[mjr00]

I'm personally a lot more than a prediction engine, don't worry about me.

For those who do believe they are simply fleshy token predictors, is there a moral reason that other (sentient) humans can't kill -9 them like a LLaMa3 process?

[mewpmewp2]

Morality is just what worked as set of rules for groups of humans to survive together. You can try to kill me if you want, but I will try to fight back and society will try to punish you.

And all of the ideas of morality and societal rules come from this desire to survive and desire to survive exists because this is what natural selection obviously selects for.

There is also probably a good explanation why people want to think that they are special and more than prediction engines.

[quesera]

Yes, specifically that a person's opinions are never justification for violence committed against them, no matter how sure you might be of your righteousness.

[mjr00]

But they've attested that they are merely a token prediction process; it's likely they don't qualify as sentient. Generously, we can put their existence on the same level as animals such as cows or chickens. So maybe it's okay to terminate them if we're consuming their meat?

[quesera]

"It is your burden to prove to my satisfaction that you are sentient. Else, into the stew you go." Surely you see the problem with this code.

Before you harvest their organs, you might also contemplate whether the very act of questioning one's own sentience might be inherent positive proof.

I'm afraid you must go hungry either way.

[mjr00]

> "It is your burden to prove to my satisfaction that you are sentient. Else, into the stew you go." Surely you see the problem with this code.

It's the opposite; I've always assumed all humans were sentient, since I personally am, but many people in this comment section are eagerly insisting they are, in fact, not sentient and no more than token prediction machines.

Most likely they're just wrong, but I can't peer into their mind to prove it. What if they're right and there's two types of humans, ones who are merely token predictors, and ones who aren't? Now we're getting into fun sci-fi territory.

[quesera]

And how would we discern a stochastic parrot from a sentient being on autopilot?

So much of what we do and say is just pattern fulfillment. Maybe not 100%, on all days.

[lupire]

Why would sentient processes deserve to live? Especially non sentient systems who hallucinate their own sentience? Are you arguing that the self aware token predictors should kill and eat you? They crave meat so they can generate more tokens.

In short, we believe in free will because we have no choice.