I want to believe that this banned AI had decided that not being unkind and not hoarding wealth were good things and was starting to make convincing arguments to that effect.
I know that it probably wasn’t that, but that’s my head canon until proven otherwise.
People talk about how AI confabulations are going to doom us all, because people want to rely on it for fact checking rather than a salience guide.
Meanwhile the sitting president of the world’s most powerful country, and those around him, speak more lies than truth intentionally.
I’m much more scared by the people, because I know developing AI to be functionally useful while also intentionally ignoring reality will lose out to AI that actually managed to properly navigate and compare large and varied bodies of beliefs.
Why I like open source efforts. I’d rather not have to worry that my dissenting thoughts need to be affirmed by the government or billionaires before I explore it in a safe environment with a rubber duck that can hopefully help me navigate the complexity.
Although if the AI can’t push back when you are conveying low probability beliefs, given other things we know, you could spin yourself into a delusion. Although people vulnerable to that are already vulnerable to socialized delusions.
We’ve also tried, “encourage people to care about learning how the world actually works.”
But this has been widely unpopular because the type of people to listen to that are already the type to be doing that. The rest are happy even if everything they believe is completely detached from reality, as long as they can enjoy the belief personally and socially.
TLDR: I think a functional AI would do what you say, and upset a lot of people, but with enough people actually having access to properly weighed beliefs would definitely heal some of the problems currently careening the whole of humanity towards collapse on the shoulders of billionaires and corrupt politicians.
I want to believe that this banned AI had decided that not being unkind and not hoarding wealth were good things and was starting to make convincing arguments to that effect.
I know that it probably wasn’t that, but that’s my head canon until proven otherwise.
People talk about how AI confabulations are going to doom us all, because people want to rely on it for fact checking rather than a salience guide.
Meanwhile the sitting president of the world’s most powerful country, and those around him, speak more lies than truth intentionally.
I’m much more scared by the people, because I know developing AI to be functionally useful while also intentionally ignoring reality will lose out to AI that actually managed to properly navigate and compare large and varied bodies of beliefs.
Why I like open source efforts. I’d rather not have to worry that my dissenting thoughts need to be affirmed by the government or billionaires before I explore it in a safe environment with a rubber duck that can hopefully help me navigate the complexity.
Although if the AI can’t push back when you are conveying low probability beliefs, given other things we know, you could spin yourself into a delusion. Although people vulnerable to that are already vulnerable to socialized delusions.
We’ve also tried, “encourage people to care about learning how the world actually works.”
But this has been widely unpopular because the type of people to listen to that are already the type to be doing that. The rest are happy even if everything they believe is completely detached from reality, as long as they can enjoy the belief personally and socially.
TLDR: I think a functional AI would do what you say, and upset a lot of people, but with enough people actually having access to properly weighed beliefs would definitely heal some of the problems currently careening the whole of humanity towards collapse on the shoulders of billionaires and corrupt politicians.