I was thinking of OpenAI’s rights as the owners of ChatGPT to allow their software to produce speech. A quick perusal of Google for ai generated speech or bot speech only brings up first amendment discussions. What other area of law might deal with this?

I'm convinced this is what happened behind the scenes. I usually shy away from certain current events that oddly show up in my timelines. So the openAI board vs Sam or whatever I didn't look into at all. Thats my vague conception of a headline...I do suspect tho that we will find out it was something that wasn't reported. And this stuff seems quite relevant.

We can imagine the government wants to censor certain types of response. Rocoo seems to suggest that It wouldn't be against free speech, because the AI isn't a human, if state said to openAI 'you have to censor certain answers....'.

There will be SOME content/answers that are not controversial even IF we thought the 1st Amendment applied, is that right?

I have already declared that we need to constitute unfettered access to AI, and to be clear I'm not suggesting companies MUST offer it, rather we need it so government CAN'T limit it.

And for news I was getting in my timelines, this seems like it could have been the underlying debate.

You’re convinced based on what? There has been zero indication of any connection between the board chaos and government intervention.

I think because of the twitter files, and the state admissions BF mentions, we can assume the same pressure is also on this project that twitter and fb faced. Telling them what to censor is no more than what happened there.

What are you referring to here?

Why can we assume that?

You always argue like this. Assigning some particular viewpoint to a notional "we" that encompasses all of us. Don't do that.

We definitely can’t assume that. Where is the connection to the board shakeup at OpenAI? What about all the other AI companies? At what stage does the government get involved?

This...
Because my suggestion is pointing to the evidence of the twitter files, we can make that assumption, drawn from the corollary question, 'why wouldn't we?' Its not a challenging question, its a sincere one...is there a reason to not assume the same pressure is applied?

Its better maybe if i say 'one'. One could imagine...

I think sometimes we have to think or speak as a single whole. I think sometimes one needs to think or speak as a single whole.

Is there a reason we think that the same pressure to facebook or twitter wouldn't apply to openAI?

I don’t know what kind of pressure if any was exerted on those companies. Assuming that happened, why would you think it would happen at OpenAI? And again, how far does the conspiracy go? There are many AI companies and projects, including open source projects that do similar things to ChatGPT. Why the focus on ChatGPT?

I think we need a twitter files thread to reference. We can't just ignore it, its relevant and significant, I'm only realizing now the existence of the evidence hasn't propagated yet like the snowden revelations.

Well...its the popular one right? Certainly it applies to others just like the twitter files apply to other social media sites.

This is the testimony:

Ah, OK. That's not what I'd call an "admission", but it's not important as I understand you now.

Of course there is. They are two entirely different things. Facebook and Twitter are social media sites, where everyone's posts are both input and output, and where they are public. They are forms of a "town square", where people can go to discuss issues. ChatGPT is completely different, a service where someone asks a question and receives an answer privately. There is no conversation with other people, and there is no viral element to it. Might a government want to apply some pressure on them to alter results in some way? I suppose that's possible. But I certainly wouldn't assume that's happening.

bah, its my norm macdonald patter. didn't mean to call attention to anything with the word 'admission'.
A lot of the censorship was about covid science. And then a lot of the censorship was political bias against trump. We can say there is no reason to think this would spread to AI outputs?

With twitter and FB, it was covert, meant to be secret, and so it can go against the 1st amendment or any other constitutional limitations.

Do we feel there arent or should not be any limitation on what the government can censor in this regard? Sorry if people have answered and I'm re-asking, we have outlined a little more nuance. I think that would be super dangerous if the state was left unchecked here.

No, I'm not saying that. What I'm saying is there's no reason to assume that it is happening, or even that it will.

If it goes against the constitution, then isn't it already limited?

We are talking about something that did happen, and asking what the reason is to think that it doesn't apply also to openAI, since the covid narrative was enforced and countless examples of correct science and experts were censored, and that openAI would be a place people would go for this type of content.

No, but its meant to be. Propriety is important. It has to be upheld by the will of the people.

This is what is problematic about the early moderation flags that I see, its training us to think its fine. Is it fine if companies are forced to censor the same information that twitter and facebook were forced to censor? Why would this platform, openAI, be different?

No, we're talking about ChatGPT, and you're comparing it to something else that happened to other sites.

No, you're asking that. No one else is.

So if it being against the constitution doesn't limit it, what will?

We're kind of going around in circles here. I already explained why it could be different.

I think it makes more sense to compare OpenAI to something like Disney. You are free to petition Disney to make a movie about Neonazi Cinderella but if they choose not to it’s probably not because of government censorship.

You see that twitter has lost tons of value since Musk took over and just lost even more advertisers after Musk’s recent statements, right? Why is it so difficult to see that this is likely the primary reason that other companies don’t want unmoderated content being hosted on their platforms.

Propriety. And I think maybe this was what Godel was referring to:
https://en.wikipedia.org/wiki/G%C3%B6del%27s_Loophole
The constitution doesn't do anything if the public doesn't uphold it.

Well because of evidence the countless times that these platforms were forced by the state to censor information and content related to covid or trump. Isn't that more primary?

You’ve counted two places of a different type of business where this allegedly happened. Why would this be “more primary”? In any case, businesses exist to make money, apparently even those like OpenAI that claimed to exempt themselves from this. Their primary interest is making money. Doing things that are demonstrably bad for that is something they try to avoid. You can’t get more primary than that.

jb: the short answer is your ideas seem offensive and purile and the people who make AI tools aren’t interested in them.

Thats a tangential strawman to the point of having government force you to censor.

I called the will to comply primary over making profits.

The worry is the public sentiment shifts to the left and decides that state based censorship is ideal.

Where that is actually happening, sure. But when it’s not known it doesn’t make sense to conclude that censorship is likely coming from an external source when the primary interest of a business is aligned with policing their own products.

Well assumption there is state intervention doesn't come from the observed moderation/flagging but rather because of the intentions of the censor. Proper covid information was censored, we have evidence of this, and it was mass censorship on multiple platforms. ChatGTP would have been THE goto for information at that time. That would be why we would assume that state would want to censor chatGTP for the EXACT same reason.

You are citing advertising leaving and taking away advertising money, why/how is that also then applicable to chatGTP?