redamxg

i build stuff and yap about AI
Platform engineer

[redamxg @redamxg]

i spend an unhealthy amount of time following new models, agents and open source AI

figured i might as well post the interesting stuff here✍️

[redamxg @redamxg]

OpenAI doesn’t hire people, it factory resets them

@thsottiaux does the same to Codex

[Matthew Berman @MatthewBerman]

My full interview with Tibo (@thsottiaux)

0:45 Tibo's Lessons from Google DeepMind
4:22 Building OpenAI’s Relentless Culture
7:23 Astra & Next Gen Models
11:18 How Fast AI Changes Developer Workflows
14:27 ChatGPT & Codex Merging
20:25 OpenAI vs. Anthropic
23:37 Why OpenAI Keeps Resetting Limits
30:25 Recursive Self-Improvement
32:00 Dangers That Caused "The Pause"
34:13 Will Ultra Fast Become the Default?
43:20 Why Everyone Needs to Try AI

[redamxg @redamxg]

Somewhere along the way, my model routing collapsed into a single provider architecture.

100% Grok now.

[redamxg @redamxg]

@SpaceXAI @elonmusk That’s why we love @SpaceXAI , Grok to the moon 🚀

[redamxg @redamxg]

@thdxr The craziest weekend we’re having*

[redamxg @redamxg]

@arakharazian honestly the chart looks wrong without @SpaceXAI
OpenAI vs Anthropic makes for a nice little two horse race
until you remember there’s a guy in the corner building the horse, I’d keep an eye on this. Time will prove

[redamxg @redamxg]

@HarshithLucky3 Keep cooking 🧑‍🍳

[redamxg @redamxg]

@thdxr we all know ox alpha is your fork @thdxr 😉

[redamxg @redamxg]

@thdxr says he knows what Ox Alpha is.

Ox Alpha says its creator is undisclosed.

One of them is lying.

I’m putting my money on the model. 👀

[dax @thdxr]

i know what ox alpha is i know what ox alpha is nah nah poo poo

[redamxg @redamxg]

is @opencode actually behind Ox Alpha? i think i know ball 👀

[OpenCode @opencode]

Ox Alpha is now available on OpenCode Go too

For the next 6 days, usage is near unlimited and completely free

It won’t count against your Go usage

[redamxg @redamxg]

@opencode 100T tokens/day is fucking insane 😭

they really just said go break that shit lmfao

[redamxg @redamxg]

okay this Ox Alpha thing is actually pretty fucking interesting.
Anonymous model drops on OpenRouter → 1M context + vision/video + strong coding → people start poking at the fingerprints, and the fingerprints look very GLM/Zhipu

if this really is an unreleased multimodal GLM being quietly tested in the wild, that's way more interesting than another random model launch

[OpenCode @opencode]

Ox Alpha (stealth model) is free for the next week

- 1M Context
- Multi-modal
- Zero Data Retention

Generous rate limits, near unlimited usage

We have capacity for 100T tokens per day, lets see what you can do

[redamxg @redamxg]

I was genuinely just trying Grok 4.6 to see what all the hype was about, i use Codex pretty heavily, so I had zero intention of switching. Then I spent way too much time in Grok Build and somehow burned 85% of my weekly limit 💀

Now I’m actually considering switching

And 4.7 hasn’t even dropped yet…

[redamxg @redamxg]

Nobody talks about this part of the "AI agents" hype enough, well mine writes its own training data.

hits a topic it doesn't know how to handle, generates a new example on the spot, saves it permanently into its own dataset. no human review.

synthetic data generation isn't a research paper anymore. it's a cron job in my repo.

is "no human in the loop" a flex or a red flag to you?

[redamxg @redamxg]

Grok 4.6 is honestly much better than I expected.

It’s fast as hell, surprisingly capable, and actually has some taste. Not Fable level, but still good.

I was already impressed with 4.5 especially with how fast it was, but the writing quirks used to annoy me. 4.6 feels noticeably cleaner. For coding, I’d put it very close to GLM-5.3. Maybe GLM has a slight edge in some SWE tasks, but I keep coming back to Grok because of the native vision, the model size and especially the speed.

And yeah, Grok is still one of the least restricted models out there, which I honestly appreciate.

The @SpaceXAI team alongside with @cursor_ai cooked hard with this one.

And I say that as someone who has absolutely shit on Grok before 😂

If 4.7 lives up to the hype, things could get really interesting.

[redamxg @redamxg]

Grok 4.6 has been my main model all day and I genuinely didn’t expect this.
There’s something about using it that feels just different, might need to rethink my whole AI stack 🤖

[redamxg @redamxg]

@moraitakis Because it’s the right move.

[redamxg @redamxg]

June 2025: Us models were ~70% of tokens on OpenRouter, the router most developers use
June 2026: roughly 30%

Nobody announced this.

[redamxg @redamxg]

@Henryf1w I’m not betting, I’m confidently using Grok.
Something about the trajectory feels different

[redamxg @redamxg]

@Medbenn7

[redamxg @redamxg]

Obsidian + Grok 4.6 ... yeah i think i found my winning combo
What's your current go to setup rn ?👀

[redamxg @redamxg]

@elonmusk @art_muse Did we unlock Muskibo? I mean Mibo.
Thanks to @thsottiaux