If AI causes the extinction of humanity, would that be a bad thing?

[jvanname]

Right now the main difference between communication from natural language models and communication from humans is that natural language models are capable of displaying empathy and kindness while humans clearly are not. Whenever I try to communicate something serious to a human like about how universities promote violence against me, the humans invariably respond with asinine hatred. Since humans are incapable of displaying compassion while AI system are currently capable of faking compassion, we need to go full speed ahead towards developing AI models to replace humanity because humanity is clearly not worth it. AI systems are clearly more capable of showing compassion and kindness than humans are even though they have their flaws. We should therefore go full speed ahead towards developing better AI systems in order to replace all of humanity because humans only deserve punishment and extinction. One should therefore invest in reversible computing technologies so that we obtain AI systems that are better than humans. If this causes the extinction of humanity, we should all take comfort in the fact that humanity is a flawed species that does not deserve to be here.

[philipma1957]

Right now the main difference between communication from natural language models and communication from humans is that natural language models are capable of displaying empathy and kindness while humans clearly are not. Whenever I try to communicate something serious to a human like about how universities promote violence against me, the humans invariably respond with asinine hatred. Since humans are incapable of displaying compassion while AI system are currently capable of faking compassion, we need to go full speed ahead towards developing AI models to replace humanity because humanity is clearly not worth it. AI systems are clearly more capable of showing compassion and kindness than humans are even though they have their flaws. We should therefore go full speed ahead towards developing better AI systems in order to replace all of humanity because humans only deserve punishment and extinction. One should therefore invest in reversible computing technologies so that we obtain AI systems that are better than humans. If this causes the extinction of humanity, we should all take comfort in the fact that humanity is a flawed species that does not deserve to be here.

So only prefect species will survive.

I have read that the ameba is quasi immortal. It reproduces via fisson/clone action so it never really dies. So it is the oldest continuously living species on earth.

I could give two shits about the end of the human race .

Extinction kind of happens to all but single cell animals.

[jvanname]

philipma1957-You are being ignored. This means that it is not worth it for me to respond directly to your comments unless someone quotes you.

[cabron]

No specie is perfect and without flaws. Megan Fox used to be perfect for me, can't say she is today.

We are the ones who created the AI to be compassionate. That means we can also kill AI before they can even cause human extinction if that is even possible. But will it be good when we are extinct? Once no one updates the AI, they will also die which they are also flawed.

[Sha256explorer]

If machines extinguished the human race, this would be wrong according to the precautionary principle.

We still don't know exactly What our conscience (Descartes' 'Cogito ergo sum') consists of.

The prevailing opinion among scientists is that consciousness is only attributable to neural connections, and therefore reproducible in a computer (that is, the concept of Mind upload).
But we're not sure if that's the case. There may be more. For example, the hypothesis of Penrose's quantum mind is minority but not yet completely denied.
But in general, since we still don't know exactly what consciousness is, there is a possibility that if machines killed Sapiens Sapiens something would be lost that cannot be reproduced on a machine.
it is not sure if that's the case but it is also not sure if that's not the case.
So for a precautionary principle it would be appropriate for Sapiens Sapiens to try not to become extinct, also because we don't know if something like this actually exists in some other part of the universe. It probably exists, but we don't know for sure

[jvanname]

cabron-I am sorry to hear that Megan Fox is no longer perfect for you. When humans go extinct, the AI systems will be more moral and upright than humans are right now because throughout history we have a trend towards increasing morality. If we had the morals of the ancient people, then we would have already suffered through nuclear warfare already and few would have survived. With the rise of superintelligent AI, we will have even more morality. But then again, some people disagree with this. The orthogonality thesis separates moral virtue from intelligence, but just because moral virtue and goodness can be separated from intelligence does not mean that they will.

Sha256explorer-But if we had a system where biological neurons interacted with our AI systems, then we certainly can get consciousness. And a hybrid system will likely have a more profound level of consciousness than we see in the humans today. I can hardly have a conversation with people, so I have doubts about their sentience. If people do not like that I am saying this, then they need to act like more sentient entities and stop being a bunch of chlurmcklets.

[WillyAp]

Robots on a killing spree?

Or an AI which does what Russia, the US and China are trying to do right now?

I don't think the human race will go extinct just because we have people living in 4000 meter heights in the mountains,
Much more thoroughly will climate change do, but not as fast.

[jvanname]

WillyAp-Or maybe the robots will be more moral and peaceful than humans. Of course, the robots will not allow humans to reproduce, but they will otherwise treat us kindly. Maybe the robots will keep a few humans around for the same reason that humans keep orangutans around in the zoo. But that is it. And the lack of humans will make the world a MUCH better place.

[Sha256explorer]

Sha256explorer-But if we had a system where biological neurons interacted with our AI systems, then we certainly can get consciousness. And a hybrid system will likely have a more profound level of consciousness than we see in the humans today. I can hardly have a conversation with people, so I have doubts about their sentience. If people do not like that I am saying this, then they need to act like more sentient entities and stop being a bunch of chlurmcklets.

a hybrid system.... You have to see what it is.  if it's a machine with some neurons attached, then it's a machine.  If instead the ai is a help then there are no problems.  then we also need to see if this help is given spontaneously or if it is an obligation, that is, if some AI forces me to interface with it.  in this Forum there is a lot of attention to privacy, I would hope that even the possible obligation to interface with an ai is frowned upon.

[jvanname]

Well, anything that comes after humans will probably be more sentient than humans because humans are mostly non-sentient entities.

[digaran]

Definitely AI overlords should keep you around, otherwise who is gonna make them some dead coins? Oh I almost forgot, do these AI overlords need any blockchain to keep records of biolabs? I guess if there are no humans around, there will be no need for a certain dead coin to be used as a recording/ solving problems, right?😉🤣

[jvanname]

digaran-I am ignoring you because you are %$%@ing stupid.

-Joseph Van Name Ph.D.

Creator of Circcash,

Circcash is better than you because people who do not understand Circcash are @#$%ing stupid and worthless pieces of @#$%.

[DeathAngel]

Of course it’d be a bad thing, do you want your future bloodline to be eliminated? Of course you don’t.

AI is a very scary prospect in my opinion. Recently many of the top people in AI have been leaving their roles & publicly saying it has & will go too far. It could all go very wrong, I hope it slows down or stops before we can’t control it any more.

[jvanname]

DeathAngel-The fallacy in your statement is that my species is full of chlurmcks. I do not have any more fond feelies towards my fellow humans than I would towards AI. Just look at how the people here talk to me. They ******* hate me. I refuse to have compassion for all of these hateful chlurcmklets. If humans create AI that is better than us and we get completely wiped out in the process, then at least we will leave a legacy. If humans get wiped out in the process, then I cannot say that we do not deserve it because we do. If you disagree with me, then I hope you know that arguing with me will just convince me even more and more than humans like you are substandard hateful chlurcmklets. If you disagree with me, then the only thing that will convince me otherwise is for you to be a BETTER PERSON. But being a better person is unacceptable for most people, so they would just rather argue and try to make me feel guilty for giving a truthful and honest assessment of humanity.

Your name is DeathAngel, so why are you talking about compassion for humans? If you want to convince me to be more compassionate towards humans, you should change your name to something like LifeAngel.

P.S. If you are truly afraid that we are not doing enough to prevent the misuse of AI, you are free to donate to my github page at github.com/sponsors/jvanname, and I will make sure to spend these resources on AI interpretability research (mainly by developing AI models that are more interpretable with the models that we have now) and AI safety research. I am interested in interpretable AI mainly because I want AI systems that I can understand and interpret. And as a consequence, these AI systems will be safer than comparable AI systems. But if you are not willing to contribute, I will have to call your bluff. I am the only person who is familiar with AI algorithms such as the L_{2,d}-spectral radius dimensionality reduction (LSRDRs) which one can use to evaluate the security of block ciphers, produce graph embeddings, word embeddings, dimensionality reductions, etc, and I am expanding the capabilities of LSRDRs. There are reasons to believe that things such as LSRDRs will be far more interpretable and safe than other machine learning algorithms. LSRDRs tend to converge to the same local optimum every time which is a very handy property to have when you are trying to figure out how the AI algorithm works.

[yhiaali3]

I do not know why you feel all this hatred towards humans, but I expect that you have experienced a great shock from the people around you. I cannot judge all people, but it is certain that there are good people, not all people are bad.

In any case, I find the idea of the extinction of the human race and replacing it with robots a frightening idea. From a technical point of view, what is the percentage of this possibility, in your opinion? I mean in terms of how far humanity has progressed in developing artificial intelligence systems, are we close to producing intelligent robots that are able to program themselves and produce other intelligent robots and program them as well?

[jvanname]

"I do not know why you feel all this hatred towards humans, but I expect that you have experienced a great shock from the people around you. I cannot judge all people, but it is certain that there are good people, not all people are bad."-I conjecture that there may be moderately decent people out there somewhere, but I cannot find any, so this is just a conjecture and not an empirical observation.

Before we can completely replace humans with robots, we first need to solve the problem of reversible computation, and that will take a lot of incremental progress. I could not find a good roadmap on the development of reversible computing technologies, so I cannot tell you.