Biggest computers in history are being built
unusual_whales@unusual_whales
Big Tech firms are set to spend over $600 billion on data centers this year, driven by AI investment, per Reuters
English
Nick Dobos
44.7K posts
Nick Dobos
@NickADobos
prompt engineer @browsercompany | I make apps. My code runs on 100mil+ phones | former iOSdev @Twitter | indie dev @mindGoblinStdio Download Hivemind⬇️
Seattle, WA Katılım Ekim 2012
8.8K Takip Edilen52.2K Takipçiler
Nick Dobos retweetledi
Oh. You can make a codex mobile app just by asking codex browser use to watch & poll a chat on ChatGPT .com and opening the same chat on ChatGPT iOS app
Hars@harsette
AM I A GENIUS OR JUST A DUMBAHH FIGURING THIS OUT LATE??? I’m using ChatGPT inside the Codex browser so Codex has live context from my ChatGPT conversation and can immediately turn that context into code or anything else. i can drag and drop live context straight into ChatGPT from codex and vice versa with ZERO FRICTION. I’m literally using AI to brief AI to build with AI. This is the most meta workflow I’ve ever used I can’t even imagine all the use cases for this This is probably just scratching the surface. Maybe everyone else already knew this and I’m just late gng or maybe there's always been a smarter way to do this who knows #Codex #AIAgents #VibeCoding #AgenticCoding #ChatGPT #AI
English
@vnovak_404 @thsottiaux I mean yeah you can of course also do both and just completely max it.
But I’m curious about the triangle of efficiency
English
@NickADobos Why not do both—you can have /goal carry an instruction to spawn a xhigh subagent I think, right @thsottiaux ?
English
What’s the difference between
GPT 5.5 low reasoning + /goal
Vs
GPT 5.5 xhigh reasoning + one shot
Both are essentially yeeting compute at a task.
But which one
- is more efficient?
- works better & produces better results?
- finishes the task?
Seems like the major difference is low would spend less time thinking between each step? And would do way more tool calls because of this?
English
@InvincibleEdge Yeah exactly, so which one is better? Slow consideration? Or rapid movement and iteration?
English
@NickADobos huge difference in how each one actually uses that compute tho
low reasoning + goal feels more like guessing your way forward while the xhigh line actually makes sure before moving
English
@rileybrown @diabrowser is getting better every week :)
A question we have had since the beginning is whether a the browser would be eaten by a ChatGPT super app
or whether the super app is in fact the browser.
English
They’re being inserted into super-apps. They aren’t (and never were) a primary product.
nizzy@nizzyabi
what happened to ai browsers
English
Show us the Codex pets you hatched.
Use /hatch to create your own Codex pet.
We’ll pick 10 favorites to get 30 days of ChatGPT Pro.
OpenAI Developers@OpenAIDevs
Customize your Codex pet with /hatch
English
@orphcorp Ohhhhh I just realized I’m way better at performing self control for others than actually doing it for myself
English
ok for those who seem confused
the crux of the matter is the difference between self-control AND *performing* self-control
if you're not careful you end up teaching the kid to downplay their needs & to merely mask convincingly enough to please authority figures
this is an important skill in itself, but it's qualitatively different from having the skill to *actually* calm one's self down
Matthew Dub@5matthewdub
i never expected to have such a literal application of romeo’s tweet
English
@salem_sofiene Design feedback:
- You might want to lose the gradients. Big sign of Ai code slop.
- layout gets all messed up on mobile
English
@banteg It’s so good.
And give it a few months and some more people trying to prompt it
we basically have full assets for any game
3d soon too
English
codex pets is secretly and example of a game asset pipeline, it's pretty cool, run it and study it. it's not perfect, but consistent enough for zero effort you put in.
Max Weinbach@mweinbach
This is adorable Codex is AGI
English
The codex /hatch skill is the coolest prompt ive seen all year.
It builds and iterates on a sprite sheet!
OpenAI Developers@OpenAIDevs
Customize your Codex pet with /hatch
English
@NickADobos repurposing the skill into something that builds game assets would be really cool
English
@leoxs22 Nonsense? No way.
1. Training data for sprite sheets
2. Handy floating notifications available in other apps with a cute touch
English
Realmente el feature es una pelotudez. Pero demuestra genial como usar workers y armar un buen flujo de revisar y corregir automáticamente
Nick Dobos@NickADobos
The codex /hatch skill is the coolest prompt ive seen all year. It builds and iterates on a sprite sheet!
Español
@AIWarper Yeah a few frames are off. I could keep throwing more compute at it, it’s seemed to fix a few other issues already. But I’m almost out of tokens 😭
English
@NickADobos Am I blind or is that running left and right animation not actually functional. Looks like the legs are just in a fixed position for each frame?
English