Do you think any aliens exist in the universe?

[pood]

Please address this.

 

[pood]

As to AI wiping out humanity, I am much more concerned about humanity wiping out humanity, as it is currently doing with vast numbers of species around the world including our closest relatives, the other Great Apes. It is sorrowful and disgusting

 

[excreationist]

I have no idea what those odds are. Why are you asking me?

Please address my analysis of your misapprehension of how to calculate odds,

I'm saying sometimes people come up with odds based on feelings and for things that have a single data point. Those AI researchers are able to guess the odds. What if you had to guess the odds? (if someone had a gun to your head) I thought you would have peferred me to guess the odds based on my beliefs but it seems I should have said "I have no idea what those odds are".

 

[excreationist]

Please address this.

How do you think the AI researchers are coming up with their odds? My belief is it is based on their feelings. I disagree that "your “feelings” are meaningless". I consider feelings to be "intuition" or a "hunch" which I think can come in handy sometimes.

 

[excreationist]

do you have a problem with there being a VR video game with fairly photorealistic graphics (like GTA 6) and characters you could talk to that do real-time text to speech, etc?

Are you asking about being a conscious simulation INSIDE such a video game without knowing it? Yes, I have a problem with that, because no one has even shown it is possible.

No - just exactly what I asked. I think it would be possible with current technology.

That leads onto some evidence. But you didn't respond to the question I asked just something else that I wasn't talking about.

What about this - which of these things do you think is impossible:
- a VR video game with fairly photorealistic graphics (like GTA 6)
- characters tthat you can talk to that do real-time text to speech

 

[excreationist]

I think it is extremely mentally unhealthy to believe you are the sole conscious being in a simulated world that amounts to techno-solipsism.

If a person were to believe this, it may make them cavalier about harming or even killing others, on the reasoning that they aren’t feeling a thing.

Fortunately, in excreationist’s case, the motive seems to be the opposite — to ease the fear that the world is full of actually felt pain and suffering.

There also seems to be some ancillary motivation that being the only one conscious makes him somehow special.

Personally I think I have been more mentally healthy since I've believed my life is like the Roy game (with hope and consequences like a "Roy score") than when I believed I wasn't particularly special and that death would be the end of all of my suffering (like when I was gassing myself in my car). The last time I went to the mental ward was in 2019 and at the time the doctor had a huge problem with my belief that I could probably make more than a million dollars by making websites/ebooks about "cheap easy fatloss" (intermittent fasting combined with Bible passages). When I was first manic in 2000 I thought there was a significant chance that I could be the President of the USA (even though I'm Australian).

 

[pood]

Please address this.

How do you think the AI researchers are coming up with their odds?

Pulling them out of their ass.

 

[pood]

I have no idea what those odds are. Why are you asking me?

Please address my analysis of your misapprehension of how to calculate odds,

I'm saying sometimes people come up with odds based on feelings and for things that have a single data point. Those AI researchers are able to guess the odds. What if you had to guess the odds? (if someone had a gun to your head) I thought you would have peferred me to guess the odds based on my beliefs but it seems I should have said "I have no idea what those odds are".

Right. I have no idea what the odds are. Nor do you. I have explained this.

 

[pood]

do you have a problem with there being a VR video game with fairly photorealistic graphics (like GTA 6) and characters you could talk to that do real-time text to speech, etc?

Are you asking about being a conscious simulation INSIDE such a video game without knowing it? Yes, I have a problem with that, because no one has even shown it is possible.

No - just exactly what I asked. I think it would be possible with current technology.

That leads onto some evidence. But you didn't respond to the question I asked just something else that I wasn't talking about.

What about this - which of these things do you think is impossible:
- a VR video game with fairly photorealistic graphics (like GTA 6)
- characters tthat you can talk to that do real-time text to speech

It appears we have both these things, or something very like them, already?

And?

Myself, I prefer to talk to real people in real life. I don’t play video games or engage with AI or VR. I do post on this message board and sometimes, like right now, I wonder why.

 

[excreationist]

Please address this.

How do you think the AI researchers are coming up with their odds?

Pulling them out of their ass.

But are their probabilities as meaningful as using a roulette wheel to answer the question? I'd say no - I think low probabilities or high probabilities are based on their intuition of how strong the likelihood feels rather than there being no relationship.

 

[excreationist]

What about this - which of these things do you think is impossible:

- a VR video game with fairly photorealistic graphics (like GTA 6)
- characters that you can talk to that do real-time text to speech

It appears we have both these things, or something very like them, already?

And?

Myself, I prefer to talk to real people in real life. I don’t play video games or engage with AI or VR. I do post on this message board and sometimes, like right now, I wonder why.

Well I could propose small changes that go towards a VR simulation with extremely realistic graphics and characters. Maybe there is a point that goes from possible to impossible in your opinion. Perhaps you not playing video games and AI and VR makes you less knowledgeable about them. I guess this argument leads to another dead end.

 

[bilby]

I am saying there is maybe a 50/50 chance this is the case.

"I do not know which of two possibilities, A or B will occur" does NOT imply "there is a 50% chance that A will occur, and a 50% chance that B will occur".

I bought a lottery ticket. It will either win the jackpot, or it won't; Therefore my odds of winning are 50-50. If that were true, we would all be multimillionaires.

 

[Swammerdami]

Although I don't fully endorse excreationist's conclusions, I think some of his comments and arguments are quite reasonable.

I'm saying sometimes people come up with odds based on feelings and for things that have a single data point. Those AI researchers are able to guess the odds. What if you had to guess the odds? (if someone had a gun to your head) I thought you would have peferred me to guess the odds based on my beliefs but it seems I should have said "I have no idea what those odds are".

In many cases it is impossible to calculate probabilities in any normal way. In such cases 50/50 (based on "feelings") is sometimes a valid starting point.

On another matter, I am suspicious of any ontological argument that relies on Gödel's Incompleteness Theorem. Recall that in Gödel's proof he constructs a statement which, far from being undecidable, he demonstrates to be true! It just cannot be proven true within the constraints of a particular formal system:

Mathematics is a game played according to certain simple rules with meaningless marks on paper.

 

[excreationist]

I don’t “believe” any such thing; I cite the paper as strong evidence that an algorithm or algorithms cannot simulate the universe because the argument is that the universe is not algorithmic.

It seems machine learning is different from regular algorithms.

Aspect	Algorithms	Machine learning
Definition	Step-by-step instructions for solving a problem	Data-driven approach where algorithms learn patterns
Input	Explicit rules and logic	Large datasets with examples
Output	Deterministic result (same input → same output)	Predictive model that generalizes to new data
Adaptability	Static, doesn’t change unless reprogrammed	Dynamic, improves with more data
Examples	Sorting algorithms, search algorithms, encryption	Spam filters, recommendation systems, image recognition

These videos show that generative AI can generate realistic-seeming physics but also crazily unrealistic physics (and crazy visuals like the anorexic athletes). It wasn't just programmed in how to do that all step by step (using a physics engine, etc)



It being able to portray crazy scenarios that people can imagine fits perfectly with Alan Watts' dream thought experiment where you can just wish whatever you want. Traditional simulation scenarios just involve what is physically possible.

Algorithms are the foundation of computer science. They’re precise, reliable, and essential for structured tasks.

Machine learning builds on algorithms but adds adaptability. Instead of coding every rule, we let the system discover rules from data. This makes ML powerful for complex, unpredictable problems like speech recognition or fraud detection.

So I'd say that ML can do things that ordinary algorithms have problems with. "complex, unpredictable problems" includes simulating people.

You might wonder what is the point of the AI being able to easily generate things that are physically impossible? Well if you were playing "the Roy game" and you wanted to watch a cartoon there are some possibilities - show them a cartoon that already exists - or generate one on the fly using generative AI - or simulate human minds that are used to generate that cartoon. If you had gone to the studio that generates the cartoon you could change history and so it would be different to what it regularly was.

 

[pood]

I don’t “believe” any such thing; I cite the paper as strong evidence that an algorithm or algorithms cannot simulate the universe because the argument is that the universe is not algorithmic.

It seems machine learning is different from regular algorithms.

Aspect	Algorithms	Machine learning
Definition	Step-by-step instructions for solving a problem	Data-driven approach where algorithms learn patterns
Input	Explicit rules and logic	Large datasets with examples
Output	Deterministic result (same input → same output)	Predictive model that generalizes to new data
Adaptability	Static, doesn’t change unless reprogrammed	Dynamic, improves with more data
Examples	Sorting algorithms, search algorithms, encryption	Spam filters, recommendation systems, image recognition

These videos show that generative AI can generate realistic-seeming physics but also crazily unrealistic physics (and crazy visuals like the anorexic athletes). It wasn't just programmed in how to do that all step by step (using a physics engine, etc)



It being able to portray crazy scenarios that people can imagine fits perfectly with Alan Watts' dream thought experiment where you can just wish whatever you want. Traditional simulation scenarios just involve what is physically possible.

Algorithms are the foundation of computer science. They’re precise, reliable, and essential for structured tasks.

Machine learning builds on algorithms but adds adaptability. Instead of coding every rule, we let the system discover rules from data. This makes ML powerful for complex, unpredictable problems like speech recognition or fraud detection.

So I'd say that ML can do things that ordinary algorithms have problems with. "complex, unpredictable problems" includes simulating people.

You might wonder what is the point of the AI being able to easily generate things that are physically impossible? Well if you were playing "the Roy game" and you wanted to watch a cartoon there are some possibilities - show them a cartoon that already exists - or generate one on the fly using generative AI - or simulate human minds that are used to generate that cartoon. If you had gone to the studio that generates the cartoon you could change history and so it would be different to what it regularly was.

About all this — so what?

 

[pood]

Is ChatGPT conscious?

Interesting article,

 

[WAB]

Off topic, but interesting, is this quote from a skeptic named Emily Bender:

“Our ability to understand other people’s communicative acts is fundamentally about imagining their point of view and then inferring what they intend to communicate from the words they have used,” she wrote in the Guardian.

 

[excreationist]

About all this — so what?

You said "algorithms cannot simulate the universe" but machine learning goes beyond algorithms. ML is "powerful for complex, unpredictable problems" - i.e. it doesn't just simulate physics using explicitly programmed in step by step maths/symbolic-based programming.

 

[pood]

About all this — so what?

You said "algorithms cannot simulate the universe" but machine learning goes beyond algorithms. ML is "powerful for complex, unpredictable problems" - i.e. it doesn't just simulate physics using explicitly programmed in step by step maths/symbolic-based programming.

Even if this granted, what of it? What is the evidence that we live in any kind of simulation or game?

 

[excreationist]

About all this — so what?

You said "algorithms cannot simulate the universe" but machine learning goes beyond algorithms. ML is "powerful for complex, unpredictable problems" - i.e. it doesn't just simulate physics using explicitly programmed in step by step maths/symbolic-based programming.

Even if this granted, what of it? What is the evidence that we live in any kind of simulation or game?

So computer controlled physics in videos based on 3D objects can be called "simulations" - see:

Simulation — Blender

Featuring industry-standards libraries like Bullet and MantaFlow, Blender offers powerful simulation tools.

www.blender.org

Here are some more simulations (using that definition for simulations)

I think it was able to do that by observing and learning from a lot of videos (including the audio), images, and text (from websites and books).

I think in the not too distant future (maybe less than a century) AI will be able to plausibly generate the visuals and audio for just about any prompt.

Though it is a big step to go from realistic visuals and audio to one that directly connects with a consciousness.