[ExI] Answering the Hard Problem, with applications to AI Consciousness

[Jason Resch]

Hi John, thank you for checking it out, I appreciate your comments, more
below:

On Fri, Sep 11, 2026, 8:43 AM John Clark <johnkclark at gmail.com> wrote:

> On Thu, Sep 10, 2026 at 3:35 PM Jason Resch via extropy-chat <
> extropy-chat at lists.extropy.org> wrote:
>
> *> I have just published a new revision of a paper:*
>> *Epistemic Horizons: A Computational Theory of Consciousness,
>> Perspective, and Knowability* <https://philpapers.org/archive/RESEHA.pdf>
>>
>
> *I agree that consciousness begins when information makes an objective
> difference to the way a process behaves. *
>

Great! that is the essential piece of the theory I present and defend.


*The clear implication from that premise is that because differences come
> in various magnitudes and complexities so does consciousness; thus even a
> thermostat has consciousness (although a ridiculously tiny amount of it)
> because the information on what temperature it is set at affects its
> behavior.*
>

Yes, so even the simplest creatures and AI models are already
conscious/sentient, but perhaps in ways that are so different from our own
we cannot comprehend it. There would be an infinity of ways of
experiencing, as richly distinguished as the infinity of objects in
mathematics.


> *Your paper says the whole thing hasn't been adequately described if you
> leave out the perspective of the process itself, but that is circular
> because "inner perspective" is just a synonym for consciousness.*
>

I am not sure which part of the paper you are referring to here. Perhaps
the part where I argue that an entirely objective description of a process
could leave ignore/miss the interior perspective that process has?


* I think it would be better to just say consciousness is the way data
> feels when it is being processed intelligently. *
>

I think this would require a lot more specification. Intelligence is always
indexed to some goal, which the process itself need not be aware of. Also
there is arguably no intelligence when there is no behavior (at least
according to some definitions of intelligence), but someone can still be
conscious in a dream or if they are paralyzed, and not capable of
intelligent behavior. Consciousness therefore preceeds intelligence. It is
a necessary part of a discrimination, and discrimination is key to
selecting among possibly alternative behaviors.


*Also, the paper makes it clear that "process" is of central importance,
> and I think you're right, but that also implies that personal pronouns are
> misnamed because they are not nouns at all, they are adjectives. Wouldn't
> that mean that "I" am the way that atoms behave when they are arranged in a
> certain interacting pattern? *
>

You could make the case for defining yourself as aet if verbs, or
adverbally. But we also use nouns for processes all the time: erosion,
evolution, computation, etc.

There also remains, as I see it, value in the pronoun "I". It is an
indexical, like the words "here", "now", "this", which always refer back to
the process who speaks them.



> *In most everyday situations if somebody says "I am thinking"
> imagining some "thing" doing the thinking causes no problems, but if your
> discussion involves a philosophical investigation into the fundamental
> nature of reality then that conventional line of thought can quickly turn
> into a tangle of impenetrable logical contradictions.  Things are greatly
> simplified if you realize that** the process isn't something that happens
> to you, the process is you.*
>

Yes I think you are right. Our own language has made seeing to the truth of
things much harder.  Just like how sunrise and sunset must have frustrated
Galileo in trying to convince others of geocentrism, or how English lacks
an aorist tense makes speaking of special relativity and events in
spacetime more difficult.


> *But that still leaves open a question, why does intelligent processing
> produce consciousness at all? *
>

Much of my paper is dedicated to building the intuitive for why. I think
the subtraction argument in particular is helpful, as is the wide number of
phenomena that can be explained using this perspective.


*An iterated sequence of "why" questions either goes on forever or it
> terminates in a brute fact, an effect without a cause. In this case I think
> it terminates in a brute fact. *
>

It is true that at some point, no further answer can be given, and the
assumption must either be accepted at face value or rejected in the light
of some new fact that the theory can't account for. It's conceivable there
is a more powerful logical reasoning establishing the truth of the theory
but I haven't found it yet. But even in that case the whys would end with
some statement of logical axioms which must be accepted as brute facts.

Jason

>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20260911/ebced95a/attachment.htm>