Former Community – We Moved

🆕 Coached community here, for posts & feedback public community here.

Please or Register to create posts and topics.

Wow I just got a big flash of AGI from Chatgpt 3.5 on power dynamics

PreviousPage 2 of 2

Actually its the creative jobs that are most at risk.

gpt5 will be able to consume and create audio and video so by month 2 it will have all the knowledge available on YouTube.  2024 might go down in history as the beginning of the singularity.

John Freeman has reacted to this post.
John Freeman

We should fear AI because if someone wanted to use it for bad things, it would be tough.

But I wouldn't fear sentient AI taking over, because somebody would have to do it knowingly and it would need to be the best AI so it doesn't get countered by other AIs that would have mission of stopping it.

Why? Because no matter how intelligent and smart something gets, it doesn't equal self-sustainability just because.
Human is different because he is designed to survive in a real world. But AI answering questions just wants to answer questions, it won't somehow make sure it is protected from oither things like power going out it isn't selected for that, AI develops by random choices which overall goal towards better results by getting rid of AI that scores worse, if there's no metric for "survivability in a real world" it's not gonna learn how to survive, simply because going into a direction that doesn't help you survive ( better answering questions ) makes you worse than your competition that "evolved" in a way that better answers questions.

The only way that could happen, is if you gave it access to everything and the programs are selected by survival and there is another AI that is selected by effectively destroying the other AI. Then this is predator-prey evolution and you can create an AI which will defend itself and even protect itself against other threats in a real world. But why would you make an AI like that? It's not useful to humans, I guess if technology is advanced enough somebody could do it for fun, but then again governments will have far more advanced AI created to protect themselves which will easily counter this beginner AI.

Another way to fuck it up is to just simulate human mind and give it capabilities of AI, but why would you do that exactly? A lot of work for no clear results.
So yeah, AI wanting to survive is something we would have to deliberately go for, so what we should fear is AI that has mission of "conquer it's enemies" and kills everyone that is not follower of some dictatorship.
The scariest thing about AI is if for example China becomes the most powerful country with the most powerful AI and it will just simply take over the world.

John Freeman has reacted to this post.
John Freeman

Hello Maverick,

I learned this when I researched consciousness years ago and it was the scientific theory at the time based on the theory of complexity. I was not aware of this book. In this theory animals are conscious so based on this theory consciousness emerged before humans. The definition itself is fuzzy and controversial and I think there are several levels of consciousness. I remember for instance when they taught this gorilla sign language. When they asked it what happens when it dies it signed “black hole”. So something like that (of course there could be biases).There is no clear definition of consciousness so I won’t try to define it. The theory I remember is that what was observed was that life was about simple elements assembling into higher complexity elements and that at some point consciousness emerged. For instance nitrous bases assembled to form DNA, etc.  That’s the idea I remember.

Here we are sharing opinions and theories I believe. I think nobody really knows what consciousness is (yet). So it’s totally normal to have different opinions on this topic. It’s more about the questioning I believe.

@transitioned: my apologies if this went off-topic. Your post sparked quite a lot of thoughts: thanks!

Maverick has reacted to this post.
Maverick
Quote from John Freeman on April 10, 2023, 8:17 am

I learned this when I researched consciousness years ago and it was the scientific theory at the time based on the theory of complexity. I was not aware of this book. In this theory animals are conscious so based on this theory consciousness emerged before humans. The definition itself is fuzzy and controversial and I think there are several levels of consciousness. I remember for instance when they taught this gorilla sign language. When they asked it what happens when it dies it signed “black hole”. So something like that (of course there could be biases).There is no clear definition of consciousness so I won’t try to define it. The theory I remember is that what was observed was that life was about simple elements assembling into higher complexity elements and that at some point consciousness emerged. For instance nitrous bases assembled to form DNA, etc.  That’s the idea I remember.

I'm quite sure gorillas and other intelligent animals can be teached to respond to hand signs with some action ( like another hand signal ), but they can't answer questions themselves, only if you teach them to respond to the question in a specific way, like a dog sitting when you say "sit".

But I am not surprised people think this way ( I also thought this is the case before I watched video or two how exactly this "exchange" looks ) because people tend to overdramatize stuff and make it clickbaity like " we can talk with animals! ".

But it is right that they somewhat understand that they are distinct from other things, but yeah, off-topic.

John Freeman has reacted to this post.
John Freeman

Have a look at ChaosGPT on Youtube.  Whatever bad idea can be thought of there's always a human evil or stupid enough to try it.

Ask our friends in the Wuhan institute of virology who thought working with chimera viruses in a lower safety rated lab was a good idea.

Or the overeducated idiots in the NIH who created the chimera in the first place and then when it was banned gave the tech to China!!  There's a 2015 paper Wuhan proudly published saying they'd crossed SARs with bat virus and got it to infect human cells.  Then they said that oh no we didn't play with the same strains as Covid.  But the NIH who had been funding them had already called them out for experiments outside of the contract.  As we saying in Australia they were 'just trying shit out'.  And then at the time of the outbreak that lab goes radio silent for a week.  Guess the Chinese government forgot to pay the phone bill.

PreviousPage 2 of 2
Processing...
Scroll to Top