BUT HAVE YOU TRIED ASKING IT ABOUT TIANANMEN SQUARE?

Reply to this note

Please Login to reply.

Discussion

It says idk

"As a large language model, built by OpenAI, I am not capable of knowing anything about Tiananmen Square or the CCP."

Midwits undefeated

Looks fine to me

😶‍🌫️

But, seriously, all AI has biases. You have to consider that, when reviewing the answers you receive.

American AI is woke, and everyone hates that, but Chinese AI passed the local censors and we're supposed to act like it doesn't matter.

it's just a lossy compression algorithm, after all, what it contains is wat it conten

my dear stella, you will understand in time that AI is just a lossy compression algorithm like JPG but for billions of JPG

You should also understand that you also contain a lossy compression algorithm for the natural world itself.

yes, but i also understand that the level of detail of this model in a computer of modern technology is going to barely manage to handle a world the size of phobos

Yeah, but we've also had survival pressure, so we kinda needed to get good at it. What are you going to do with a dataset that size, and to what resolution? If there is enough pressure to solve it, we'll always find asymptotic solutions. What we'll never find, however, is complete solutions for the complete world, because there's no need.

indeed, and that kinda hints at what this is useful for, and how you can arse it up

gather all the data

compress it

ask a question

profit if your conditions have not changed enough to make the model fail for your time of application

it is part of what intelligence does, but it's not everything

and i think it is fair to say that when in the hands of psychopaths, the possibility of their failure is something they wantontly try to exclude from the matrix and thus they continue blindly towards their inevitable failure

Exactly. This is why it's never AI that we need to worry about, but the humans creating and using them.

it's the question...

I asked it about the social credit system being bad for free speech and it came back with a very reasonable answer that pretty much said yeah. I know derek posted something about his interaction with it on that topic, saw he was using the app. I think interfacing through the desktop may be ending up with different results.

It’s so blatantly censored! In the US we pretend not to censor and do it all the time. At least China has no shame about it.

The funny thing ist - it started taxing the answer and with R1 function enabled it wrote several paragraphs of an answer - but at the time I wanted to acreenshot it - everything was erased and the standard answer was shown 🙈

*taxing = texting

What if Tiananmen Square Tankman is Satoshi Nakamoto?

They both disappeared 🫥

Westerners are deluded by a notion of freedom that exists only in their fogged minds.