Ian
raindrift.bsky.social
Ian
@raindrift.bsky.social
Philosophical zombie
There’s a thing @jonathanstray.bsky.social said to me once: “Corruption and integrity are both self-reinforcing.” I usually imagine this on an individual level, but of course it applies to societies too.
May 14, 2025 at 6:00 PM
It honestly reminds me of me when I was 21, and would confidently state as fact things that I was *pretty sure* were true.

Knowledge felt different then: it was harder to verify, and at that stage of my life accuracy mattered less. So being unreasonably confident somehow felt more ok (it was not).
April 15, 2025 at 10:39 PM
Wow I did this and got it exactly 100% wrong. What if I’m a robot?
April 12, 2025 at 6:24 PM
At least in tech startup land, being weird is seen as an asset since no normal, rational person would found one. The iconoclast is celebrated. The real torment nexus moment comes when establishment institutions start using this same ai for stuff like child custody and parole decisions.
March 23, 2025 at 4:33 PM
…the same meme, but with Musk riding on Trump’s back as Master Blaster.
March 1, 2025 at 5:54 PM
It's cute how they think "our youth" will see any of the money they're stealing from those old folks (who paid into social security their whole working lives, btw), vs it just going into the pockets of a few people who are already too rich to even spend what they already have.
February 25, 2025 at 10:45 PM
What a great idea! Also notable: you can now get many cat meds in transdermal form from a compounding pharmacy. Some cats hate pills but don't mind having a little goo rubbed on their ear each day.
February 19, 2025 at 12:36 AM
I love the music videos too! They often have such a well-defined form. I’ve always thought it would be fun to make a giant photo booth that would show you and your friends how to dance and then make you one.
February 7, 2025 at 7:17 PM
They say it dies in darkness…
February 6, 2025 at 5:24 PM
Could it be wind at the destination airport?
February 6, 2025 at 5:21 PM
Of course it could also be that someone read the paper, then built a similar counterspeech bot and set it loose. Impossible to know for certain. And the misguided research explanation doesn't hold up for every example here. Eg. I doubt an ethics board would approve the insult-everyone-bot.
December 6, 2024 at 1:49 AM
I showed this to a colleague who does research in this area, and he pointed out that the bot in these examples sounds a lot like the one from this paper: arxiv.org/pdf/2411.14986

This could be a replication attempt but with a faulty hate speech detector, so it tries to argue with friendly posts.
arxiv.org
December 6, 2024 at 1:43 AM
“Every tool is a weapon, if you hold it right” —Ani DiFranco
November 28, 2024 at 9:47 PM