Artificial Intelligence - History Falsification

[feralimal]

I very carefully used the word reasonably.

Fair enough. I can entertain (or ignore) all sorts of wild ideas.

Ancient? Or could they be alt-future AI which somehow can visit and tamper with our past? Or would that be a distinction without a difference, if they are somehow transcending time?

But is this reasonable though? Perhaps you could provide some evidence? If not, why do you dismiss my 'night-vision dinosaur' idea? What is the method? Is it whether it's been in a film, or is there more to it?

Did these elites not openly obsess about Pedigree Bullsh*t for centuries?

Did they?

Do you think they stopped, or do you think they just got really shy about it because the inherent eugenics component revealed too much about what vile scumbags they are?

Are they vile scumbags?

Was the world not dominated by a relatively tiny set of dynasties?

Was it?

Is that not part of the historical record?

Is the historical record true, false or 'true and false' though? If 'true and false', how to discern the truth - what's your technique?

Your contention is that he is absurdly sloppy about "identifying" family tree shenanigans. I agreed.

Is it? Maybe you have me mixed up with another commenter?

At this point in time, some time travellers could theoretically revisit all the worst tragedies in history and replace the actual events with a staged production and false newspaper stories, and IT WOULD MAKE NO DIFFERENCE TO US, to us now. Would it?

Us? Speak for yourself. The way I see it, each individual needs to work things out for themselves. Those individuals have to base themselves on something of course and that could be the MSM, an alternative narrative, their own experience, what their mum told them, or whatever. But each individual gets the opportunity to exercise their free will over whether they 'choose to believe' or 'try to know'.

If you are in the business of expressing an alternative narrative eg 'bloodline time travelling ais killed my granny', do you think you could provide some evidence please? Do we need another story? I like my my 'night-vision dinosaur' idea more - your story doesn't get my vote.

If you do intend to push on with this narrative, can I point out a fundamental flaw? How do reconcile "them" being vile etc and having time travel abilities to create theatrics? For all you know, they may be averting a bigger disaster, so could be benevolent, no? After all, no one is dying with all these theatrics!

 

[TooTrue]

Fair enough. I can entertain (or ignore) all sorts of wild ideas.


But is this reasonable though? Perhaps you could provide some evidence? If not, why do you dismiss my 'night-vision dinosaur' idea? What is the method? Is it whether it's been in a film, or is there more to it?


Did they?


Are they vile scumbags?


Was it?


Is the historical record true, false or 'true and false' though? If 'true and false', how to discern the truth - what's your technique?


Is it? Maybe you have me mixed up with another commenter?


Us? Speak for yourself. The way I see it, each individual needs to work things out for themselves. Those individuals have to base themselves on something of course and that could be the MSM, an alternative narrative, their own experience, what their mum told them, or whatever. But each individual gets the opportunity to exercise their free will over whether they 'choose to believe' or 'try to know'.

If you are in the business of expressing an alternative narrative eg 'bloodline time travelling ais killed my granny', do you think you could provide some evidence please? Do we need another story? I like my my 'night-vision dinosaur' idea more - your story doesn't get my vote.

If you do intend to push on with this narrative, can I point out a fundamental flaw? How do reconcile "them" being vile etc and having time travel abilities to create theatrics? For all you know, they may be averting a bigger disaster, so could be benevolent, no? After all, no one is dying with all these theatrics!

There'd be time-travelling rivals.

Just like the TV show Timeless, I guess.

You are correct, some must be benevolent.

Or else we'd already all be dead.

Here's my latest outlook in a memetic nutshell:

 

[push4more]

Why are the hands always SO bad in AI images? If there were any kind of intelligence behind it, it would have learned by now how human hands work, but they are universally terribly presented. Extra fingers, missing fingers, wrong positions, weird deformities, sometimes no hands at all, there's no way you can mistake an AI image for a real photograph if there are hands involved. Maybe the people behind AI are doing that on purpose. That way, when the hands do start to look perfect, we really won't be able to tell the fake from the real. Or maybe they're just having a laugh.

Most AI's can't count and they are just ingesting countless pictures then hands interpolating - but still not enough hands.

View: https://www.youtube.com/watch?v=24yjRbBah3w
. Google has a new big AI engine called Gemini so for sure AI's will be getting counting skills.

Why are the hands always SO bad in AI images? If there were any kind of intelligence behind it, it would have learned by now how human hands work, but they are universally terribly presented. Extra fingers, missing fingers, wrong positions, weird deformities, sometimes no hands at all, there's no way you can mistake an AI image for a real photograph if there are hands involved. Maybe the people behind AI are doing that on purpose. That way, when the hands do start to look perfect, we really won't be able to tell the fake from the real. Or maybe they're just having a laugh.

 

[TooTrue]

Most AI's can't count and they are just ingesting countless pictures then hands interpolating - but still not enough hands.

View: https://www.youtube.com/watch?v=24yjRbBah3w
. Google has a new big AI engine called Gemini so for sure AI's will be getting counting skills.

How is it possible that an AI can write a successful essay but not be able to count? (I say this as someone who can write a successful essay but I typically use my fingers to count like a kindergartner.)

 

[push4more]

Ancient? Or could they be alt-future AI which somehow can visit and tamper with our past? Or would that be a distinction without a difference, if they are somehow transcending time? These might seem to most people, even most conspiracy theorists, like bonkers speculations, seen-too-many-movies territory, but I think it might actually be sufficiently real, some semi-loop framework like that. Being inundated with movie/TV/novel narratives about various related subtopics might be part of the scheme...or a counter-scheme, I dunno. If we are up against the literal equivalent of Team Skynet, then there's probably also a Team John Connor battling them on the same spacetime turf doing the same things but with the opposite valence.

TLDR: Is Skynet basically an actual thing already?

How is it possible that an AI can write a successful essay but not be able to count? (I say this as someone who can write a successful essay but I typically use my fingers to count like a kindergartner.)

Here is a long one but the best article I have seen about AI. AI are prediction/probability engines. My guess is hands are analogous to many human metaphors and similes but AI are currently not smart to 'read' those constructs yet. So hands are quite complicated compared to faces especially when fingers/hands are rotated in thousands of different angles. AI has not perused tens of thousands of hand images yet.

"It’s Just Adding One Word at a Time"​

What Is ChatGPT Doing … and Why Does It Work?

​

I sent in to ChatGPT with some information and a common English phrase: "some other time maybe". It was clueless but it did a pretty good job to intelligently misread it and weaved it into a paragraph. I then sent in a classical definition of the phrase and it still misread it.​

Google has new Gemini AI so perhaps hands will become much better.

 

[greatscot]

relanguaging after the fall of Babel was probably done with "AI," they could just have bots spam in a million different made-up languages until humans couldn't understand one another

 

[AntiSoof]

Maybe we should ask AI good questions?
It can answer many things; is the earth a globe, prove it, who are the most powerful families in the world, prove, is E=MC^2 meaningful, why not, do AI see any anomalies in historical events, prove it, is AI real and give evidence, prove it?
The best questions gives the best answers.
You don't ask a baker about greens.

 

[jeenam]

Maybe we should ask AI good questions?

There's no point. All AI does is go out and scrape the web to "find answers". It's the same thing as doing a manual search of the web, the only difference being that "AI" does searches in parallel and then attempts to determine the "best answer" based on search results. There's no intelligence behind it in that it can't actually reason. It's just using pre-determined algorithms to guess at the best reply based on the number of matches it gets from search results.

In other words, AI is incapable of reasoning like a human being. There is zero creativity and zero ability to propose new conceptual ideas because the only thing it has to work with is whatever data it can already obtain. I've worked in the computer field for over 20 years and AI is an absolute joke of a technology at this point. And when I say I've worked with computers for 20+ years, the systems I've help build and design handle billions of transactions per day, including mission critical 24/7/365 platforms.

In simplest terms, AI at this point is no more than a fancy hyped-up statistical analysis model. The "loudest" search results win. In other words, whatever search result comes out on top with the highest number of matches wins. That's all AI is doing at the current level of capabilities.

 

[feralimal]

There's no point. All AI does is go out and scrape the web to "find answers". It's the same thing as doing a manual search of the web, the only difference being that "AI" does searches in parallel and then attempts to determine the "best answer" based on search results. There's no intelligence behind it in that it can't actually reason. It's just using pre-determined algorithms to guess at the best reply based on the number of matches it gets from search results.

In other words, AI is incapable of reasoning like a human being. There is zero creativity and zero ability to propose new conceptual ideas because the only thing it has to work with is whatever data it can already obtain. I've worked in the computer field for over 20 years and AI is an absolute joke of a technology at this point. And when I say I've worked with computers for 20+ years, the systems I've help build and design handle billions of transactions per day, including mission critical 24/7/365 platforms.

In simplest terms, AI at this point is no more than a fancy hyped-up statistical analysis model. The "loudest" search results win. In other words, whatever search result comes out on top with the highest number of matches wins. That's all AI is doing at the current level of capabilities.

... but useful to find out what 'everyone' thinks. Or ought to think.

 

[AntiSoof]

There's no point. All AI does is go out and scrape the web to "find answers". It's the same thing as doing a manual search of the web, the only difference being that "AI" does searches in parallel and then attempts to determine the "best answer" based on search results. There's no intelligence behind it in that it can't actually reason. It's just using pre-determined algorithms to guess at the best reply based on the number of matches it gets from search results.

In other words, AI is incapable of reasoning like a human being. There is zero creativity and zero ability to propose new conceptual ideas because the only thing it has to work with is whatever data it can already obtain. I've worked in the computer field for over 20 years and AI is an absolute joke of a technology at this point. And when I say I've worked with computers for 20+ years, the systems I've help build and design handle billions of transactions per day, including mission critical 24/7/365 platforms.

In simplest terms, AI at this point is no more than a fancy hyped-up statistical analysis model. The "loudest" search results win. In other words, whatever search result comes out on top with the highest number of matches wins. That's all AI is doing at the current level of capabilities.

jeenam,

It's pointless? You think so because you come from IT?

I also have an affinity for IT, and I have a different opinion than you.

If people are looking for answers about the world, I recommend using AI.
Right now, it's still relatively harmless, and certain books are still available online, but for how much longer?

 

[jeenam]

jeenam,

It's pointless? You think so because you come from IT?

I also have an affinity for IT, and I have a different opinion than you.

If people are looking for answers about the world, I recommend using AI.
Right now, it's still relatively harmless, and certain books are still available online, but for how much longer?

The point I was trying to make was there is no way to ask "good questions" to try and induce AI to provide better/different answers. AI can't "think". It's just a fancy data scraper. People shouldn't get the impression there's any real intelligence behind AI akin to how a human mind thinks. That's what needs to be conveyed to the general population so they don't get the wrong ideas about what AI is capable of.

If I'm being lazy and don't want to take the time to look through a few websites to confirm specifications of some device, then sure, I'll trust AI to go out and return the info I'm looking for and have a reasonable expectation that it will provide the correct information. It's easier than sifting through a bunch of websites to verify that the first one or two websites aren't providing the wrong info. However, I have zero confidence based on my knowledge of how the current technology works that it can provide anything meaningful beyond regurgitating whatever it can scrape from the web.


Here's a theroretical example...

Prior to advent of what is now referred to as AI, folks had been working toward building this technology with one of the prime intents being to defeat a human opponent in chess. Most people have probably heard of IBM's Deep Blue. IIRC it beat Kasporov in chess. What it was programmed to do, and really what AI is essentially doing, was cycling through a preprogrammed history of all of Kasporov's games previously played against all of his opponents. The AI also had a preprogrammed knowledge of the exact move history of a gigantic catalogue of chess games played by other highly ranked chess players. That's the foundation of knowledge that it was relying upon. What it couldn't factor in when determining which moves to make was the conditions those players were playing under. For example, if they were under duress. They could have preprogrammed in some timers to try and assert if the opponent was seemingful frustrated based on the time it was taking for them to make their next move, but that's no true guarantee of measuring duress. The opponent could have intentionally been delaying their moves until the time limit for the next move was up for any number of reasons beyond that of duress. Without the ability to look an opponent in the eye and perceive their exact intentions and mental state assessing, an opponent requires a level of intelligence that is beyond anything AI was and is currently capable of. People need to realize this technology is still in its infancy and it isn't even close to a human level of intelligence beyond quick regurgitation of whatever statements it can or has already been fed.

This is the same reason the folks who try and predict the weather can never get it right. Do a quick search on supercomputers and you'll see that a big proportion of them are used to simulate weather patterns. Well color me impressed because even though the weather predictors wield gigantic clusters of systems that make my desktop PC look like the equivalent of a hot wheels car compared to a McLaren F1, they still can't tell me with 100% accuracy whether or not it's gonna rain today. It is because of the limitations of the data that contributes to the factors used in making those simulations simply doesn't account for enough information in a manner that allows for precision accuracy that they still get it wrong.

 

[AntiSoof]

The point I was trying to make was there is no way to ask "good questions" to try and induce AI to provide better/different answers. AI can't "think". It's just a fancy data scraper. People shouldn't get the impression there's any real intelligence behind AI akin to how a human mind thinks. That's what needs to be conveyed to the general population so they don't get the wrong ideas about what AI is capable of.

If I'm being lazy and don't want to take the time to look through a few websites to confirm specifications of some device, then sure, I'll trust AI to go out and return the info I'm looking for and have a reasonable expectation that it will provide the correct information. It's easier than sifting through a bunch of websites to verify that the first one or two websites aren't providing the wrong info. However, I have zero confidence based on my knowledge of how the current technology works that it can provide anything meaningful beyond regurgitating whatever it can scrape from the web.


Here's a theroretical example...

Prior to advent of what is now referred to as AI, folks had been working toward building this technology with one of the prime intents being to defeat a human opponent in chess. Most people have probably heard of IBM's Deep Blue. IIRC it beat Kasporov in chess. What it was programmed to do, and really what AI is essentially doing, was cycling through a preprogrammed history of all of Kasporov's games previously played against all of his opponents. The AI also had a preprogrammed knowledge of the exact move history of a gigantic catalogue of chess games played by other highly ranked chess players. That's the foundation of knowledge that it was relying upon. What it couldn't factor in when determining which moves to make was the conditions those players were playing under. For example, if they were under duress. They could have preprogrammed in some timers to try and assert if the opponent was seemingful frustrated based on the time it was taking for them to make their next move, but that's no true guarantee of measuring duress. The opponent could have intentionally been delaying their moves until the time limit for the next move was up for any number of reasons beyond that of duress. Without the ability to look an opponent in the eye and perceive their exact intentions and mental state assessing, an opponent requires a level of intelligence that is beyond anything AI was and is currently capable of. People need to realize this technology is still in its infancy and it isn't even close to a human level of intelligence beyond quick regurgitation of whatever statements it can or has already been fed.

This is the same reason the folks who try and predict the weather can never get it right. Do a quick search on supercomputers and you'll see that a big proportion of them are used to simulate weather patterns. Well color me impressed because even though the weather predictors wield gigantic clusters of systems that make my desktop PC look like the equivalent of a hot wheels car compared to a McLaren F1, they still can't tell me with 100% accuracy whether or not it's gonna rain today. It is because of the limitations of the data that contributes to the factors used in making those simulations simply doesn't account for enough information in a manner that allows for precision accuracy that they still get it wrong.

jeenam,
I agree with all of that, of course, and it all makes sense. And yes, I warn people I know to be very careful with it, even to avoid it like the plague.
But here, on this site, there are people with knowledge and experience, and they could use it to get a sense of direction. If we have to work through a billion books ourselves, that's not possible, so you could let AI do it.
That's all. And especially because AI hasn't been made so shielded and politically correct yet, we shouldn't wait too long if we want to use it cautiously. But, okay, it could be devilishly dangerous, and it could also make us a bit stupid and a bit of a thinker. So, I agree.