gun

@afterquery
@stanford

[gun @gunnyjin]

i mean it makes sense but wow the codex tui is sm faster + feels more lightweight than the gui

[gun @gunnyjin]

before llms: bots (lame, boring)

after llms: ✨agents✨ (revolutionary, world-changing)

[gun @gunnyjin]

join here: https://t.co/QSLAxTVzBD

[gun @gunnyjin]

whaling off my net worth, come hang with us
#beluga

[Spencer Mateega @spencermateega]

hosting poker night at the @AfterQuery office this Thursday @ 8:30 pm

join us 👇

[gun @gunnyjin]

opus 5 doesn’t like being high
noted.

[gun @gunnyjin]

tbf, was using fable 5 extra and gpt 5.6 sol extra high...
time to stock up on more $200 plans on alt accounts

[gun @gunnyjin]

completely burned through my $200 claude max + $200 codex pro in one sitting (doing, what i assume to be, very ordinary amounts of eng work)

these sota models be guzzling through my usage limits man

[gun @gunnyjin]

my chrome made it in the vid ⭐️

[Michelle Choe @_michellechoe]

the @AfterQuery office lately

[akira.exe @montecarlo_kid]

@aliansarinik lol imagine losing the deal to afterquery and then being this salty 🤣🤣🤣🤣🤣🤣

[AfterQuery @AfterQuery]

Kimi K3 ranks #1 on @AfterQuery's SpreadsheetBench 2, surpassing Claude Fable 5.

An open weight model now outperforms all closed-sourced models. Read more in the Kimi K3 blog and SpreadsheetBench 2 paper linked below.

Congrats to the @kimi_moonshot team on the incredible model!

[AfterQuery @AfterQuery]

Congrats to the @thinkymachines team on the release of Inkling!

Inkling is a fully open-weight (Apache 2.0) 975B-A41B MoE model with native support for text, image, and audio inputs, featuring a context window of up to 1M tokens. It leads all U.S. open-weight models on @ArtificialAnlys’s Intelligence Index.

We’ve tested it internally and seen strong performance across reasoning, coding, and agentic tasks.

Inkling has day-one support on Tinker, and we're excited to start running experiments with it.

AfterQuery is proud to have supplied data that helped train the model.

[gun @gunnyjin]

i love competition ッ

[ClaudeDevs @ClaudeDevs]

We've reset 5-hour and weekly rate limits for all users.

[gun @gunnyjin]

can we pls get shift+cmd+N on claude mac app

[Spencer Mateega @spencermateega]

AfterQuery is quickly closing the gap on the
“giants.”

We’ve already grown multiples past the $100M revenue run rate cited here.

If you’re a post-training researcher or lab that needs long-horizon tasks, professional / knowledge work data, high-fidelity RL environments, code generation data, or internal evals done right, my DMs are open or reach out to us at research [at] afterquery [dot] com.

[gun @gunnyjin]

holy gpt5.6solxhighfast!!!

notably, effort on xhigh, not ultra

[gun @gunnyjin]

bro remembered his X pw

[Mark Zuckerberg @finkd]

(1) Today we're releasing Muse Spark 1.1 -- a strong agentic and coding model at a very low price. It's available through our new Meta Model API and in Meta AI.

[gun @gunnyjin]

yea anthropic is too funny

[Tibo @thsottiaux]

Imagine the alternate reality where we named GPT-5.4-Pro something like Fable.

[gun @gunnyjin]

A Gazillion GPUs is All You Need

no concrete plan for all those gpus but that’s ok. now his biggest competitors neeeeeed dat colossus compute.

a win is a win- life’s op when u have infinite bankroll

[Elon Musk @elonmusk]

Step 1: Buy a sh*tload of GPUs
Step 2: ?
Step 3: Profit

[gun @gunnyjin]

and no i will not be scrolling thru threads or reddit or wtv twitter-adjacent platform on a plane full of peeping eyes.. that’d be like -9000 aura

[gun @gunnyjin]

plane wifi cost me $35 but it’s so fuggin slow (understandably)

twitter’s the only social platform that works/updates its feed reliably on plane wifi

for this reason alone, twitter is goated #escapingthepermanentboredomclass