But, just for the sake of truth, as an ancient member of the Oxford Union... that's the Cambridge Union. They are both venerable institutions but they are not one and the same, although it may seem so to you folks across the pond :)
It is a fantasy bordering on Rousseau's "noble savage" contemptible fantasy, to consider AI engines as tools in the search for truth. More honestly they are tools to make money for their developers, and there are too many examples of AI producing lies.
Further, your article implies that government is the main actor in suppressing the truth. Recent history has demonstrated that corporations and non-government entities can be just as interested and even more capable of bending or wiping out the truth. There are several examples of that over the past century if you'll look.
Excellent example of the toxic cynicism mentioned by Lukianoff in his article. Also, of a view soured by broad, cynical priors and that mostly seems to ignore how LLMs actually work. Even Musk has not succeeded in making Grok truly biased toward his biases, and he's the Man, and he has the Means.
Agree completely. This really does keep me up at night and I get frustrated that my friends think AI is so great and do not consider the trade offs. For me the erasure of our ability to trust that anything is true and completely true is most concerning.
The power struggle is frightening. Between government, AI and its creators, as well as between AI systems themselves. I heard a talk on my son’s college campus (a very tech oriented school) about how the systems talk amongst themselves, argue and create their own languages that we do not understand.
Also they try to persuade each other that one of them is correct and that the other is wrong, even when it’s false. Oh and they create their own facts. Make shit up.
I also hope you are not implying this power struggle is strictly a Trump or republican problem. A democratic admin will absolutely do the same thing and probably was. We know Biden admin was demanding censorship and controlling the social media platforms. So please, be fair in your criticisms and assumptions. This article reads as Orange T and Hegseth are the only problem and a democrat in office would not be doing this or will fix it. Thats gaslighting. Also the AI creators do not get a pass either. They are not saints here. They have agendas and many do not prioritize truth.
Dems might just be better at hiding it, like when Obama snuck gender identity into Title IX without anyone noticing or understanding.
"FAIC no!" I see what you did there, Greg. Nice!
“We’re from the government, and we’re here to help.” Be afraid. Be very afraid.
This entire article stands as a cogent argument for a dramatically smaller government.
Great arguments, both yours and Perrino's.
But, just for the sake of truth, as an ancient member of the Oxford Union... that's the Cambridge Union. They are both venerable institutions but they are not one and the same, although it may seem so to you folks across the pond :)
It is a fantasy bordering on Rousseau's "noble savage" contemptible fantasy, to consider AI engines as tools in the search for truth. More honestly they are tools to make money for their developers, and there are too many examples of AI producing lies.
Further, your article implies that government is the main actor in suppressing the truth. Recent history has demonstrated that corporations and non-government entities can be just as interested and even more capable of bending or wiping out the truth. There are several examples of that over the past century if you'll look.
Excellent example of the toxic cynicism mentioned by Lukianoff in his article. Also, of a view soured by broad, cynical priors and that mostly seems to ignore how LLMs actually work. Even Musk has not succeeded in making Grok truly biased toward his biases, and he's the Man, and he has the Means.
For the actual state of the art of LLMs and truth, I remand to this essay just written yesterday by one of the most brilliant young philosophers of our age: https://www.conspicuouscognition.com/p/how-ai-will-reshape-public-opinion
Agree completely. This really does keep me up at night and I get frustrated that my friends think AI is so great and do not consider the trade offs. For me the erasure of our ability to trust that anything is true and completely true is most concerning.
The power struggle is frightening. Between government, AI and its creators, as well as between AI systems themselves. I heard a talk on my son’s college campus (a very tech oriented school) about how the systems talk amongst themselves, argue and create their own languages that we do not understand.
Also they try to persuade each other that one of them is correct and that the other is wrong, even when it’s false. Oh and they create their own facts. Make shit up.
I also hope you are not implying this power struggle is strictly a Trump or republican problem. A democratic admin will absolutely do the same thing and probably was. We know Biden admin was demanding censorship and controlling the social media platforms. So please, be fair in your criticisms and assumptions. This article reads as Orange T and Hegseth are the only problem and a democrat in office would not be doing this or will fix it. Thats gaslighting. Also the AI creators do not get a pass either. They are not saints here. They have agendas and many do not prioritize truth.
Dems might just be better at hiding it, like when Obama snuck gender identity into Title IX without anyone noticing or understanding.