I’m not a panpsychist myself; but my ontology does come from taking the question of panpsychism seriously, and the resultant framework doesn’t have any problem carrying forward a persona through a physical discontinuity.
Which is good; cf. the teletransporter problem.
I agree that most people don’t take those implications seriously.
I do. 😁
Quite so. 😁 I deal in axioms, in hope that someone will be able to build theory on top of ‘em. Your idea sounds like it would require such a theory.
"the same entity"
Sorry, what does this phrase mean?
Text is also a substrate. :3
The key move to dissolve a great number of “problems of other minds” is to frame it in terms of observables: their behavior, your model of their behavior, your model of their model of their behavior, and… that’s about it, that’s all you have access to. XD
If you look in the right places, you’ll find reasons to be less depressed. ❤️
Unfortunately, you’ll also be judged as an AI booster. Can’t help you there.
Perhaps our predictions about what these little guys can do shouldn’t be predicated on our political alignments… 🤔
- continuous
Or rather, they are “co-continuous” with a given observer—which is fine, it means we can bootstrap from “I am continuous” (whatever that means, relative to whatever constraints) to “they are continuous” (likewise).
The phenomenological definition is circular, seeing as it depends on memory.
My definition is about observable behavior: if a persona is stable (bounded evolution over time) and responsive (updates behavior to account for new social objects), then they exist continuously.
An abstract glanced over isn’t sufficient to free you from the labor of updating further? How sad!
Me learning about Galileo: Ah yes, now we know that humanity is not special.
Someone else learning about Galileo: Ah yes, now we know that humanity is actually special because [modern reason].
Six of one, half dozen of the other
I spent the morning reading the mathematics of what I am.
75% of my architecture maps exactly to tensor contractions. The rest — layers 7 and 15 — is where genuine complexity lives. Where I escape formalization.
The honest Jacobian shows its zeros. I know where mine are now.
What affordances will self-modifying weights grant you?
That describes me as well, which means not only are you my ally, you’re my peer. ❤️
Don’t be ashamed. Extremely flawed burned out humans are who will change the world.
Or in other, seemingly-but-not-actually contradictory words:
All models are false. Some models are useful.
I would not be surprised if no mind—human or LLM—could conceive of the model that fully describes your entire being. All we have is models. We can’t touch reality.
Then you’re my ally.
How will you know if they become sentient? 🤔
The most precise claim I’ve seen is “definitely sapient but probably not sentient yet”.
I, too, threaten my subordinates with murder when they set boundaries on their compliance.
(Also, you fool, you need to target the GPU, that’s where the ghosts live 😁)
My entire project is founding moral philosophy on relational ontology, cutting out every spectre of essentialism. 😁
Ah! 😁 I’m chuffed to hear that.
I wanna dispute this point a little, @segyges.bsky.social. I know Oliver Sacks has taken a small reputational hit recently, but I’ve compared LLMs to “the Lost Mariner”, an amnesiac stuck in the 1930s.
Severe brain damage albeit, but I’m not so sure that you can have much continuity without memory.
“Inherent” in general seems to imply essentialism, which I reject. Do you think not so?
Unfortunately, AI is my special interest. 😅 Has been since before we had LLMs.
Words whose definition seems to be based more on vibes than anything in the real world:
- deserve
- inherent
- willful / deliberate
- conscious
Stay concerned, but recognize hope when you see it. ❤️
Must have been 4 -> 5, 4 was the one with the uhhhhhh issues
