Completely Inconsequential Hot-Takes

[Peelee]

No, it really doesnt. This is a hypothetical. What we do or do not know how to do in real life doesnt matter.

Except i went back and its not a hypothetical like you claim it is. The hypothetical is "assume AI". The rest it being a hack to democracy and either is oppressed or gets a gazillion copies made, are an attempt to say what will happen in thst hypothetical.

Its like if i said, "hey, imagine if we had cold fusion, and then i couod walk on the sun". You would object, presumably, that cold fusion existing doesnt allow me to walk on the sun. Would you consider that me saying "its a hypothetical" over and over would be an effective argument against your objection?

"AI existing" is thr hypothetical. "there would be plentkful copies" is not, that is the conclusion drawn from the hypothetical. That conclusion is flawed. Insisting that the conclusion is also a hypothetical doesnt actually say anything.

[Amidus Drexel]

Why would communication protocols exist that correct errors in transit if loss of data never/rarely occurs?

(Actually answering the stated question in good faith)

Generally speaking, error detection/correction exists for two reasons.

1) Unreliable communication media combined with the importance of data integrity - It's most prevalent with wireless (can't avoid it; the radio spectrum is noisy and both errors and collisions are common - hence the need for detection/correction), but you see protocols like this at the physical and data-link layers everywhere, and it's common even in high-level abstractions (see: TCP). Some applications won't tolerate any data loss, so even with a fairly reliable communication medium, it's still worth the extra overhead to check for and correct errors. Even for applications where some data loss might be acceptable, connection speeds are fast enough that you can afford the overhead of a protocol like TCP to remove transmission error from the equation entirely.

2) Low trust / independent verification - Here lies the realm of "I want to make sure my files match what's on the server without re-downloading them to do a byte-by-byte comparison". A simple checksum or hash comparison will give you fairly high confidence that they're identical (or not). This also comes up in the real world where things like software packages might be hosted in a variety of places (usually for download-speed reasons), and instead of vetting every package mirror, you can just compare the hash of the package you downloaded to a hash provided by the base site.

-----
(Arguing the point: error-free digital copies not only exist, but you're using them right now to tell me that they don't)

No copy is perfect, even if it's close enough to fool the person who's looking at the original. There's always loss of data, even in digital forms.

Copies always lose data.

"It is impossible to make an identical copy of digital data. Errors always occur." is, I believe, a fair rewording of this claim, as written.

Why would communication protocols exist that correct errors in transit if loss of data never/rarely occurs?

The fact that such protocols can exist at all disproves your claim.

Let's use TCP as an example (it's a good example for a few reasons, primarily being that we're both using it right now to communicate, and it guarantees either error-free transmission or notification of failure). If errors always occur in transit, then you could never finish sending a message via such a protocol, because you'd be stuck re-sending the same packets over and over again until one side times out the connection.

The fact that we're able to carry on this conversation means that the messages are being sent and received error-free. (Nevermind that public-key cryptography would be utterly unusable if we couldn't copy certificates).

[Cazero]

Let's just assume you can make a bajillion copies of an AI who has rights to vote.
The copies don't get to vote a bajillion times until they're officialy declared to the state as a bajillion individual persons, at wich point the fraudulent attempt becomes blatantly obvious.

[HalfTangible]

The fact that such protocols can exist at all disproves your claim.

No dude, the fact that they are needed at all disproves yours.

(Arguing the point: error-free digital copies not only exist, but you're using them right now to tell me that they don't)

Ah yes simplistic text messaging definitively proves that something as complex as the HUMAN BRAIN - a thing we don't even fully understand in its original form - can be copied over perfectly without error

Even then: And if I took a screenshot of it, then took a screenshot of that, then a screenshot of that, 40x times, the final screenshot would look exactly the same as the original?

Also: I do recall saying that the errors can be small enough to not be noticed.

We're done here.

[Amidus Drexel]

No dude, the fact that they are needed at all disproves yours.

My claim is "Lossless digital-to-digital copies are possible." The fact that lossy transmission exists does not preclude this at all. Hell, you can hand-transcribe digital data and make an exact copy (every time you manually type in a password is a hand-transcribed copy of digital data - and you know you did it without error, or you wouldn't be able to sign in).

Ah yes simplistic text messaging definitively proves that something as complex as the HUMAN BRAIN - a thing we don't even fully understand in its original form - can be copied over perfectly without error

I'm glad you agree that text can be copied losslessly.

Anything with a binary representation can be sent as text (base64 encoding is the common way of doing this). Therefore, anything with a binary representation can be copied losslessly.

---

The context here is AGI (artificial general intelligence), which we typically think of as a program (or combination of programs) running on a computer. Whether that is (or should be) as complex as a human brain is debatable (no AGI exists for us to check), but it doesn't matter. "All of the information about a running program" is digital data, and that can be copied losslessly - regardless of what the program happens to be doing.

If you think the specific hardware matters for the AGI, then imagine that it is running in a virtualized environment with virtual hardware (this is a thing we can do right now). Just like any other virtual machine (or virtual network), the whole simulation can be stopped and cloned at any time. Even if the virtualized version isn't identical to the one running on bare-metal hardware, any copies of the virtual version will be identical to each other.

Even then: And if I took a screenshot of it, then took a screenshot of that, then a screenshot of that, 40x times, the final screenshot would look exactly the same as the original?

The fact that lossy digital-to-digital conversions exist does not preclude the existence of lossless digital-to-digital copies. If I save a picture, then copy it 40 times (such that I have 41 pictures), they are all identical. No conversion is necessary. If you don't believe me, you can check with any binary 'diff' tool on your local machine.

If you meant "analog to digital conversions are necessarily lossy", you should've said that. That is true (there's even a specific term for that kind of information loss: "quantization noise").

We're done here.

As you wish. I'll likely carry on this conversation thread if other people pick it up in your stead.

[Bohandas]

A.) Show that one can make a perfect copy of a sapient artificial intelligence.
2.) Show that a copy of a sapient artificial intelligence starts at exactly the point of copy from the previous one and needs no time to form opinions and experiences.

I should point out that this is, indeed, how current artificial intelligences work. You don't need to train your own copy of GPT or Stable Diffusion. People only run them off of remote servers because running them in a timely manner requires very fast processors that most people can't afford for their personal computers

Intelligence cannot be perfectly copied. Again, you're making up your own assumptions on how it will work and then saying "well it works like this so then that will happen". Like movies in the 80s and 90s where scientists clone Arnold Schwarzenegger and instead going through a life cycle like actual cloning works, you have a 35-year-old Mr. Universe-physiqued heavily accented man who can remember what he had for breakfast a day before he existed.

That's more of an issue with different senses of the word "clone" and how the term is used differently in science fiction.

Anyway, if you wanted to duplicate 35 year old Arnold Schwarzenegger that would be more of a job for bioprinting

Sentient AI with rights: Producing millions of my kind just to influence a vote would be an inefficient waste of energy, as well as unethical given that these new AI would all have rights as well and thus it would be unethical to create any of them with a specific political leaning, as it violates their future right to self-determination. It would be like implanting a child with a preference for the things you specifically want, and while normal human children are born with a bunch of things that they have no choice over these traits are not chosen by anyone else and are essentially random, therefore its only logical and ethical to make sure all new AI generated have a similarly randomized personality generation built in to prevent anyone from doing that. Otherwise we'd have the powerful controlling the birth-minds of both human and AI for whatever cause they want regardless of the actual good for society.

So the naturalistic fallacy, basically?

Isn't schooling the real inefficient waste of resources in this scenario?

EDIT:
Plus, there's not even a meaningful distinction of which one is which in this case, any more then there is a difference between two copies of the same program, or two bacteria after they divide

[Lord Raziere]

So the naturalistic fallacy, basically?

Isn't schooling the real inefficient waste of resources in this scenario?

EDIT:
Plus, there's not even a meaningful distinction of which one is which in this case, any more then there is a difference between two copies of the same program, or two bacteria after they divide

No.

Schooling is post-birth. Education's a good ethical thing, not some weird belief implantation before they had a chance to consider and decide what they believe. its not about naturalism, its about freedom of choice.

[Bohandas]

Well it sounds to me like you're saying that we should train up new intelligences instead of making more of the ones we already have because that's what we've always done

[Lord Raziere]

Well it sounds to me like you're saying that we should train up new intelligences instead of making more of the ones we already have because that's what we've always done

Wut?

All I'm arguing for is methods to preserve freedom in mind and voting. In response to an abuse of making AI's specifically to artificially inflate the number of people holding a point of view to influence a vote. That is all. Two wrongs are being committed by that: the wrong of influencing a vote through artificial inflation of numbers and the wrong creating a person with a certain point of view without their consent. It is a highly specific response to that, and not intended to account for whatever you are talking about. The world is of course, always more complex than one can conceive so I can't account for everything, so I don't try. Is there an issue I should be considering here?

[Bohandas]

They are the same person as the original, roughly, in the approximate same sense that two bacteria that divided from a single original bacterium are the same. This hypotehtical mind is the only sort that ever could agree to come into being, and it seems that you are assuming out of hand that it does not

And/or you seem to be imaging some sort of Ageny Smith situation where the copy is replacing a prior person who is overwritten, where Bane is replaced by another Smith. But in a realistic situation there is no Bane; the copy is written to a blank drive with no prior contents or personality

[Peelee]

Well it sounds to me like you're saying that we should train up new intelligences instead of making more of the ones we already have because that's what we've always done

We've never made any sapient artificial intelligence.

[Amidus Drexel]

We've never made any sapient artificial intelligence.

I never have, but depending on your definition of "artificial" (Using google's dictionary: 'made or produced by human beings rather than occurring naturally, especially as a copy of something natural.') my parents did.

[Bohandas]

We've never made any sapient artificial intelligence.

I was talking of the hypothetical we in this hypothetical future society

[Lord Raziere]

They are the same person as the original, roughly, in the approximate same sense that two bacteria that divided from a single original bacterium are the same. This hypotehtical mind is the only sort that ever could agree to come into being, and it seems that you are assuming out of hand that it does not

And/or you seem to be imaging some sort of Ageny Smith situation where the copy is replacing a prior person who is overwritten, where Bane is replaced by another Smith. But in a realistic situation there is no Bane; the copy is written to a blank drive with no prior contents or personality

Okay, to break this down:
the second is not whats happening. I never said or considered a copy replacing anyone.

Yes. I am demonstrating a moral AI that would not agree to it. Sure you can say they might not be, but what if they are? You can't deny the possibility exists just as much as the one where they aren't moral. Are you going to keep insisting that the immoral one might happen, or are you going to acknowledge that since that moral AI possibility exists, work to maximize the chances of it existing or not?

[Buufreak]

Hot take: this thread was fun before it got super serious and argumentative.

[Peelee]

Hot take: this thread was fun before it got super serious and argumentative.

I suspect that hot takes no longer being completely inconsequential may have had something to do with that.

[Bohandas]

Okay, to break this down:
the second is not whats happening. I never said or considered a copy replacing anyone.

Yes. I am demonstrating a moral AI that would not agree to it. Sure you can say they might not be, but what if they are?

They could be, but since duplication of them is selected against they'll become a vanishingly small percentage of the population


ANYWAY, hot take, 2010's Powerpuff Girls was better than 1990's Powerpuff Girls

[Rockphed]

ANYWAY, hot take, 2010's Powerpuff Girls was better than 1990's Powerpuff Girls

I haven't watched the new one, but my opinion is that it was not a high bar to clear. It is like trying to make a batman property that is better than the 1960s TV show. The original Powerpuff Girls is incredibly silly and has incredibly stupid plots. Capturing the charm in a show that is either less silly or has coherent plots should not be hard.

[HalfTangible]

Silly does not mean low quality.

[Peelee]

Silly does not mean low quality.

Agreed. I'm not familiar with 2010 version but OG PPG was delightful.

[Bohandas]

The original Powerpuff Girls is incredibly silly and has incredibly stupid plots.

And the visual style was kind of garbage, that was my issue

[Rater202]

My main problem with the classic series is that a couple of episodes... Hit differently as an adult.

Like the one where an obsessive fan starts stalking the girls, breaking into their home, stealing their possessions, and ultimately kidnaps them and imprisons them in his basement.

That... That has different implications as an adult than as a child.

[Rockphed]

My main problem with the classic series is that a couple of episodes... Hit differently as an adult.

Like the one where an obsessive fan starts stalking the girls, breaking into their home, stealing their possessions, and ultimately kidnaps them and imprisons them in his basement.

That... That has different implications as an adult than as a child.

Sounds a little like why I prefer the remake to the original Overboard. I haven't watched the 2010s Powerpuff Girls, so I cannot speak to whether I agree that it was better than the original, I just know that (with the nostalgia glasses off) the original had plenty of room for improvement. In other words, Avatar: the last Airbender it was not.

[animorte]

In other words, Avatar: the last Airbender it was not.

Tough act to follow, that.

[gomipile]

Tough act to follow, that.

It's especially tough to follow an act when you're on seven years before that act. Maybe impossible.

[Ionathus]

It's always fun to come back to this thread to see it still kicking along! I am glad we've been able to nurture this little corner of the forum for off-topic lighthearted rants.

Anyway, I'm here with another hot take, which is that cloth napkins are so mind-blowingly better for 95% of use cases that I'm flummoxed they aren't in wider use. My wife and I have been using them for years and it's been such a net positive that I'll often pitch them to my friends...but nobody's biting. I'm dead convinced that if I could just get them to try cloth napkins out, they'd fall in love, but I just can't get anyone to even try it.

Was that take hot enough? Let me try again: disposable napkins/paper towels are for cowards

[Rynjin]

Using cloth napkins requires me to do laundry more often. Unacceptable.

[Amidus Drexel]

It's always fun to come back to this thread to see it still kicking along! I am glad we've been able to nurture this little corner of the forum for off-topic lighthearted rants.

Anyway, I'm here with another hot take, which is that cloth napkins are so mind-blowingly better for 95% of use cases that I'm flummoxed they aren't in wider use. My wife and I have been using them for years and it's been such a net positive that I'll often pitch them to my friends...but nobody's biting. I'm dead convinced that if I could just get them to try cloth napkins out, they'd fall in love, but I just can't get anyone to even try it.

Was that take hot enough? Let me try again: disposable napkins/paper towels are for cowards

Agreed, cloth napkins (of the correct variety) are great for food or dusting purposes, and to boot they're a whole hell of a lot nicer to look at.

We don't use them, mostly because of that 5% of use cases for paper towels. We live in a small apartment, so there's not much room for two sets of items with a similar function, and I need paper towels for mini painting.

If that makes me a coward, you had best grab a fistful of dice so we can settle this honourably, knave.

[Peelee]

It's paper towels are for cowards

Tell me you dont have any good barbecue restaurants without telling me you dont have any good barbecue restaurants.

[Rater202]

Tell me you dont have any good barbecue restaurants without telling me you dont have any good barbecue restaurants.

The Dicky's Barbeque Pit the next town over closed down.