I share interesting things I find, what I’m learning, and projects I’m trying. Follow me if you like tech, creative stuff, and learning new things. INFJ-T
Joined March 2018
- Tweets 1,052
- Following 1,028
- Followers 187
- Likes 1,051
58 Photos and videos
Pinned Tweet
2 Dec 2024
Created a video how to train Flux LoRAs using my convenience script utilizing ai-toolkit on @quickpodio
youtu.be/0I06JHyvuwQ
1
7
3,300
Jun 10
If OpenAI drops a model with Fable level pricing, that might be my sign to finally end my AI psychosis because the spending is getting insane
Now I just need Alibaba to release a GPT-5.4 xhigh level model for us peasants
1
108
Jun 2
Glad to know that we now have better options in Bedrock.
Jun 1
Now generally available, @OpenAI GPT-5.5, GPT-5.4, and Codex on Amazon Bedrock.
Deploy frontier AI models with automatic scaling through Bedrock's next-gen inference engine.
With these offerings on Bedrock, customers can:
◽ Build autonomous agents that handle multi-step coding, data analysis, and knowledge work with GPT-5.5 and GPT-5.4
◽ Bring Codex, the OpenAI coding agent, into your dev workflow and route all model calls through Bedrock. Pay per token, scale as you go.
37
Jun 1
The USD $100 Codex plan was honestly very worth it this month.
The 5x to 10x Codex usage boost already made it feel generous, then the limits got reset 4 times too lol.
That basically meant I could leave Codex /goal running for long stretches on my personal projects without constantly thinking, "am I about to hit the limit again?"
At this point I am half treating the reset generosity as a Codex feature that you can't find elsewhere. Thanks @thsottiaux
This is why I love Codex.
1
1
291
Anthropic's new Opus 4.8 scores 3.6% lower than GPT 5.5 on Terminal-Bench 2.1. Available to compare side-by-side in Cline now.
(They also announced a plan to release new models with higher intelligence than Opus after adding stronger cyber safeguards in the coming weeks.)
44
58
977
64,501
Aivan Monceller retweeted
May 21
Now, I'd rather wait for Codex / Claude limits to refresh than to code myself.
It's bad, guys.
30
6
175
8,577
Aivan Monceller retweeted
May 16
remember when programming was free
138
161
2,132
258,862
May 18
A surprising number of brilliant people I know have gone through a C# chapter
May 18
did you know the entire opencode team used to do C#?
much to think about
41
Aivan Monceller retweeted
May 14
Free $60 for new accounts in @DevinAI
Thanks @cognition
Giving it a spin today
1
4
11
1,783
May 13
multiplexing over SSH used to be a game of "which keybinding will break my scrollback today?"
@Zellij_dev Zellij is the first tool that actually feels native over a remote connection.
no more ctrl b [ just to see a log.
web sessions that don't lag.
layouts that don't explode.
the "old way" just became obsolete.
54
Aivan Monceller retweeted
Apr 21
Coding agents turn code from artifact into materialized view.
The base tables are the ideas — algorithms, protocols, semantic contracts. The code is one projection. For decades we treated the artifact as authoritative because regenerating was expensive. It isn't anymore.
4
4
21
2,845
Aivan Monceller retweeted
May 8
actually useful skills also come with little programs to execute.
12
2
167
14,391
May 3
Eight years in, custom-elements-ts just got its biggest update: a reactive html/render() runtime with deeply-proxied @State, in addition to the 6-decorator API I shipped in 2018.
Native Web Components in TypeScript. Zero dependencies. Framework free.
geocine.github.io/custom-ele…
1
72
May 2
if you aren’t running e2e tests on your TUI, you’re just ship and praying.
i finally got this full terminal emulator automation harness working on my local. no more manual clicking to see if the UI broke.
how are you guys testing terminal layouts? or are we all just "eyeballing it" in 2026?
Apr 18
Terminal automation e2e testing solved
Now as simple as snapshot, click, type:
– wterm renders terminal-in-html, every cell in the a11y tree
– agent-browser automates pages via the a11y tree
Here's opencode in one browser driving Claude Code in another
1
95
Apr 29
the era of the "code monkey" is officially over.
if you're still measuring your value by lines of code, you're already a bottleneck.
the next generation of 10x devs won't ship features. they'll ship the systems that ship features.
from operator to architect.
1
1
68
Apr 26
gpt 5.5 is so good that even the local model guys started talking about it
lately opus 4.7 sounds so retarded next to gpt-5.5. i did not expect this but i am so back. so so fucking back baby
70
Apr 24
I like where this is going. Not sure when my peers will realize that Codex is superior, maybe by the time they do, I’ll already be exploring the next big thing.
Apr 23
We tried a new thing with NVIDIA to roll out Codex across a whole company and it was awesome to see it work.
Let us know if you'd like to do it at your company!
66
Apr 24
Will try GPT 5.5 and new Codex update today, numbers tell me its definitely better.
Apr 23
Introducing GPT-5.5... together with a ton of new Codex features (more on those in the next hour).
Included in all paid plans and coming to API soon. Update your Codex app or CLI to use it.
openai.com/index/introducing…
87