ChaseRensberger

interesting this is so close to the iPhone Duo release.

interesting this would come out so close to the iPhone Duo release.

I think it's probably fine in anthropic's case but I flinch at the "underpromise and overdeliver" line. Heard way too many people say that and what they really mean is "underpromise and deliver."

this should be fun

ive been really enjoying my iPhone duo so far, its hard to imagine going back to a normal phone. it makes it so much more fun to schedule appointments, read blog posts, vibe code, update my calendar before a trip, watch movies, etc...

I feel like I make this point a lot at work. We are not running out of problems. Let’s move the goalposts.

no shortage of answers to this question but the common/short explanation is

- it says codex but not what model, it could have been an abliterated model or something more open to this sorta thing

- convince the model youre not doing the thing youre doing (a bunch of ways to go about this, some short and some long)

its probably possible to convince any model to do any thing if youre clever enough about your approach, kinda just the way this stuff works. General intelligence = infinite ways to direct it.

also tbh it coulda just been fine with it since the user owns the hardware and phrased it in a non malicious way.

id still argue that (human) education and intelligence is not worthless even in this world. learning is fun, a joke is often funnier the more clever it is, movies are better when they are well thought out, reading is more interesting when you can acknowledge the parallels to things youve read before, etc...

still, i think getting ASI to distribute itself properly is going to be harder than people give credit. there is still a lot of problems to solve. i dont think im going to wake up one day and we are shuttling people to jupiter and back.

lmao

i think this is a bit naive unless youre letting a neural interface run opus 12.6 and shutting down the rest of your brain to redirect energy towards local compute.

smart people are better at taking action based on directive than not smart people (im assuming).

> Education and intellectual work isn't really worth anything if AI succeeds

I feel like this is a crazy take. Smart (and motivated) people are better at adapting to a different world than not smart people.

all good points, i think it really is just a preference thing, i keep config files in github via chezmoi and then just chezmoi apply on different machines.

i agree things get a bit messy but i try to keep the OS in a state where I can reinstall at any time without any trouble.

config files on github, pacman -S any tools i need, important docs all in the cloud already, couldnt get much simpler than that imo

i do get the desire to have a reproducible OS, just dont know that its something i need

pacman and AUR are just so peak, nix os is cool but if you dont care a ton about reproducible builds, idk

im sure nixos is fantastic but honestly i just love arch so much i cant imagine ever using anything else.