As yall have seen me post and post, I’ve been doing freelance AI training to keep myself afloat while looking for a steady job.
It occurred to me a few days ago that when training the AI, you’re almost forced to think in a certain way that doesn’t really jive with me, and my brain might not be wired for.
I suspect I’m not the only one, but if yall would read it, I’ll make some time to get something more formal writted!
Innit it’s being trained in ways that don’t really accomodate anyone outside of the very specific techbro american culture where they’re being trained too.
AI moderation for example HATES communities that are culturally british, or australian, or new zealand, the english-speaking countries that are generally much more relaxed about language they use and I’m not just talking about the slurs either. I mean just generally american “civility” bs is built into the analysis and it’s fucked. Nothing I’ve written here would be considered bad by a Brit but if I put this in a tiktok or youtube comment it won’t post, it’ll be automoderated for harassment or some shit.
Literally everyone outside an extremely narrow white techbro american interpretation of correct social behaviour is going to be screwed by all this stuff. It’s genuinely culturally insensitive and intrusive. I can’t really say how this differs for non-English speaking countries, I assume the AI training is not as good or not implemented there. Generally the issue is that all “English” speakers are expected to interact by american standards and not their own. This is effectively going to train a generation of people into acting like yanks and I fucking hate it.
Sorry, went off on a tangent here but yes I’m extremely interesting in the neurodiversity angle on the same topic because I am frustrated by this rubbish literally every single day.
Innit it’s being trained in ways that don’t really accomodate anyone outside of the very specific techbro american culture where they’re being trained too.
Le Epic Grok is a glaring example of that. It attempts to be funny in a way that only the most tedious and insufferable techbros think is funny.
There’s just so many things, subcultures mainly, that are harmed by this. Stuff that anyone from that subculture would understand in an instant is just completely not understood. Do you think AI can understand when someone is doing ironic
shit vs not? Hell no. It doesn’t understand the extremely specific nature of GCJ circlejerking. Reddit turns their auto-AI shit off for communities it knows damn well it just does not work on… For now.But not everywhere does. And subcultures are ruined by it because they have to put up with conforming to what a crappy AI wants instead of having real human beings around to understand context.
If someone bothered to train a treat printer on Hexbear it would PPB like a drunk cowboy firing from the hip.
Sounds like it’s speeding up the process that social media kicked off with Anglo countries becoming increasingly more American. I’ve noticed it more here with people who consume a lot of Anglo media. Everything from their mannerisms to way of speaking becomes noticeably more American, even applying American stereotypes of our culture to themselves.
That is so fucking grim. The rot of American cultural hegemony has already so thoroughly infected so much of the globe. To know that it’s not just getting worse but accelerating, killing even faster any actual culture the world over, it really drives me into a doomer emotional state.
It’s not as bad in non-Anglo countries although you do notice it from people who consume a lot of Anglo media.
Other countries have their own sphere of cultural influence in their respective linguistic spheres. I think Brazil is in a similar position with how its social media heavily influences the Portuguese speaking world. In Guatemala we’re more influenced by Mexico and Spain with how dominant they are on social media.
But the Anglo world is definitely becoming increasingly American with the US being as dominant and omnipotent as it is. The EU is also heavily Americanized but they’re a bit of a special case.
One interesting thing I have noticed is that many of the “early adopters” are freaking out over how good “AI Search” is now. I have literally heard people say, “Google search used to be awesome, but now it sucks, but this new AI Search that I’m paying for is just as good as Google search used to be. I love it!” Which is insane to me. They do not seem to realize that a corporation has 1) given them something for free, 2) made the free thing shitty for no reason, then 3) charged them money for the thing that used to be free using a third-party app. And the tech-bros are EXCITED to be paying for this thing that used to be free. Not even realizing the amount of wasted energy, human suffering, and environmental damage is being caused by this unbelievably wasteful system. Truly a cult-leader based economy among the top 1% of earners.
Would love to hear a more insider take on it.
For a slightly more insider take, I’ve barely even used the former, but the free version of https://www.perplexity.ai/ is definitely better than free Google (low bar) due to the ads, scams, etc. on the latter these days that they don’t even care about
I’m convinced the Microsoft Recall shit is going to be used to watch people’s screens at work and generate reports about everything people do at their desks. The boss wants you to function like a machine. They want work performed at a constant, measurable rate, with no variation through the day.
The boss wants you to function like a machine. They want work performed at a constant, measurable rate, with no variation through the day.
Amazon Fulfillment Center toiling metrics for everyone!
Quite a few years ago, the “algorithms are made of pure objective logic and therefore have unbiased output” bazinga speakers were saying anything they could about facial recognition software deciding that beautiful people were all mysteriously one particular skin color
.The only thing new is the amount of hype, the even more feverish Singularity™ fantasy beliefs in play, and the drastically increased carbon waste output and energy demands.
I find the pretend-friendly text presentation of the treat printers (especially Le Epic Grok) to be abrasive to read, like a very obnoxious fake friend that keeps telling the same unfunny jokes.
Yes, that would be of interest! When I look at the “AI” stuff I get the feeling that it will be used to discriminate against us, but I don’t know any specific details. Your experience with it is much more direct, so please tell us
YES please.
So sick of this AI bazinga bulletin that’s gotten to the point where new laptops are shipping with a dedicated Copilot key
Ping me when you post?
Edit: reply with your favorite emoji if you want pinged
Yes pls, this
:ohnoes:
100%. It’s legit starting to reach unimaginable levels of bad as more and more continue to uncritically accept Silicon Valley’s latest scourge on humanity
Gib corgislop
Yes I was literally saying this to someone recently.
Ping me when you post that effortpost, please!
I think it’s an observation many people have made independently, you’re definitely not alone in thinking that way. I think Paris Marx said something to this effect on his newsletter I think.
Would definitely love to read your perspective on it!
P. S. Just read this great article by one of my favorite authors: https://aeon.co/essays/can-computers-think-no-they-cant-actually-do-anything might be useful or interesting for your post.
Do tell, sounds interesting
yes please
Let’s get it! Ping me when it’s done, I’d love to read it.
