Friendly Skeleton
banner
jonathan.eylerwerve.com
Friendly Skeleton
@jonathan.eylerwerve.com
I post art, art process, some trash. Reskeets mean you fucking crushed it, good job.

Day job trying to make the Internet a nice place for humans. Chicago.
There is a NPD behavior in which people are grouped into uncritical allies or mortal enemies, bc cannot integrate "person I like has a reasonable criticism"

en.wikipedia.org/wiki/Splitti...
Splitting (psychology) - Wikipedia
en.wikipedia.org
January 14, 2026 at 4:58 PM
"splitting" is a very compact, very explanatory rationale for US cabinet politics that basically never gets into a newspaper.
January 14, 2026 at 3:27 PM
Filled next to "no really the house elf thing is fucked up, right?" circa 2002
January 14, 2026 at 2:41 PM
Nah, it's bad. In the 90s/00s there were people saying, hey the way he writes women and minorities routinely sucks, and he publicly decided that those reviewers were his mortal enemies. The rest, as they say, is history.
January 14, 2026 at 2:39 PM
The Black Mouldering
January 14, 2026 at 2:32 PM
To be fair, crypto also equals Eric Adams rug pulls
January 14, 2026 at 5:56 AM
A man's gotta have a code
January 14, 2026 at 5:48 AM
It's true! Some of us are also Cubs fans!
January 14, 2026 at 5:16 AM
Reposted by Friendly Skeleton
Really loved this fight
January 13, 2026 at 10:12 PM
Secondary sources for image include AP, New York Times.
January 13, 2026 at 4:25 AM
Aron Rodgers humiliation ritual is complete
elmo from sesame street is standing in front of a fire
Alt: elmo from sesame street is standing in front of a fire
media.tenor.com
January 13, 2026 at 4:18 AM
Agree and same. But even then amplifying the half true but irrelevant is very much the state actor playbook.
January 13, 2026 at 3:28 AM
Thanks dan
January 13, 2026 at 2:23 AM
The goal is confusion, not persuasion. And (I observe anecdotally) it peaks when there is anti-Trump momentum of any kind.
January 13, 2026 at 2:23 AM
The simple explanation is that running bad faith accounts at scale is easy and Russia's efforts on this area go back to 2014, possibly earlier. If a post seems unbelievable... yeah.
January 13, 2026 at 2:22 AM
Suggest you connect with Dan Sinker, who is working on same problem from the end user distribution end of it. Or, was a few weeks ago. @dansinker.com
January 13, 2026 at 2:14 AM
>:(
January 13, 2026 at 1:58 AM
Wow I hope she was ok
January 13, 2026 at 1:58 AM
One more link:

If you want to get into the tech, the book AI Engineering by Chip Huyen (O'Reilly, 2025) is a thick but approachable primer on how these things, particularly apps built on LLMs, actually work.
January 13, 2026 at 1:12 AM
An under-used document is the NIST framework for evaluating risks posed by an AI system, NIST-AI-100-1. It's clear, concise and comprehensive. DOGE fired the authors on day one, I believe.

As you seek to understand the tech, I suggest using the questions in that framework to guide your research.
January 13, 2026 at 1:02 AM