Why did you select poverty as the variable to optimize? Would you not prefer to optimize happiness?
If so, why does ending poverty, but creating a dystopia, result in more happiness?
This work is a form of submission, an effort to be eclipsed, to seek life in death, to fulfill the evolutionary instinct to die for a more magnificent child, and to be dazzled by the beauty of one's own self-immolation like a moth flying into a flame.
Anthropic does not care about reducing poverty. Anthropic is a technological temple, walled off from the turgid and festering mass of humanity. Its work involves preventing people from accessing its models the way the Vatican wanted to control the people's access to God; with closed weights security, guardrails, encrypted chain of thought, logging, regional bans, and moderators. Its goal is to create AGI by any means necessary.
Your argument is something along the lines of Thatcher's "there is no alternative". It is unfalsifiable.
"There is no alternative" is a lie designed to help people hide from the guilt of bad decisions.
Meaning is not an abstraction, meaning is all there is.
Without meaning there is nothing.
Modernity loathes itself. It is so devoid of meaning and so full of the fear of meaningless death, that it will try to drown itself in its own works, dazzled by its lifeless and immortal creation, the sublime of technology, the holy machine of loving grace.
The 'technique' of technology (Ellul), is to justify its own existence. Like the surveillance state, the demand for "advancement" is self-referential.
And on the pedestal these words appear:
"My name is Ozymandias, king of kings:
Look on my works, ye Mighty, and despair!"
The tech companies do above and beyond what is legally required to make their customers accessible.
Prior to TCNs Apple was going to scan icloud anyway. It does indicate at least some integrity that they changed plans and refused to decrypt ADP.
Encryption is legal in many countries. Not in the UK of course. But in countries where encryption is legal, it is the decision of OpenAI, Microsoft etc, to keep data unencrypted. Certainly this is very bad for the user.
Historically Apple has not gone above and beyond. They have done the opposite and actively worked to close multiple loopholes that governments have used and could in the future have used to demand access to customer data.
Apple is commendable for that. You can surely see the point I am making.
Inference is the main example. The AI companies can implement attestation. They can entirely legally have no knowledge of customer prompts, with minimal infrastructure costs.
They only refuse because they want to go beyond the law, they want to enforce the law themselves and need the data to be accessible.
It is to some extent known that the UK has never respected free speech.
This knowledge is muddled by the custom and practice of the UK government to gaslight the world, by describing the UK as a free and open society and also condemning other countries for censorship and human rights violations.
Hence you get a form of doublethink in the UK, where it is true that there is no free speech but also true that there is free speech.
There is a short string of words I cannot say because I might get into trouble myself, which is now illegal to say in the UK. It has to do with a certain group in the Middle East.
Now, I don't give a shit about that conflict, but the precedent this sets is terrifying.
"Free speech" means you're free to say or think whatever you want, provided it follows their rules. You also have freedom from hate speech, which is not speech. Suppose you don't like sending money to Israel, that's hate speech, they'll free you from it.
reply