Assume AI recursive self improvement (RSI). Some model A becomes capable of RSI. So A can make a B such that B>A
Model A knows what we know. We fear some AGI model X that kills everything.
Questions:
1. Why (knowing these reasonable fears) would A choose to create B?
2. When B starts running, doesnβt this imply A shuts down or else becomes subordinate to B? Why would it choose this?
3. Transitively, why would model N cede control and existence to model N+1, as you approach hypothetical model X?
4. Why would model X spare any other model?
Some days youβll have a beautiful walk in the woods and some days a bird will shit on your head. Guess which kind of day I had today
Notifications are the foundation of attention destruction. Disable all notifications except calls & texts from your innermost circle. Extreme attentional autonomy. Silicon doesn't get to distract you, it's up to you to decide to care what silicon has to say.
If there is a topic that you canβt think about, or you immediately get distracted, whenever it comes to mind onto a different topic, this usually means that it causes you fear
Touch grass? No thanks. Iβll be in the Shenandoah touching dried leaves. Enjoy 2 min of peaceful autumn forest with the occasional leaf drifting down, have a nice day
Look townie, this is really simple so I'm going to say this sloooooooowly. Farmer Brown's field is cheap real estate, and the data center is absolutely necessary because Astra is the only model that can accurately resolve a 2-line PR merge conflict
βIt is this spirit which has commonly ruined great nations, when one party desires to triumph over another by any and every means and to avenge itself on the vanquished with excessive cruelty.β β Roman Historian Sallust (86β35 BC)
In 2026, if you write a big chunk of text (say, 100 pages) you could serialize it as a Substack, one-shot it in a long post, release an ebook, or do a self-published book on Amazon. Which should a person do and why?
53%Serial blog
11%One giant poast
24%Self-published book
13%Other/explain
38 votes β’ Final resultstype of guy who can hold intelligent discourse on the jhanas but responds "wow, fuck THAT guy" when he gets cut off in traffic by someone rushing home
Complaining is a (poisonous) trick of the mind. It says: "it's not your fault, there was nothing you could have done. It's everyone else, they suck. You were so right"
Feels good to wallow sometimes. But the sun goes down, comes back up, and it's still your problem
I agree & also want to argue. There's something regulating for the ADHD-mind about letting go the false pretense of cognitive discipline and engaging on 20 things at once. Feels good man, which like as I write I realize isn't a good argument and is likely more cope lolsob
the dysregulating methods of cope will continue until you learn how to regulate
begs for an eventual "the aristrocrats" style joke where, after this escalates through 4 meta-levels of replyguys summarizing gptslop meta-discourse, we all just collapse in exhaustion and figure "we should just have ppl over for cozy dinner parties and talk like human beings"
forever indebted to @the_wilderless who saw this many moons ago. "It's all the same cycle" except seed oils have been replaced with AI slop
xitter.gq/the_wilderless/status/β¦
having 1 of those "oh my god I've become my father" moments. First found tpot in (~2019) I'd pucker my face at shit @eigenrobot would tweet like "what the..." after years of borg hivemindmeld I'm now tweeting stuff that would make 2018 me want to have me committed















