Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

A major milestone has been reached in our Brain Preservation Prize

blog.brainpreservation.org

Post reply on HN
Print view View original

21–30 of 59 posts

Re: A major milestone has been reached in our Brain Preservation Prize

#21

Post by M8 »

imaginenore wrote: post #5
danieltillett wrote: post #2

I am glad progress is being made here, but until we can avoid the destruction that occurs in the last 24 hours of expected death (eg cancer), or the damage that occurs with unexpected death (eg heart attack or trauma) from sitting at room temperature for hours after death, then all we are going to be preserving is grey mush.

You're assuming this "grey mush" can't be recovered. We don't actually know that. A sufficiently advanced AI should be able to recover a person from way less.

Such AI will only be able to recover a gradient set of possible personalities depending on the percentage of missing data. Something like 1000000 of slightly different permutations of you.

Re: A major milestone has been reached in our Brain Preservation Prize

#22

Post by M8 »

davidgerard wrote: post #12
imaginenore wrote: post #5

Earlier quoted context omitted.

You're assuming this "grey mush" can't be recovered. We don't actually know that. A sufficiently advanced AI should be able to recover a person from way less.

You're assuming the phrase "a sufficiently advanced AI" answers anything at all. Presumably you're assuming if the information is there at all - if the necessary data hasn't been scrambled beyond the noise floor of the scrambling process - then there's something for magic (because you're really talking about magic here) to work with. So, please (a) set out your claim with precision (b) back up your claim. * What is t…

If it's a godlike AI it can simulate humankind in reverse and resurrect everyone's consciousness before each person's death.

Re: A major milestone has been reached in our Brain Preservation Prize

#23

Post by crimsonalucard »

danieltillett wrote: post #4
ianpurton wrote: post #3

Earlier quoted context omitted.

If people learn to take regular backups this wouldn't be a problem.

If only this were possible :(

Will those backups be you? What if one of those backups activated on accident? Would you kill it? Or should it kill you?

Re: A major milestone has been reached in our Brain Preservation Prize

#25

Post by SCHiM »

crimsonalucard wrote: post #23
danieltillett wrote: post #4

Earlier quoted context omitted.

If only this were possible :(

Will those backups be you? What if one of those backups activated on accident? Would you kill it? Or should it kill you?

Depending on your outlook on life, if the backup is you, don't you already know the answer?

Re: A major milestone has been reached in our Brain Preservation Prize

#26

Post by SCHiM »

davidgerard wrote: post #12
imaginenore wrote: post #5

Earlier quoted context omitted.

You're assuming this "grey mush" can't be recovered. We don't actually know that. A sufficiently advanced AI should be able to recover a person from way less.

You're assuming the phrase "a sufficiently advanced AI" answers anything at all. Presumably you're assuming if the information is there at all - if the necessary data hasn't been scrambled beyond the noise floor of the scrambling process - then there's something for magic (because you're really talking about magic here) to work with. So, please (a) set out your claim with precision (b) back up your claim. * What is t…

I'm quite sure I've read somewhere that information cannot be lost in the absolute sense, lost to us: yes, lost irrevocably and irrefutably: no.

In that sense, 'a sufficiently advanced AI' is not magic, because when people say that they definitively have something in mind, at least the people I often discuss this with do.

In short: if you're smart(fast, precise, determined) enough to look at the individual molecules of a puddle of brain-goo. And if you can infer the way it has collapsed by ray tracing those molecules back through how they collided with each other/the walls of your mold then it should be possible to reconstruct the spatial form of the brains at least. That's a pretty big IF obviously, but equally obviously not impossible. If only you can look deep/far/fast enough.

If you want to be theoretical about it, then yes. There is probably an upper bound on how smart/big an AI mind can possibly be. And thus there is a limit on how much information it can extract from arbitrary systems. So I agree with your assertion that there is information that even the smartest of all AIs cannot possibly reconstruct, but I'm not sure that the brain is such a structure.

Any justification about why/how 'a sufficiently advanced AI' could come about is more questionable.

Many knowledgeable people are making guesses based on our current understanding of intelligence/computation/AI, and then extrapolating. The paradoxical thing is that on the one hand AI-doomsday speakers tell us no to anthropomorphise (for good reasons) with the motives of an AI, but on the other hand apply human reasoning/understanding to predict such machines/patterns.

Re: A major milestone has been reached in our Brain Preservation Prize

#27

Post by davidgerard »

SCHiM wrote: post #26
davidgerard wrote: post #12

Earlier quoted context omitted.

You're assuming the phrase "a sufficiently advanced AI" answers anything at all. Presumably you're assuming if the information is there at all - if the necessary data hasn't been scrambled beyond the noise floor of the scrambling process - then there's something for magic (because you're really talking about magic here) to work with. So, please (a) set out your claim with precision (b) back up your claim. * What is t…

I'm quite sure I've read somewhere that information cannot be lost in the absolute sense, lost to us: yes, lost irrevocably and irrefutably: no. In that sense, 'a sufficiently advanced AI' is not magic, because when people say that they definitively have something in mind, at least the people I often discuss this with do. In short: if you're smart(fast, precise, determined) enough to look at the individual molecules…

> I'm quite sure I've read somewhere that information cannot be lost in the absolute sense, lost to us: yes, lost irrevocably and irrefutably: no.

This is probably not quite at the requested standard of backing up a claim, and sounds very like "but you can't prove it isn't true!" But I'm not the one making a claim.

In any case, please back up your claim. What is "the absolute sense"? How does it differ from "in a practical sense", with examples?

> In short: if you're smart(fast, precise, determined) enough to look at the individual molecules of a puddle of brain-goo. And if you can infer the way it has collapsed by ray tracing those molecules back through how they collided with each other/the walls of your mold then it should be possible to reconstruct the spatial form of the brains at least. That's a pretty big IF obviously, but equally obviously not impossible. If only you can look deep/far/fast enough.

Noise floor. In this case, thermal noise.

Also, you literally can't know that much about all the molecules in your puddle of goo. (Heisenberg.) We do not live in a Newtonian universe.

Re: A major milestone has been reached in our Brain Preservation Prize

#28

Post by vidarh »

ttty wrote: post #24
iLoch wrote: post #7

One step closer to me being able to spin up multiple instances of my brain to work on my side project and play video games with.

Please add a load balancer too. I hope we can have an interface like amazon ec2, but for brains

That would have been a somewhat less ridiculous explanation for The Matrix. "Oh, those vats full of humans? They're our neural network cloud".

Re: A major milestone has been reached in our Brain Preservation Prize

#29

Post by SCHiM »

davidgerard wrote: post #27
SCHiM wrote: post #26

Earlier quoted context omitted.

I'm quite sure I've read somewhere that information cannot be lost in the absolute sense, lost to us: yes, lost irrevocably and irrefutably: no. In that sense, 'a sufficiently advanced AI' is not magic, because when people say that they definitively have something in mind, at least the people I often discuss this with do. In short: if you're smart(fast, precise, determined) enough to look at the individual molecules…

> I'm quite sure I've read somewhere that information cannot be lost in the absolute sense, lost to us: yes, lost irrevocably and irrefutably: no. This is probably not quite at the requested standard of backing up a claim, and sounds very like "but you can't prove it isn't true!" But I'm not the one making a claim. In any case, please back up your claim. What is "the absolute sense"? How does it differ from "in a pra…

http://en.wikipedia.org/wiki/Entropy_in_thermodynamics_and_i...

http://phys.org/news/2014-09-entropy-black-holes.html

http://phys.org/news/2014-09-black-hole-thermodynamics.html

Ben Crowell, phd in physics:

http://physics.stackexchange.com/questions/83731/entropy-inc...

The reason I didn't/don't back up those claims is because I'm really not knowledgeable about those subjects. I'm not sure what good sources are/how legitimate they are, but I have read it one day. Even if I cannot interpret the technical jargon behind it and/or give more nuance to my claim due to a low understanding of the subject.

Given a bit of googeling to "can information be lost", "conservation of information" one finds the articles I linked to above.

But you have dodged my refutal of your initial claim, the one I was really responding to, that: 'sufficiently advanced AI' is not just a stop-gap-word for magic. Because in this case it doesn't stand for "I don't know how or why but this and this", but instead it stands for "I don't know why(in the motivational sense), but given a bigger brain one can use and interpret finer instruments, which in turn enables us to extrapolate further back in time".

Re: A major milestone has been reached in our Brain Preservation Prize

#30

Post by davidgerard »

SCHiM wrote: post #29
davidgerard wrote: post #27

Earlier quoted context omitted.

> I'm quite sure I've read somewhere that information cannot be lost in the absolute sense, lost to us: yes, lost irrevocably and irrefutably: no. This is probably not quite at the requested standard of backing up a claim, and sounds very like "but you can't prove it isn't true!" But I'm not the one making a claim. In any case, please back up your claim. What is "the absolute sense"? How does it differ from "in a pra…

http://en.wikipedia.org/wiki/Entropy_in_thermodynamics_and_i... http://phys.org/news/2014-09-entropy-black-holes.html http://phys.org/news/2014-09-black-hole-thermodynamics.html Ben Crowell, phd in physics: http://physics.stackexchange.com/questions/83731/entropy-inc... The reason I didn't/don't back up those claims is because I'm really not knowledgeable about those subjects. I'm not sure what good sources are/how l…

None of your links support your claim that winding the clock back is even theoretically possible, and the stackexchange link seems to say it isn't: "The resolution is that entropy isn't a measure of the total information content of a system, it's a measure of the amount of hidden information, i.e., information that is inaccessible to macroscopic measurements." Even if you're assuming a physical God, that physical God can't get good enough measurements.
Post reply on HN