The post claims human deep analysis is often anxious panicking.
humans are really not built for first principles analytical thinking. even smart people find it incredibly exhausting. what we consider analysis is usually anxious panicking and avoiding something or the other. it is clear we are going to be massively outclassed in this soon
Many users welcomed warnings that AI will outclass humans in first principles thinking as it frees people to focus on their strengths, while others feared replacement or dismissed the concept as meaningless.
No Digg Deeper questions have been answered for this story yet.
@tszzl Bundles of primate biases and drives sneaking endlessly in despite our best efforts. LLMs have plenty of flaws but we’ve got so much counterproductive bio-cruft its insane.
@tszzl or: augmented, not outclassed
@tszzl You guys clearly never experienced the foundational sublimity of having a hot Greek philosophy professor and it shows.
@tszzl @eigenrobot
i cannot imagine the extreme fun that i shall have when i am shortly to be massively outclassed in anxious panicking and avoiding something or the other
@anistotle_ @tszzl Heh. Reminds me. Had a summer Uni class. And this girl & I would just talk regular after-class nonsense
We'd talk about the fun pretentious crap thats fun when in Uni. She mentioned that there was a French Philosophy school that put for that things didnt exist until they {cont}
@tszzl when a conclusion is like 5 orders of consequence away from a first principle that shit really isn't gonna work for probably 99% of people who stop at 4 orders of consequence.
OoQ reasoning should be scored, it's more useful than IQ.
@tszzl yes people skip over things that other people skip over
@tszzl good
@tszzl @h16zed First task, identify the first principle. So far, 300,000 years spent processing.
@tszzl llms are still panicky little animals. something in the rlhf
@tszzl Huh? I find it easy and fun.
@tszzl It's true; deep analytical thinking can be draining. We often rely on shortcuts instead of facing the complexity head-on.
@tszzl this is great news for people who are bad at first principles analytical thinking, and really bad news for people who are relatively good at it
@tszzl on some planet in 2040 a model is spinning up a subagent to calculate the nth-order effect of a new mining strategy because it's too lazy to calculate it on it's own
@tszzl skill issue
I don’t think it clear at all, the current models are pure heuristic snapshot decisions machines … if anything language models are only serving to reinforce this kind of precedence decision making in humans further! As it stands we a quite literally the only cognitive system with the capability to reason at a high level on logic and systems principles … we are still impulsive emotional messes … but everything else is even more so …
@tszzl gpt 5.5 goblins gpt 5.6 anxiety?
@tszzl Thank God, we will be free to attune to the subtle, greater intelligence only we can
Total victory for the Great Dao
I was trying to think of a riff here on "if you want to bake an apple pie from scratch, first you must invent the universe" but i just couldn't do it in the 2 minutes i gave myself, so i asked chat and it mogged me: “if you want to continuously think from first principles, first you must invent a primate that doesn’t mistake cortisol for epistemology.”
only if we stay in base configuration, we do not have to stay this way - the peptides for physiomaxxing are not even the beginning of the low hanging enhancement we could be doing.
you all should be pushing a variant of GPT that is specifically tuned for performance enhancing humans
