Oh please let it be a fully egalitarian future…
Attached is a perfectly acceptable social media post and two entirely understandable LLM responses to the post and a reimagining of the post.
One relevant data point in terms of how soon we’d conceivably be refactoring language models and updating our own views:
From their research, Sáenz and Poston Jr expect the United States to have progressed to overall white minority demography by 2044.
Of course, a white minority doesn’t necessarily equate a dismantling of a power structure favoring white folks.
It shouldn’t need to be said on the fediverse that racists need not reply - expect most folks will approach this academically as intended.
PS: if you hate this post, please try to be as nice as you can about it - e.g. toss me a polite sentence after you downvote
Who cares what AI thinks?
Like, what’s the point of the screenshot?
That’s what you’re doing.
No one cares AI said something dumb, and you caring makes it seem like you actually believe AI is anything more than a chatbot.
It’s an averaged-out view of what “the internet” thinks, with extra massaging to make sure it isn’t too offensive. I think the chatbot’s take here pretty much encapsulates the general US public opinion.
Why thank you!
Overall, GPT-4 here was a distraction. Didn’t much get responses to the question posed as a result.
Thanks again :)
You could have just told us you don’t know, we didn’t need an example
Don’t know what? Do you mean about the question in the OP? I think in the USA’s case it will be much like South Africa, where there’s still tremendous resentment against Afrikaners even though they are an even smaller minority and no longer form government. I think it will take perhaps up to a century for white Americans to be treated as a minority by others.
Edit: and for the other angle of OP’s question, we could look to the ancient world. Rome, for example, didn’t care much about one’s skin colour or place of birth. Cultural assimilation was what made you acceptable or not. Speaking good Latin and being a good citizen was enough.
One example was more than enough, I get it
toxic comments like this stifle discussion, ironic considering your username
There multiple fundamental mistakes with that…
Would explaining them be more productive?
Sure, but I don’t get paid for this.
You want to be productive, take the time to explain all the ways their wrong. You don’t have time for that?
Let everyone else at least know it’s wrong.
But what you just did was completely pointless
literally do not care about the subject matter or debate whatsoever, I was just pointing out that your behavior is not acceptable
LLMs work by predicting which word comes next based on training data (“the internet”) and the model is then tweaked so that it doesn’t sound like 4chan. How is any of that incorrect, o wise one?
Just about every high schooler with an internet connection is doing their homework following instructions from AI models. You know AI is just autocomplete on steroids, but the vast majority of its users don’t. Several lawyers have already been caught putting AI confabulations into official court documents, for instance.
Not caring about bias and the broader impact of technology is what got us Facebook and Tiktok.
I goofed, I was too lazy man.
We all know that it will be perceived differently to raise up Black people versus trying to do the same for white people. Same as a women’s rights versus a men’s rights group - many cringe at the thought of the latter.
I believe people here have no problem with the social media post in the screenshot. I also believe it would rub us the wrong way to read a white dude mentioning that he was into a flight attendant if he specified she was white.
Instead of finding a real person with views on the subject, of which there are many, I took the shortcut of using a tool whose output was just as I counted on. (The responses didn’t strike me as dumb - I called them entirely understandable.) Then I wanted that to be used for quick reference as a starting point for brainstorming about the future.
May make more sense reading the latter half here so you have something interesting (hopefully!) to think about instead of industrial autocomplete.