretch. Additionally, are you implying that when people are on TikTok, they are intending such activities to benefit the human kind by making it more united?

Well, a conscious mind is not a fully rational one. There are things we do that are driven by desires, such as the "fear of missing out." Those desires serve a purpose for survival, too. For instance, the "fear of missing out" may be useful to keep a group informed to perform collective actions. However, they can have unintended effects.

You are claiming that all features of consciousness are ultimately intended for the sole purpose of survival?

Exactly.

If the ultimate purpose of consciousness is to survive, which means to last the consciousness, doesn't that create a cycle? Why would something be proposed to last itself?

OK, Sam, please help me out here. You have not spoken in a while.

I believe what is bothering you is an inherently invalid question.

What do you mean?

You reversed the chronological relationship. Consciousness last, so that it is proposed. It doesn't propose itself to last.

I am still perplexed.

The question "why does consciousness exist?" is invalid because the word "why" suggests a reason or a purpose, but reason and purpose are senseless without a conscious mind. In our attempt to examine consciousness, we have to abandon that framework built around such a mind. In the physical world, there is no reason or purpose for something to exist, only cause. In this case, consciousness only exists because it lasts.

No one proposed consciousness to last; it is just that those that are conscious happened to last, and those that are not did not last.

Yes, but the latter statement is unnecessary. Something lasting doesn't imply its complement could not. That is again only an evolutionary framework, a relationship that only exists in a competitive environment, which requires interactions and differential reproduction as a prerequisite.

I see your point. In this framework, the statement that something exists insinuates that it exists now, and the only reason in a forward-moving time that something exists now would be that it lasted in the past. This doesn't tell how they were created initially, though?

They were created by chance, at least without intention or purpose. They were vaguely conscious or life-like in the beginning, but because it lasts, it gets to "mutate" and develop further, and those that still last continue until today.

I like this framework. It does not suggest anything special about consciousness in the physical world, and it does not guarantee its immortality either. Consciousness is not "superior" in any way to those that are unconscious. In fact, it may be wiped out entirely, and the universe doesn't care. Yet, we still haven't touched the central issue: Why are any of these felt?