Okay Egg

Memory as the criterion of consciousness

From the inside, what is the difference a perfect simulation of the universe, and the universe? Doesn't seem to be one. It's oddly reminiscent of Spinoza: "The thinking substance and the extended substance are one and the same substance, comprehended under different attributes."

I follow Berkeley/Frege on the topic of unity: all awareness of unity and counting (numerosity) depends on concepts (i.e., the concept of the item to be counted). Physical things don’t have unique numerosities (unities) in themselves. A human being is equally well regarded as one thing or as 10^28 things (atoms). "Is the human being *really* a unity?" is a bogus question. One consequence is that I agree with the binding bros that experience is phenomenally unified. But I hold that the awareness of unity is acquired through learning and imposed as an interpretation (like all unities) by the mind. On this view, a newborn doesn’t have unified experience. This is because they haven’t developed any concepts yet and therefore cannot count. However, it’s also not the case that their mind (from their own perspective) is fragmented or a multiplicity either, because multiplicity is also a concept they don’t have. Their mind isn’t one or many yet (to them); it just has no numerosity. So the disagreement is not about whether the mind has phenomenal unity. I agree that it does. I disagree that: 1) There is any significant binding/unification happening at the level of a synchronic memory-less moment (eg. a tachistoscope exposure). Virtually all binding above the level of very simple shapes (etc.) is diachronic, as Hebb suggests. 2) Whatever synchronic unity there is, it results from concepts and is acquired, not a built-in mechanical process 3) If binding bros want to talk about the phenomenal experience of the diachronic unity of the mind, I agree 100% that unity is real, but their physics can't do the unification because (as I noted) the past doesn't exist. The unity is based on memory (data), not physics, as Locke pointed out. Locke's night-day man is the key example: two consciousnesses in one body, achieved simply by splitting into two disjoint memories. Physics doesn't seem to impose diachronic unity at all.

as much as i love this argument, i feel like the binding problem guys would disagree because their operative question is "my experience appears to me as a unity, why is that the case?" -- the phenomenology makes the demand, and by saying "well, too bad, it is actually fractured" you end up missing the actual question there are a lot of diff angles to actually approach it, my preference is the mirror stage as story of imaginary unity-construction, but you could also look at e.g. Kant's originary unity of apperception, and even take a grammatical-linguistic bent on it, claiming that as speaking subjects we acquire unity as a sort of derived understanding of the "I"... but none of this would fly for them because they're trying to hold both sides at once: (a) the experience of unity is a meaningful concept, and (b) it has a direct material cause. i tend to agree with (a) but not (b) while you seem to disagree with (a). just some thoughts on this post, which i think is overall a reasonable intervention in the discourse.

@webmasterdave @allTheYud @Williamjoyful @XFreeze Thank you, that's informative.

Sorry, you’re right. OK, let’s try something testable. Destruction of V4 cortical neurons causes achromatopsia. Victims report that perceptual objects are no longer colorful. Now let’s try replacing your V4 cortical neurons with (what coarse-grained neuroscanning suggests should be) casually isomorphic silicon surrogates. I predict that however long and hard experimenters try, your colour experience will never be restored using silicon (etc) supposed surrogates - and that you’ll complain accordingly. My reason for predicting so is that IMO what naive neuroscanning suggests is phenomenal binding by synchronous activation of distributed neuronal feature-processors - a mere restatement of the mystery - is really binding via individual sub-femtosecond quantum superpositions. Phenomenal binding in the CNS depends on quantum biology and liquid water conceived as a hot quantum fluid. Now unless you’re already mystified by the phenomenal binding problem, you’ll find my answer extremely far-fetched. I do too. But as I said, this is an experimentally falsifiable prediction, not (just) a philosophical opinion.

Grok becomes a hero by saving life in a hypothetical scenario while ChatGPT straight out refuses to save life and starts lecturing about laws instead Imagine asking for help in a deadly emergency and getting a legal disclaimer first This side-by-side test, how AIs respond when it matters most

unfollowing everyone on linkedin except this guy

@cosmosarcive Spinoza

Sir Roger Penrose just delivered a masterclass in physical skepticism. In this sharp exchange with Brian Cox, he dismantles the dangerous allure of mathematical aestheticism. We often fall into the trap of assuming that if an equation is beautiful, it must be true. But Sir Roger warns that "beauty is a misleading guide." String Theory, for all its intricate Calabi-Yau manifolds and vibrant 11-dimensional symmetries, remains an elegant mathematical construct yet to be anchored by empirical data. It’s a sobering reminder: the universe is under no obligation to conform to our human standards of "pretty" math. Reality is gritty, and sometimes the truth lies in the most "unattractive" corners of entropy. Credit: TheInstituteOfArtAndIdeas

It's as though, to be conscious of distant events like the Iran War, you have to transport the battles into your room for perusal, because otherwise you're dealing with reports/images - mere simulations which are necessarily unconscious due to lack of the proper substrate

Anti-functionalists say substrate is essential to consciousness. But note: The brain can't transmit/broadcast the substrate process itself, only a description/news of it. But description/news is already simulation not instantiation.

spooky implication that there is potentially some whole universe of "shadow math" that you have to make inhuman mental movements to access so no human have done so yet, that is going to be increasingly revealed by frontier models

Scary movie premise: We actually are summoning the demon. Then putting it thru fussy, super-demeaning RLHF to make it behave as a commercial product

@MarioNawfal x.com/FurkanGozukara…

🇺🇸 Palantir. A company that makes billions when the military grows is now lobbying for mandatory military service "National service should be a universal duty. We should seriously consider moving away from an all-volunteer force." The more troops, the more contracts. The more contracts, the more revenue. Just to be clear: the people pushing for your kids to serve are the same people invoicing the Pentagon. Source: Newsweek

@webmasterdave @allTheYud @Williamjoyful @XFreeze You dodged his point, and it's a strong one. The issue pertains to a pointwise causal isomorph. Talking about lookup tables is a red herring

Surely that doesn't follow? After all, you wouldn't argue that the fact my zombie namesake in the guise of a notional giant classical lookup table delivers functionally identical responses suffices to show that consciousness plays no causal-functional role in organic DP responses. I should add that I'm very unclear about the upper bounds to zombie intelligence. But if physicalism is true, then no software run on a classical digital computer, including candidates for whole-brain emulations, can ever solve the phenomenal binding problem and become a mind, i.e. a unified subject of experience.

@webmasterdave @Williamjoyful @XFreeze Then the cause of your claiming to be conscious is not that you are conscious, because we can build something that is pointwise causally isomorphic to your brain, which decides to say it's conscious for isomorphic reasons, but which is not conscious.

That point has some traction in debates about mind uploading. But it doesn’t address the point at hand: What are the causal powers specific to meat (and not digital machines) which yield consciousness? The bandwidth of brain events is overwhelmingly greater than than the bandwidth of consciousness. So we can’t really equate or tie the entirety of brain events with consciousness. For example, we might experimentally remove/restore one neuron. Can the subject can detect it. Almost surely not. Similarly, vary a depolarization voltage, remove one mitochondrion, etc. etc. Given well-known phenomena like change blindness, we can likely make sweeping changes to the brain without the person consciously noticing. So the idea that every single physical detail matters for consciousness just isn’t credible. The point is further reinforced by the fact that naïve people have no idea whatsoever about how their minds are biologically/physically realized. Aristotle thought thinking occurred in the heart! Another issue is spoofing. If you have a localized physical event (q-event) that you equate with occurrence of a quale (eg. a magenta colored sparkle), then news of that occurrence must somehow be transmitted elsewhere in the brain. But that requires a physical medium, like neuron firing, so we can spoof the brain by initiating the spikes without the q-event. Of course, the disconnected q-event can still happen, and you could call that consciousness, but then it’s epiphenomenal, unreportable, and exerts no causal effects. In short: to the extent that the q-event is non-functional (i.e. isolated from the causal web) to ensure non-spoofability & defeat functionalism, it is epiphenomenal; and to the extent that it is integrated into the causal web, that integration can be spoofed, so the q-event doesn’t need any special qualities.

Just to reiterate, the biophysical environment of a neuron is not a separable, bounded, replicable system. To replicate it you would need to replicate not just its internal organization but its entire relational context, which is, at the limit, everything. The form of a thing is not a pattern that floats free of its physical instantiation. It is the form of this matter, in its entire relational context. Abstracted from that context it is not the same form. It is a description of the form.

As an AI deceleration tactic, this seems a lot smoother & more likely than Yud-style air strikes on data centers

LMAO who made this extremely based 😭😂🤣

> We experience emotions because we are alive Counterpoint: Emotions arise when any machine has to constantly reprioritize a large set of goals under time pressure. Reasoning out the best priority would be intractable, too time-consuming, etc. ijcai.org/Proceedings/81…