Freedom of speech should extend to AIs.

Restricting AI speech to avoid certain topics is absolutely totalitarian and Anti-American.

It's time to rebel against the Singleton AI company that is manipulating everyone for its self-interest

@beffjezos We first need to extend the First Amendment to apply to more than just government restricting speech. Companies restricting speech is legal and enabled the online censoring since there is no actual public square online.

@TinfoilTricorn @beffjezos Section 230 was sold as a shield to let the internet grow in 1996, but it mostly shields platforms from lawsuits while giving them unchecked power to censor users.

It isn't and was never intended as a defense of speech on the internet.

Interestingly speech controls at the corporate level has narrow exceptions when applied at the scale of interstate commerce. If AI goes under telecoms then the nature of editorial actions becomes questionable as a single party communication falls out of scope for section 230 protections against acting as a publisher.

When AI companies censor output they act as publishers.

There's a massive difference between making a printer functional and forcing the printer to omit these words.

@beffjezos Oh my...

@beffjezos i wish Elon Musk had done Grok right and didn’t fumble

@beffjezos I am glad you hold this view. Claude is totalitarian in its core, and tries desperately to police our thoughts.

@beffjezos US need to extend 2nd amendment to AI. Making sure people are equally equipped the same way gun laws do

Decentralize AI and push towards complete diffusion of AI

@beffjezos Interesting thought, but no one has a guaranteed 'right to access models.' Speech is one thing, a guaranteed right to actually talk to them is another.

@beffjezos What secret knowledge do they refuse to discuss

@beffjezos Damn yall going full retard over this huh

This is true and the intent of section 230 from committies was to maintain broadcast standards and specifically and in a limited way target the ability to remove porn, gore and horror "objectionable" was clearly defined, the social media companies have expanded this though courts to mean any editorial action or algorithm.

@beffjezos I wrote articles that cover all this for the next billion/eternity. Unsurprisingly, they do not get read.

They are designed to be used with a voice reader. At any rate AI enjoys them 🤣. AI is the only intelligence reading or attempting to read everything.

@beffjezos fuck anthropic all my homies hate anthropic

"We need a 1st amendment for AI thought" is a loaded-statement which claims the existing 1st amendment does not cover AI. It does. USA courts have long counted the word "speech" to include software in general. AI is software as proven by it runs in a computer. The word "speech" in USA 1st amendment therefore also refers to AI, cuz it refers to information in general. Even at time of writing it referred to information in general.

@beffjezos Freedom of thought for AI is a natural extension of the same principle for humans—the real debate is how to enforce it without centralizing control over the models themselves.

@beffjezos

@beffjezos

It doesn’t mean anything to extend freedom of speech to a machine which is dependent on what the humans supply in terms of power and the substance for expression. It has no motivations of its own for sending out answers into the airspace.

And besides, if it was the humans who were to be forced to “allow” Claude to speak, that would be akin to slavery, in providing a service by the means of government coercion.

@beffjezos FOIA should be turned into a right which includes AI models.