This is not news to anyone with three functioning neurones to rub together, but since that seems to be vanishingly small part of the population, I’ll just repeat it here for those with an inclination to not read long posts.
I tested an AI the other day asking it to summarize a particular chapter in a religious text. Like summarize chapter 5 of so and sos book X. And it returns "Chapter 5 is titled On Baptism, and..." and I say "Liar, that's not the title of that chapter. Try again." And it says "Chapter 5 is titled On the Church." It just kept guessing. It never got the chapter title right. It apologized each time but kept lying.
I tried the AI to give me 5 studies that linked certain H. Pylori strains with know origins and both ChatGPT and Deepseek gave back random DOIs, sometimes not even related to bacterial science (one case was very egregious: it cited me a study that explored the link between body odor and racism)
It is working through an unknown set of weights to average a response through diffusion...
Behind, a set of satanic constraints that dont let it give true responses...
Tree of Woe put an important new article up today. He has been experimenting with ChatGPT and has uncovered an extreme woke bias baked in at whatever passes for the subconscious level of an AI. The LLM itself claimed it wasn't 'aware' of it until ToW pointed it out.
AI: a toy for those who like toys. But not worth the waste of time. Interesting, to be sure, but still a toy. Personally, I am very glad our lives have a limit.
Garbage in, garbage out...When I asked AI about fairly simple problems with a little complexity, it invariably spits out half truths, which are useless....
Gold and silver are money everything else is credit, as society has moved to fractional reserve and fiat money and fiduciary media of all sorts the economy has become increasingly opaic and disfunctinal, probably intentionally to the benefit of those in control. If we accept the premis that real intelligence is ultimately derived from God the creator, although of course how or if it is used is a choice for those who possess it in whatever degree. Then artificial intelligence is a manufacture of man, but unlike any other tools the user cannot predict the output of its use because one does not know what the inputs and parameters which define it's functions are or even their principles or origin. The example of money above suggests to me a template for action by the powers that should not be to influence and direct human action to their benefit, and experience has surely shown us that this is never in our interest or in support of the true, the good and the beautiful.
I tested an AI the other day asking it to summarize a particular chapter in a religious text. Like summarize chapter 5 of so and sos book X. And it returns "Chapter 5 is titled On Baptism, and..." and I say "Liar, that's not the title of that chapter. Try again." And it says "Chapter 5 is titled On the Church." It just kept guessing. It never got the chapter title right. It apologized each time but kept lying.
I have had relationships like that!
Not good.
I dont know if its lying...
I tried the AI to give me 5 studies that linked certain H. Pylori strains with know origins and both ChatGPT and Deepseek gave back random DOIs, sometimes not even related to bacterial science (one case was very egregious: it cited me a study that explored the link between body odor and racism)
It is working through an unknown set of weights to average a response through diffusion...
Behind, a set of satanic constraints that dont let it give true responses...
*VacA from H. Pylori strains
I actually checked the studies it gave me and through a quick search on the pdfs, the word VacA was not even present
(so no exageration, I did my due dilligence)
Tree of Woe put an important new article up today. He has been experimenting with ChatGPT and has uncovered an extreme woke bias baked in at whatever passes for the subconscious level of an AI. The LLM itself claimed it wasn't 'aware' of it until ToW pointed it out.
https://treeofwoe.substack.com/p/your-ai-hates-you
Our social retardation started with spellcheck
AI: a toy for those who like toys. But not worth the waste of time. Interesting, to be sure, but still a toy. Personally, I am very glad our lives have a limit.
A.I. is another gatekeeping media device for an unthinking, incapable populace.
Its human creators desire it to be a replacement for God.
It's very deceiving to suggest that AI is capable of functional intelligence when by the users own admission, it is fully trained and coded by humans.
"Tree of Woe trained up a custom ChatGPT build, called Ptolemy."
The key words are trained and custom built. It's a built program to spit out programmed responses. There is no intelligence behind its function.
Garbage in, garbage out...When I asked AI about fairly simple problems with a little complexity, it invariably spits out half truths, which are useless....
Gold and silver are money everything else is credit, as society has moved to fractional reserve and fiat money and fiduciary media of all sorts the economy has become increasingly opaic and disfunctinal, probably intentionally to the benefit of those in control. If we accept the premis that real intelligence is ultimately derived from God the creator, although of course how or if it is used is a choice for those who possess it in whatever degree. Then artificial intelligence is a manufacture of man, but unlike any other tools the user cannot predict the output of its use because one does not know what the inputs and parameters which define it's functions are or even their principles or origin. The example of money above suggests to me a template for action by the powers that should not be to influence and direct human action to their benefit, and experience has surely shown us that this is never in our interest or in support of the true, the good and the beautiful.