How our possible simulation could work

2PhiloVoid · Monday at 7:02 AM

JohnClay said:
(thread was originally called Evidence for being in a possible simulation)

I think the possible simulation could use generative AI rather than using the old method of simulating every atom in the universe. Note that using the old method to simulate the Sun would involve 10^57 atoms which is a 1 with 57 zeroes like this: 1,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000.

Early versions of the kind of technology that could be used: (see the first minute)

It could use generative AI similar to this from 9 months ago which can generate pretty realistic characters:

An AI from 2 years ago that could generate super realistic voices and laughs in realtime - and it can see through his phone's camera:

I am a fan of this old video of Elon Musk:

Welcome, Life's Player

The possibility of being in a simulation, video game, or matrix...

lifesplayer.com

So I think there would be billions of cheap simulations in the future - using generative AI to generate the sounds, visuals, and other sensory information. It would usually involve one player and the rest of the characters would be AI generated but I just assume everyone is real anyway. Elon Musk thinks it would take less than 10,000 years for technology to get to that point.

A video about this scenario can be found by looking for rick morty roy game on youtube:

https://www.youtube.com/results?search_query=rick+morty+roy+game

Note that Rick and Morty video contains swearing. It also runs faster than real time so it takes minutes or hours to simulate a life of many decades. There is also another scenario where you begin with god like powers (easy to do in a simulation) and eventually decide to live a normal life where you forget your original identity.

1.2.3. - Mythology of Hinduism - Pt. 1

I wonder, I wonder, what you would do if you had the power to dream at night, any dream you wanted to dream. And you would of course be able to alter your time-sense and slip, say seventy-five years of subjective time into eight hours of sleep. You would, I suppose, start out by []

web.archive.org

Note to be "indistinguishable from reality" like Elon Musk says, it would be impossible to 100% prove that it is a simulation.

Unfortunately, John Elon Musk has Aspergergs and despite the fact that he is very smart, I don't think he reads much of the analytic philosophy which challenges his flimsy claims that "it would be impossible to 100% prove that it is a simulation." He's not thinking things through, which is odd for a guy who claims to start with "first principles."

I'm sorry, John. I'm not trying to give you a bad day, but all this referencing to the reification of the supposed nature of absolute simulations is, for the most part, utterly missing reference to the deeper considerations that come with with the robust study of Epistemology and Analytic Philosophy.

It's time for people to wake up and step off of the Techno-Train being pushed by the Transhumanist elites.

Warden_of_the_Storm · Monday at 7:04 AM

AV1611VET said:
Isn't that their goal?

AV1611VET said:
If that's their goal, and they're passionate about building a better mousetrap, then it may just very well happen.

Who are they? And what do they want and why?

AV1611VET · Monday at 7:09 AM

Warden_of_the_Storm said:
Who are they?

The creators of virtual reality toys.

In our days, back in the '60s, we had these:

Warden_of_the_Storm said:
And what do they want and why?

Money

Warden_of_the_Storm · Monday at 7:16 AM

AV1611VET said:
The creators of virtual reality toys.

View attachment 381382

In our days, back in the '60s, we had these:

View attachment 381383

Money

I'll actually give you that one, AV. That is a substantially better answer than most I've seen for this sort of talk.

And I had one of the view-finders when I was younger in the late 90s, so they were going strong for a LONG time.

2PhiloVoid · Monday at 7:17 AM

Warden_of_the_Storm said:
I'll actually give you that one, AV. That is a substantially better answer than most I've seen for this sort of talk.

And I had one of the view-finders when I was younger in the late 90s, so they were going strong for a LONG time.

Me too. @AV1611VET ^_^

AV1611VET · Monday at 7:24 AM

Warden_of_the_Storm said:
And I had one of the view-finders when I was younger in the late 90s, so they were going strong for a LONG time.

You're into retail, right?

Do they sell those View Masters today?

Warden_of_the_Storm · Monday at 7:25 AM

AV1611VET said:
You're into retail, right?

Do they sell those View Masters today?

I've seen it sold but not at the store I work at. I've seen a few in museum gift-shops and the like, same with zoos and such.

JohnClay · Monday at 4:40 PM

Warden_of_the_Storm said:
A fake video is not a simulation.

A fake video is a step towards a simulation - where computer graphics started off with images, then video, then video games... a video game is a form of simulation. So compare old computer rendered images with this upcoming video game - which can be played on a machine costing less than $1000.

Earlier in post #57 you said:
I honestly see no reason to even entertain the idea that AI can be used to create a simulation that would be faultless to such a degree that it could be mistaken for the real thing

Well that image looks pretty close. Do you think games will never get to the point that some people think they look real?

Warden_of_the_Storm said:
I am keeping to the original letter of the OP where you specifically mention a fake reality, a fake world, like a fake sun. That's the sort of simulation I'm talking about here.

I'm saying it would be possible in the future.

Warden_of_the_Storm said:
And yet it's being treated as something to be worshipped by those who claim it can do things like create a robot uprising or a simulation.

If they really did worship it wouldn't that mean they'd obey whatever the AI told them?

Warden_of_the_Storm said:
That in no way at all answers my question.

So you can't work out why they'd play the Roy game? Maybe for personal growth or entertainment.

JohnClay · Monday at 5:03 PM

2PhiloVoid said:
Unfortunately, John Elon Musk has Aspergergs and despite the fact that he is very smart, I don't think he reads much of the analytic philosophy which challenges his flimsy claims that "it would be impossible to 100% prove that it is a simulation."

Do you think it is possible to one day create a simulation that is indistinguishable from reality? If it is indistinguishable from reality then it can't be proven it is a simulation because you can't prove base reality is a simulation because it isn't. It is more immersive if it is indistinguishable from reality. If this is a simulation I believe that there is no way for me to prove that it is - so it seems like reality.
Note it isn't really possible for people in a simulation to compare it to base reality unless they had access to information about it.

2PhiloVoid · Monday at 5:27 PM

JohnClay said:
Do you think it is possible to one day create a simulation that is indistinguishable from reality?

Not really. No.

JohnClay said:
If it is indistinguishable from reality then it can't be proven it is a simulation because you can't prove base reality is a simulation because it isn't. It is more immersive if it is indistinguishable from reality. If this is a simulation I believe that there is no way for me to prove that it is - so it seems like reality.

I'm going to stick with Hillary Putnam's suggested analytic answer to your question, which in a nutshell is a semantic refutation of Global Skepticism. That, and the fact that I think materially, there are physical problems with entrapping people in a complete way:

https://bigthink.com/13-8/the-key-problem-with-the-brain-in-a-vat-thought-experiment/

My ethical question is why would we support anyone who wants to create technology to permanently place other people into an unnatural Matrix?

AV1611VET · Monday at 5:28 PM

JohnClay said:
Do you think it is possible to one day create a simulation that is indistinguishable from reality?

No.

People tried that with LSD and got nowhere.

JohnClay said:
If it is indistinguishable from reality then it can't be proven it is a simulation because you can't prove base reality is a simulation because it isn't.

I can prove it.

Just turn the monitor around -- or take the helmet off -- and you'll literally get a reality check.

JohnClay said:
It is more immersive if it is indistinguishable from reality.

Until you either die of thirst, starvation, or the electricity goes off.

JohnClay said:
If this is a simulation I believe that there is no way for me to prove that it is - so it seems like reality.

You have been proving simulation isn't real since October 27, 2006.

Else you would have died almost twenty years ago.

Warden_of_the_Storm · Monday at 5:30 PM

JohnClay said:
A fake video is a step towards a simulation - where computer graphics started off with images, then video, then video games... a video game is a form of simulation. So compare old computer rendered images with this upcoming video game - which can be played on a machine costing less than $1000.

JohnClay said:
Well that image looks pretty close. Do you think games will never get to the point that some people think they look real?

It looks NOWHERE NEAR close to real. Like, you can tell from the textures of the skin that it's a video game. The surf of the water is absolutely flat too.

JohnClay said:
I'm saying it would be possible in the future.

Saying isn't showing.

JohnClay said:
If they really did worship it wouldn't that mean they'd obey whatever the AI told them?

Some people have committed suicide because AI told them to. Now that is obviously not worship but you can't deny that is worrying.

JohnClay said:
So you can't work out why they'd play the Roy game? Maybe for personal growth or entertainment.

I think that using such a game would just be an escape from life and an avoidance of real world issues. And I also do not care for Rick & Morty. I think it's an overhyped show where someone's attempt at being mature is being how a young teenager imagines an adult is.

Warden_of_the_Storm · Monday at 5:34 PM

AV1611VET said:
No.

People tried that with LSD and got nowhere.

Actually, LSD was created with the intention of helping people with breathing problems. The creator just happened to accidentally ingest some of it and then found out it does not help with breathing problems at all.

JohnClay · Monday at 5:46 PM

Warden_of_the_Storm said:
It looks NOWHERE NEAR close to real. Like, you can tell from the textures of the skin that it's a video game. The surf of the water is absolutely flat too.

But it is much more realistic than the original Toy Story movie which took 45 minutes to 20 hours per frame using $7-12 million worth of computers. GTA 6 takes about 0.015 seconds per frame (60fps) and the PS5 costs less than $1000. Do you think that game is at the limit to what is possible in computer graphics? What about in 10-20 years time? Note the PS6 is coming out fairly soon and that would make even more impressive games possible. Then PS7 etc.

Warden_of_the_Storm said:
Saying isn't showing.

Well I was trying to show that GTA 6 is really impressive and that it is likely that graphics would improve even more in the future. I think that's a reasonable statement.

Warden_of_the_Storm said:
Some people have committed suicide because AI told them to. Now that is obviously not worship but you can't deny that is worrying.

But that involves the AI talking to them. I don't think the AI would be talking to people in the simulation or if they did people would assusme they're mentally ill. Or it could be generating the speech of the characters though most people would think it is the people rather than an AI doing that.

Warden_of_the_Storm said:
I think that using such a game would just be an escape from life and an avoidance of real world issues.

Sometimes a game is played to seek problems especially rage games that are designed to frustrate people as much as possible. So you answered your own question...

Warden_of_the_Storm · Monday at 5:53 PM

JohnClay said:
But it is much more realistic than the original Toy Story movie and that took up to 20-30 hours per frame using 110 workstations. That game takes about 0.015 seconds per frame (60fps). The equipment used in the Toy Story movie would have costed around $7-12 million whereas the PS5 is under $1000. Do you think that game is at the limit to what is possible in computer graphics? What about in 10-20 years time?

JohnClay said:
Well I was trying to show that GTA 6 is really impressive and that it is likely that graphics would improve even more in the future. I think that's a reasonable statement.

Of course the graphics will be able to massive improve in 10 to 20 years. That's not unreasonable. But you seem to be ignoring the Uncanny Valley effect that will always be present in things like video games, and that will always let people tell that it's a video game.

JohnClay said:
But that involves the AI talking to them. I don't think the AI would be talking to people in the situation or if they did people would assusme they're mentally ill.

The people who committed suicide did so because AI chat programs told them to commit suicide. And other acts too:
Deaths linked to chatbots

JohnClay said:
Sometimes a game is played to seek problems especially rage games that are designed to frustrate people as much as possible.

That is true. But since your idea is from Rick & Morty, I ultimately discard it because I do not care for that cartoon in the slightest.

JohnClay · Monday at 6:16 PM

2PhiloVoid said:
My ethical question is why would we support anyone who wants to create technology to permanently place other people into an unnatural Matrix?

I think it could be like the Roy game or Alan Watt's dream thought experiment (see OP) where the player chose to live a simulated life like this and forgot about that choice. Elon Musk is talking about billions of "set top boxes" so the owner of them would be choosing to play that game. Rather than being like the Matrix where billions of people were born into the simulation.

JohnClay · Monday at 6:29 PM

AV1611VET said:
I can prove it.

Just turn the monitor around -- or take the helmet off -- and you'll literally get a reality check.

Your muscles would be paralysed in a brain connected simulation like they are for dreams and sleep paralysis - except be more permanent.

AV1611VET said:
Until you either die of thirst, starvation, or the electricity goes off.

In the OP I talked about scenerios where it only takes several minutes or hours to simulate decades of a life. That would determine how realistic and expensive the simulation is. Maybe a "serious" simulation like this would take longer if people want it to be more realistic.

AV1611VET said:
You have been proving simulation isn't real since October 27, 2006.

Else you would have died almost twenty years ago.

I don't understand what you mean. Why would I have died almost 20 years ago?

JohnClay · Monday at 6:48 PM

Warden_of_the_Storm said:
Of course the graphics will be able to massive improve in 10 to 20 years. That's not unreasonable. But you seem to be ignoring the Uncanny Valley effect that will always be present in things like video games, and that will always let people tell that it's a video game.

It looks like in the future many video games would get their graphics/speech/behaviour from generative AI (e.g. the first video in the OP). Like I said I was fooled by a Bernie Sanders AI video and in the future games like in the first video would be able to replicate those kinds of graphics and speech - and you'd be able to talk to them. So I disagree that it "will always let people tell that it's a video game".
Note the ONLY reason I thought the video of Bernie Sanders was created by AI was because he said "A few hours ago, my office received a letter from a one zero year-old boy named Matteo in Wisconsin".

Warden_of_the_Storm said:
The people who committed suicide did so because AI chat programs told them to commit suicide. And other acts too:
Deaths linked to chatbots

Yeah that is scary - though many AIs are supposed to be designed to be "safe" and refuse to do certain things.

2PhiloVoid · Monday at 7:30 PM

JohnClay said:
(thread was originally called Evidence for being in a possible simulation)

I think the possible simulation could use generative AI rather than using the old method of simulating every atom in the universe. Note that using the old method to simulate the Sun would involve 10^57 atoms which is a 1 with 57 zeroes like this: 1,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000.

Early versions of the kind of technology that could be used: (see the first minute)

It could use generative AI similar to this from 9 months ago which can generate pretty realistic characters:

An AI from 2 years ago that could generate super realistic voices and laughs in realtime - and it can see through his phone's camera:

I am a fan of this old video of Elon Musk:

Welcome, Life's Player

The possibility of being in a simulation, video game, or matrix...

lifesplayer.com

So I think there would be billions of cheap simulations in the future - using generative AI to generate the sounds, visuals, and other sensory information. It would usually involve one player and the rest of the characters would be AI generated but I just assume everyone is real anyway. Elon Musk thinks it would take less than 10,000 years for technology to get to that point.

A video about this scenario can be found by looking for rick morty roy game on youtube:

https://www.youtube.com/results?search_query=rick+morty+roy+game

Note that Rick and Morty video contains swearing. It also runs faster than real time so it takes minutes or hours to simulate a life of many decades. There is also another scenario where you begin with god like powers (easy to do in a simulation) and eventually decide to live a normal life where you forget your original identity.

1.2.3. - Mythology of Hinduism - Pt. 1

I wonder, I wonder, what you would do if you had the power to dream at night, any dream you wanted to dream. And you would of course be able to alter your time-sense and slip, say seventy-five years of subjective time into eight hours of sleep. You would, I suppose, start out by []

web.archive.org

Note to be "indistinguishable from reality" like Elon Musk says, it would be impossible to 100% prove that it is a simulation.

I have to love the fact that the A.I. remembers that, as the first video says, "dog's have b-holes." How quaint! Just what I need. :ahah:

sjastro · Monday at 7:52 PM

When AI systems become less dependent on human-supervised learning and increasingly learn through self-supervised or reinforcement-learning methods, they can discover solutions that are not constrained by conventional human reasoning. In closed domains such as chess, certain field of mathematics and in producing computer algorithms this has already resulted in superhuman performance.

This was demonstrated by AlphaZero in chess.

Bridging the Human-AI Knowledge Gap: Concept Discovery and Transfer in AlphaZero
Artificial Intelligence (AI) systems have made remarkable progress, attaining super-human performance across various domains. This presents us with an opportunity to further human knowledge and improve human expert performance by leveraging the hidden knowledge encoded within these highly...

arxiv.org

Here is an outline of the paper.

How our possible simulation could work

Smash'em!!!

Well-Known Member

SCIENCE CAN TAKE A HIKE

Well-Known Member

Smash'em!!!

SCIENCE CAN TAKE A HIKE

Well-Known Member

Married Mouth-Breather

Married Mouth-Breather

Smash'em!!!

SCIENCE CAN TAKE A HIKE

Well-Known Member

Well-Known Member

Married Mouth-Breather

Well-Known Member

Married Mouth-Breather

Married Mouth-Breather

Married Mouth-Breather

Smash'em!!!

Newbie

Bridging the Human-AI Knowledge Gap: Concept Discovery and Transfer in AlphaZero​

Similar threads

Bridging the Human-AI Knowledge Gap: Concept Discovery and Transfer in AlphaZero