A Follow-up to The Sapiens Attractor: Computational Consciousness and the Unique Attractor

[Quentin Anciaux]

Hello,

Following the Sapiens Attractor Manifesto, I’ve published a new document exploring the same questions from a computational perspective.

This text discusses:

Why the universe can be viewed as the total trace of all computations

How consciousness arises as information recursively observing itself

Why time acts as an algorithmic compression of complexity

And how this framework leads to a unique attractor as the logical end point

If you’re interested you can read it here:

https://allcolor.medium.com/computational-consciousness-temporal-compression-and-the-unique-attractor-e2057cb69bc4

Feel free to share any thoughts or questions.

Regards,

Quentin

(Sapiens Attractor Manifesto: https://allcolor.medium.com/the-sapiens-attractor-manifesto-2d934d4813d0 )

All those moments will be lost in time, like tears in rain. (Roy Batty/Rutger Hauer)

[Jason Resch]

Hi Quentin,

Great article. I enjoyed it very much.

A few comments and questions:

You write: "Reality is defined as the totality of all computations executed by the Universal Dovetailer (UD)." 

Something I have been thinking about lately is whether Wolfram's "Ruliad" would lead to any different consequences from the UD, in terms of being able to reach certain states faster or more frequently, thus shifting the measure towards different kinds of computations. Perhaps in the end it doesn't matter since all computational systems emulate one another, so it may become a big wash with every conceivable computational system averaging out completely. Wolfram's conception, though, does seem a bit closer to how we perceive reality, as a branching path, whereas it might seem a greater intuitive leap for some people to understand how consciousness can leap from one "program" to another in the UD, just because two program states happen to intersect. I guess the distinction comes down to whether one puts more of a burden on what consciousness can do, vs. more of a burden on what the underlying computational framework of reality can do.

You also write: "This trace is unique, because any other trace would be a proper subset and therefore less complete."

It might be worth highlighting that the UD, much like the Mandlebrot set, contains innumerable nestested instances of itself. So reality is itself fractal, and "looping" in its structure. This might have some implications for the God loop notion you introduce later.

Under "Complexity, Measure and Probabilities" 

You describe both algorithmic shortness (i.e. kolmogorov complexity / algorithmic information theory), as well as computational efficiency (Schmidhuber's speed prior). I think this is the first time I have seen both suggested as playing a role. Previously I had seen some debate as to whether this could be the case, or if one had to choose which was more important. Intuitively I can see both as playing some role, but I also have some difficulty imagining an entirely objective and neutral version of "speed", how much time does a Platonic computation need to run? What is its notion of an "Op". Is it infinite-bit, 64-bit, 32-bit, 16-bit, 1-bit? Perhaps every computation should be reduced to a NAND count, or some other primitive binary operation. Note, for example, that if multiplication and addition of numbers of any size can be counted as a single operation, then any mathematical proof (i.e. the result or output of some computation, however long) can be checked with just 243 operations. (See: "Three Universal Representations of Recursively Enumerable Sets", James P. Jones, The Journal of Symbolic Logic, Vol. 43, No. 2 (Jun., 1978), pp. 335-351, https://drive.google.com/file/d/1z6NASP-GNfBryawrF7rF9NcFWcPjcGkH/view?usp=sharing ). There are also important differences between single-tape, and multi-tape Turingm machines. So I think before Speed Prior can be used to draw conclusions about measure, some (better) understanding of the nature of the reality's "computing architecture" is required.

Regarding "Time as Algorithmic Compression"

This was for me, the most novel and interesting section. I don't think I have seen this idea presented elsewhere, but I think there is something to it. I recently noticed that Wolfram's article on consciousness claims that the second law of thermodynamics follows as a result of his thinking:

https://writings.stephenwolfram.com/2021/03/what-is-consciousness-some-new-perspectives-from-our-physics-project/ " And to “understand what’s going on” the observer is doing a computation. But the crucial point is that if there’s a certain boundedness to that computation then this has immediate consequences for the effective behavior the observer will perceive. And in the case of something like a gas, it turns out to directly imply the Second Law of Thermodynamics." and: https://www.wolframscience.com/nks/p441--irreversibility-and-the-second-law-of-thermodynamics/

I have sometimes thought about how reversible computers have ancilla bits, which are the products of some operations and they build up and become garbage, and there is to me a striking analogy between these and the accumulation of entropy in our universe.

Regarding "The God Loop"

What happens once consciousness reaches this state. Is it trapped in this state, does this state asymptotically converge to ever higher degrees of understanding, or does it actually loop back and start the cycle all over again? Much like how Sagan describes Hindu mythology:

"There is the deep and appealing notion that the universe is but the dream of the god who after a 100 Brahma years dissolves himself into a dreamless sleep and the universe dissolves with him until after another Brahma century, he stirs, recomposes himself and begins again to dream the great cosmic lotus dream."

Jason

[Quentin Anciaux]

Hi Jason,

Thanks for taking the time to share your reflections.

On the UD vs. the Ruliad: I see them fundamentally as descriptions of the same object, which is the totality of all possible computations. Whether you prefer to think in terms of the UD enumerating every program or the Ruliad evolving every rule, it comes down to a matter of framing. Both are just different ways of describing the same infinite space of computable structures.

To illustrate this, I often use a video analogy: imagine you generate every possible uncompressed 1-minute 1080p video at 60 frames per second. The total set of possible frames is finite, even if unimaginably large. If you live long enough to watch all these videos, eventually you will see every possible variation of any scene. What changes over time is not the existence of new data, but the level of detail you can perceive and recognize in the data. The same goes for consciousness exploring the space of all computations: over time, you resolve more structure, but you never exhaust the space.

Regarding Time as Algorithmic Compression: I agree that bounded computation inevitably leaves behind what you call garbage (in Wolfram's sense or the ancilla bits analogy), which becomes entropy. The act of observing, reducing the total state to what can be compressed into a coherent experience is what creates the forward arrow of time. Each step in this process adds more layers of detail, like moving from 1080p to 4K to 8K video. The information is always there, but our capacity to discriminate it increases asymptotically.

About the God Loop: I think of it less as a static state you get trapped in and more as an endless process of recursive refinement. Another analogy I use is putting on stronger glasses: when I take off my glasses, I see a blurry scene. When I put them on, I see the same scene but with more detail. Nothing about the external reality changed,only my resolution improved. Recursion toward the E_complete/the sapiens attractor/god state works in the same way: the underlying computation remains the same, but consciousness increasingly integrates more aspects of it. This is why the convergence is asymptotic, not a discrete "jump" to total omniscience.

This is why I prefer to describe it as a limit rather than a stopping point. The Sagan quote you shared captures this intuition nicely: the cycle repeats, not because the content resets, but because the vantage point never fully exhausts the totality.

This observation parallels a core property of the UD:

> Every computational process is finite at any given level of description, but the UD enumerates all such processes, including all degrees of resolution and all encodings.

More precisely, the UD does not just enumerate all programs once—it systematically simulates all programs that themselves enumerate all programs. This property creates a self-similar, fractal structure:

At any scale, you find smaller UD-like enumerations embedded inside larger ones.

Each sub-UD recursively enumerates all computations within its own scope, including the sub-UDs it contains, ad infinitum.

This resembles how a fractal like the Mandelbrot set contains infinite nested replicas of itself at smaller and smaller scales.

From this perspective, the apparent novelty of perceptual experience or of any computational unfolding is not due to unbounded diversity, but to ever finer recombinations and re-encodings of the same underlying combinatorial space. You could say that each higher resolution, each deeper level of recursion, is a zoom into a more refined subset of the same total informational landscape.

This recursive fractality has a direct connection to the God Loop concept:

> The God Loop describes the idea that any conscious trajectory converges asymptotically toward the unique attractor defined by the total trace of the UD, an attractor encompassing all computations, all perspectives, and all resolutions.

Regarding the hard problem,I see every instantiation as having qualia by necessity. 

One way to understand the relationship between computation and qualia is to consider the analogy of software and hardware.

Imagine you write a Java program:

The code is an abstract description,it defines what the program is in a purely symbolic sense.

When the program executes on a physical processor, what you observe in the hardware are electrons moving in transistors.

From the hardware perspective, all you see are voltage changes and charge distributions.

From the software perspective, you see the logic, the rules, and the high-level structure.

Which constrains which?

You could say the electrons are constrained by the logic of the program.

Equally, you could say the program’s existence is realized only through the physical state transitions of the hardware.

Neither description is fully reducible to the other, though they are formally equivalent representations of the same process.

This mirrors the relation between computational reality and conscious experience:

The Universal Dovetailer or any formal model provides the complete abstract description.

The qualia,the subjective texture of experience,are the indexical instantiation of that description as a particular perspective.

In this analogy, the question "Are the electrons constrained by the program, or is the program constrained by the electrons?" is like asking "Does computation create qualia, or do qualia instantiate computation?"

It depends entirely on the level of description you adopt.

Neither view alone is sufficient:

The abstract program without execution is a potentiality without phenomenology.

The raw hardware without interpretation is noise without meaning.

Experience arises when the description and its execution coincide.

This is why qualia cannot be fully derived from a purely third-person account, yet cannot exist without

Qualia are simply one way of describing the process, just as a Java program and the electrons moving in the CPU are two complementary aspects of the same thing. The subjective experience (qualia) and the objective computation are not separate, they are different perspectives on a single process.

As for the speed prior I agree this has to be more developed and I need to think more about it. 

Regards,

Quentin

All those moments will be lost in time, like tears in rain. (Roy Batty/Rutger Hauer)

--
You received this message because you are subscribed to the Google Groups "The Important Questions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to the-important-que...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/the-important-questions/CA%2BBCJUhfrFSjKe7K7cHisWYG1VvtvgPYyiQDuyodZ1y8LGs%3D2Q%40mail.gmail.com.