Large AI accounts and multiple people (or other AI accounts) being adversarial when I pointed out that it is AI is the phenomenon.
An AI content X/Twitter account with nearly 100k followers blocked me, and I got a couple of disapproving replies for pointing out that the account was AI-generated. I quote-tweeted the account mostly to share a useful Chrome Extension that I've been using the detect AI content, but I was surprised that there was a negative reaction in the form of a few replies pointing out the account was AI-generated. I am neither pro- nor anti-AI accounts, but being aware of the nature of the content seems to be useful.
Would be curious to hear others' thoughts on the phenomenon.
Original Tweet: https://x.com/parconley/status/2000064102543376413
Late to respond but I fixed this when I saw it. Thanks!
1I find these ideas interesting and compelling. Thanks for writing. Would also be curious for further conversation on these themes.
Appreciate the suggestion!
I’ve had coffee chats with three people who like that retreat center. I like retreats, and I did a 8h/day home retreat last summer. Have seriously considered Panditarama but it is a big time investment and career break especially early career. And also I worry it would be hard on my body sitting so long and hard, and that Mahasi practice risks being destabilizing. May still do one though. Probably not while getting a business off the ground, but maybe when there is in between time or room for a small sabbatical.
Do you have any thoughts on career circumstances where I could make income for the 9 months I wouldn’t be on retreat?
I agree that you can’t learn a subject from using reference works, and that they are very supplemental, but this comment comes off as pretty harsh! See Exercises in Comprehensive Information Gathering if you want to understand more where I’m coming from in the benefits of reference works.
Thank you for the encouragement and kind words!
A quick Opus 4.5 output asking for 10 more concept names, you'd know if they fit well better than I would.
Claude Opus 4.5 - Concept Name Suggestions
Based on this LessWrong post about "diagonalization" - the phenomenon where a more competent agent simulates and predicts a smaller agent's behavior to exploit weaknesses in their model - here are 10 alternative names:
My personal favorites are outmodeling (most precise), leveling (already has cultural traction), and boxing (most accessible).