Anthropic can’t help themselves…

Anthropic: Hey, you can't steal the data we stole! That's unfair :(

I love Claude but hate Anthropic.

They've distilled the entire Internet and the world's supply of books into a model of about a terabyte, yet they get upset when others do anything similar

Anthropic: "There's nothing to explain. You're trying to kidnap what I've rightfully stolen!" (Princess Bride)

G2i, because nothing screams "we care about our hiring" more than video interviews with AI. What a dystopic future.

"You're absolutely right" is something MS Copilot Chat (GPT 4o/5) spits out as well on the regular when I point out something wrong.

This is part of why there needs to be some regulation over this and over obtaining data, they shouldn't steal from regular people that did things before AI was something in the EULA and then that shouldn't steal from each other, but if they can steal data for training they can't complain if others steal data from them, even if it is not stealing, it would be more like a distil, since they can't get access to their original dataset

Anthropic tryna stop China. Meanwhile Moonshot and Zai: “I am so far beyond you. Lightning shoots from my fingertips”

Doing shady things "for the greater good" is pretty much the archetype of a movie villain. I'm glad you're holding them to the fire, and I've been avoiding them (and their models) ever since they cut off Windsurf... I don't consider them or their models reliable.

How TRAE can get back at anthropic: 1. Support the open agent protocols suggested by Zed and Google 2. Build an open source tool that implements claude code sdk as the open agent protocol 3. Done