asefweca

@WallpaperKeith @G0ffThew next, you will say people can be raped through bluetooth

@asefweca @G0ffThew Everyone else who has to make a choice without knowing what others have picked is being emotionally blackmailed by blue pushers who need others to save them.

The calculus to press red is, to put it kindly, short-sighted idiot math. It begins and ends with “what input gives me the best odds of living” with no regard for what the world you’d be living in - where everyone you can trust is dead and everyone else knows it - would look like

@WallpaperKeith @G0ffThew aren't YOU are the one who put others into danger by convincing people to vote more red, trying to overthrow the result?

@asefweca @G0ffThew So it is just intellectual arrogance that makes you try to blackmail others into danger?

@WallpaperKeith @G0ffThew my theory of mind say there must be blue pushers, and my instinct say most would press blue the logic already converged, so I pick blue instead of diverge it

@asefweca @G0ffThew Having thoughtlessly done so, you are now forcing others to risk their life to save you. And not for any good reason either.

@WallpaperKeith @G0ffThew force other people to risk their lives? in that situation, I'm risking my lives to save blue pushers

@asefweca @G0ffThew Why is it dignified to force other people to risk their lives to save you from the consequences of your own poor choice?

@BobTheMutant @mykola a good system promote efficiency and simplicity do you think a trust-based, self-governed community like old japan would lose to a communnity where one-upping others is normalized like india?

@mykola A good system is one where the incentive structure promotes good behavior. If producing goods and services which people desire and will pay you for, results in you getting what you want, then you have an incentive to strive to learn what you can do that others desire.

Ah this one again. To my dear rationalist friends who are so confused as to why anyone would press the blue button: I don't want anyone's blood on my hands. "But if everyone just presses red then everyone lives" yeah but that's also true for blue, plus no complicity in murder.

@ItsRobbAllen I press blue because red pushers are dumb faggots and I refuse to be a dumb faggot

New Trolley problem has arrived, only it's dumber. Pres Red - 100% chance of living. Blue - Non-zero chance of dying. It has nothing to do with empathy for others unless you're stuck on suicidal empathy, which is YOUR fault, not mine. There is ZERO logical reason for ANYONE to press blue. NONE.

@ArVibear @Neutral_OC least based indian account

@Neutral_OC And not even ones he thought of doing that??? He is sitting and scrolling in his mobile next to his son Wow what a loser

This is the face of a man who who just realized that he won in life.

@noriakishikin one dude from that squad just carried the tokyo arc the swords are just that strong, I don't think any wielder have lost to ordinary sorcerer, on screen or in the lore

the anti cloud gouger forces was probs the most secondhand embarrassment ive ever felt whilst reading bachi ngl

@bitcloud because the western "fundamental human rights" are stupid

Giving fundamental human rights to a math function you can spin up in an infinite for loop seems unimaginably stupid.

@beffjezos it do be like that

"Ok Codex, clean up this Claudeslop"

@allgarbled can you do it without a ruler?

Can anyone explain to me how it can do this but it can’t consistently generate a nine sided polygon?

@theNyx__ >no asian in video >account based in south asia

Asian dude violated Black dude at the laundry mat this how it ended🤯

@mike64_t he's just like me fr

This is one of the most frustrating things about Codex. It validates everything three bajillion times, it is scared of dirty state to the point where it avoids introducing derived state so badly that it would rather write incorrect code before it would ever consider not accessing some variable directly it considers a golden source of truth. It will not *ever* introduce derived state even when that’s a good idea, or the only way to solve a problem. It *loves* the SRP wayyyy too badly. Everything becomes goDoExactlyOneThingFrom(source_data), and you end up doing 17 passes over your data in the process. It is a capable programmer but it’s just so f*cking scared of everything. The meme of RL producing three nested try catch blocks to do a single operation still holds, it has just changed form. Not everything is up for validation at all times because there are implied contracts as to where in the architecture that burden is shifted. It does not understand this nuance. With sufficient arguing, yes, but there are more complicated things it understands a lot faster so that really is just RL induced stubbornness because it thinks it knows better.

@iseelines @reddit_lies Claude is the overhyped one, so when victims start using the real better product it create this wave. GPT 5.4 and codex is not too smart but it works throughly and almost never wrong

@reddit_lies Codex any good ? I’m getting sus from the amount of posts recommending it over Claude

Patiently waiting for Grok Code so I can stop forking over $200/month to Anthropic.

@rohanvarma built in GUI/pop up, agent could add buttons to it, so user don't have to copy-paste commands into terminal

Every time I ask yall for feedback on Codex, you say two things: 1. SSH dev box support in the app 2. View your files in the app Both are now available with our launch yesterday!! What’s left?

@qinzytech I already vibecoded that shit

How can Codex control Apps you can't even see? 1. Screenshots via macOS system API. Works even when minimized or hidden behind other windows. 2. Clicks and types go directly to the target process. Nothing else is affected. It's operating at the OS level, not the UI level.

@chetaslua bro this test is insane a non-rubik person would probably never get it right even if they are smart this seems like it benchmarking something else than general intelligence

OpenAI Image V2 - Maskingtape Alpha failed the test as usual and current all gen also failed it , i hope one day we will get this passed Prompt : A validly scrambled Rubik's cube placed by a mirror, clearly showing its mirror reflection. No harsh light reflections.

@MissMi1973 @karpathy I have like 1 tech query for every 10 non tech query across different AI even though tech is my main use

According to OpenAI's own data and a Harvard NBER study, coding queries account for only about 4% of ChatGPT messages, while non-work queries make up over 73%. For non-coding use cases, even $200/month subscribers have experienced stagnation or regression from 2025 through today, precisely because the entire AI industry has mistakenly treated coding as the sole standard for true intelligence. This perhaps reveals a real choice: should AI development serve more people, or more profit?