Do we understand that Elon CAN'T break the first amendment?

I think he might be challenging you to another test-off.

Then once again I don’t understand your issue. You are allowed to say whatever you want to it. How are you not?


It wouldn’t be weird for the mods to enforce the rules about circumventing the filter here. And it’s not that you just might, you have now multiple times done it unnecessarily. You can make your point without doing it.

If you mean what you can say to it, yes. But, by that I mean protected from the government restricting you, not the company.

How does humoring racist conspiracies “balance” anything?

I don't know about you guys, buy I want to hear more about the robot and chez and the photocopier.

This wasn't meant to be some "gotcha"; I sincerely wasn't sure what your understanding was on private companies and free speech. And here, I'm not certain if you simply mean that Twitter isn't subject to freedom of speech restrictions, which of course they aren't just as OpenAI isn't, or if you actually think they have no limits as to what you can post. In case it's the latter:

https://help.twitter.com/en/rules-an...ot%20permitted.

Do you mean in the same way that OpenAI can't? If so, of course!

We agree on this then. But you don't know that twitter and facebook were forced to comply with the government breaking the 1st amendment.

That seems meant to comply with the law, or do mean it goes beyond that?

I do. So we are also asking if private AI should be different, should the government have more control over AI than twitter etc.

So you say. I would not support the government doing that and I don’t think you’re going to get many people disagreeing with you.

You don't believe many people think the government should regulate AI? You still haven't convinced me you aren't one of those people.

It depends on how powerful it is. If we’re just talking about chatbots, generative media, etc then no. But it should be much more involved in the pursuit of AGI. A super intelligent AI is much more like an atom bomb than a product like twitter.

In terms of violating the first amendment, no.

lol Twitter Files.

To me the biggest difference is ChatGpt is a tool, it is not a website used to communicate. I hate not having control of how I run programs. If I want it to rewrite Mein Kampf in the style of dr seus or write an erotic story about a toaster falling in love with a dishwasher I want to have that freedom. To me it is like if my pen refused to move if I tried to draw a pair of titties. Now if I repost this new rhyming Mein Kampf on 2+2 in order to indoctrinate middle age poker players and the mods take it down, I have no issue with that.

Maybe you shouldn't become dependent on a private chatbot to express yourself? idk what to tell you man, people have been writing erotic stories and racial slurs since the dawn of time, I think you'll be fine.

Sure I would love to self host my own LLM but I am not nearly rich or smart enough to do so with any of the sophistication of something like chatgpt.

LLMs have no understanding. They are just predicting the next token. Likewise their censorship has no understanding either. Human mods can do a fairly good job of taking down objectionable content. The way LLMs do it is like chemo, they are hobbling their own system. Even trying to get it to answer some very basic questions has gotten increasingly difficult.

To me it is like being handed a dull knife. Yeah I guess it is harder to stab people but it also makes cooking hard. Guess I shouldn't become dependent on private knives.

First of all, "we" are not asking, you are. Secondly, can you clarify whether this is something you think is happening now (government having control over ChatGPT), or something you're worried about in the future? Earlier, you alluded to controversy around OpenAI, but never clarified what that was. I don't know if you meant the Sam Altman vs the Board silliness, or if this is something about Biden's Executive Order regarding AI, or something else. But in your OP you mention ChatGPT censoring you, which doesn't seem to line up with either of those things. Some clarity would be good.

Well, that is a difference, but I don't think it changes anything. If someone wants to create a tool, they can put restrictions in if they want to, and you can choose whether or not to use it.

Of course, and I can choose to go on a forum to vent about the restrictions being lame.

Sucks man, but you're not entitled to the use of someone else's intellectual property or tools.

Yup and I never claimed otherwise. To me the discussion of whether or not chatgpt can put restrictions on is not very interesting, of course they can. What is more interesting to me is should they.

Well, of course. OP posted about whether ChatGPT should be allowed to restrict things, and I've been responding to that. You brought up the "it's a tool" point in response to that, so I responded in the same context my post was made - whether it should be permitted. If you just want to vent, go for it.

Got it, ty. I should have been more explicit that I don't think OpenAi should forced to do anything. Do you feel chatgpt restricting users is useful?

It's interesting in that it gives answers for private use, so on one hand it shouldn't be a problem what people enter or what it outputs, because they're not publishing it publicly. But I can see three reasons they might want to restrict things:

1. If people decide they want to continually feed it a bunch of hate speech, and it's learning from that, it could affect the output for others in a way they don't want.
2. If people choose to generate hate speech and then post it elsewhere, it's not going to be a good look for ChatGPT.
3. Regardless of the optics, they may not want to be a party to the reposting of hate speech they generated for their own moral reasons.

Ok, so that surprises or u challenge my assumptions a little.

Why so funny? Its a reference. Like 'snowden revelations'.

https://en.wikipedia.org/wiki/Twitter_Files