apekros

@gabrielchua share the love to adelaide?

Hello Sydney 👋 🇦🇺 GPT-5.5 is here, and the OpenAI Codex & Startup teams are bringing the vibes, tokens, and builder energy to town next week. Catch us here: > OpenAI x January Capital x Lyra x Relevance AI Hackathon [luma.com/aq1yr5vc] > Builder session with Sydney Computing Society - SYNCS/ University of Sydney > Vercel x OpenAI Builder Day [lnkd.in/gU5FH_nX] > Coffee, Coworking and OpenAI Codex with Build Club [luma.com/1hi8t8kw] > OpenAI Codex Hackathon - Sydney, co-hosted with University of Technology Sydney, UTS Startups and Arafat Tehsin, Codex Ambassador for Sydney [luma.com/or8icykr] > Going Global: AI Founders Lunch with OpenAI, Crane Venture Partners and Liminal [luma.com/i5j5egfp] > January Capital x Accel x OpenAI x Airwallex Pitch Day [luma.com/hdw4cp7j] I’ll also be in Melbourne on 30 April for our Codex Community Meetup, hosted by Dr Sam D., Codex Ambassador for Melbourne, and the crew at MLAI: luma.com/yhc5wr8h I also heard Thomas Jeng will be recording an episode of the Startup 360 (Startup Daily) podcast with Simon Thomsen Most of all, I’m excited to finally meet my wonderful teammates in Australia IRL 💙 PS: Some of the events are RSVP-only or registrations have closed, but happy to chat if you’re building something really cool with Codex, GPT-5.5, or GPT Image 2.

we are so back

@Dimillian I do this with nix on both my MacBook and home machines :)

Am I the only one? I have a bunch of config repo, and a bootstrap script and any new machine I can just run bootstrap and it'll pull and install all that

@thsottiaux dms locked - would love to chat about this, could you shoot me a dm so I can reply with my work? :)

Codex is hiring across San Francisco, Seattle, New York, London and full remote. Apply online or DM me with evidence of exception work (full-stack, rust, low level system, distributed systems work, ...). Codex is at the center of everything we do at OpenAI and we are accelerating.

@nullableVoidPtr I have sent you a DM!

@apekros Little baby around it, need to pick up macros and metaprogramming eventually, but always willing to learn, especially when I don't have to worry about time.

slowly beginning to think the world doesn't have a place for me

@gdb wish we could but it’s getting routed to 5.2!

codex momentum is strong, and many people are feeling just how big of a leap 5.3 is. if your organization hasn't tried codex yet, it's worth revisiting.

@rekram11 @danielb0901 @Lsenerman @thdxr @adamdotdev are they still terminating against for this breaching TOS or has that been worked out?

@danielb0901 @Lsenerman @thdxr @adamdotdev u dont need it in ur config and shouldnt have it in there, should work w out

Hello there @adamdotdev @thdxr I'm getting this error on both desktop and cli opencode. I tried downgrading to v1.1.35 and nothing works, any idea? I did not found anything on GH issues about it. Thanks

@dillon_mulroy this hurts

presenting SAFE - the Scaled Agent Framework. this is the only reliable way to manage your ai agents at scale.

@burcs super nice art on the post! did you generate it?

@thdxr can we get an OC black drop with a friendlier time for people in Asia 😅

this is coming too

@HSVSphere can now be a little simpler if you just replace the anthropic auth plugin with v0.0.7 as well

This Nix expression lets you use your Claude Code plan with opencode. And it's managed properly, btw. It is that easy.

@nlothian @GeoffreyHuntley can’t believe the one time I wasn’t in Adelaide I missed him speak!

I saw @GeoffreyHuntley speak at DDD Adelaide in Nov 2025. I've written 2 agent implementations, an agent coordinator and endless random utilities since. I was already AI Pilled, but November seems like a lifetime ago.

I often think about Juan Chanson, in Vinge's "Marooned in Realtime" (where the concept of the Singularity was invented). He talks about how as he worked on Mercury, communication from Earth became increasingly sparse and incomprehensible.

@cloud11665 incredibly based

2025 recap, this probably was the biggest year of my life, I’ve: - Moved to the US on the O1 - Made a lot of new friends - Bought an apartment in Poland - Joined OpenAI, where I’ve done the most engaging work of my life (both research and engineering wise) and shipped big E2E speedups for Sora - Spent more time on other hobbies (CAD design and 3d printing) - Spent a month in Japan going to DJ sets of my favorite artists - Imported 2 rhythm game cabinets from Japan to Poland (500kg combined lmao) Looking forward to what 2026 has to bring!

@jparkjmc i do not understand any upside to asking for less in this scenario

“i will work for free” is an instant no hire btw a bad hire will kill your early stage startup the hourly rate is almost a non-factor in comparison

@kennethnym i will be, want an invite?

nvm getting married in 2026 no matter what

@thdxr @aidenybai time to win

if you followed @aidenybai trying out all the AI PR review tools you'll see 1. he couldn't get most of the tools immediately working 2. it's very hard to make a tool that works for everyone who tries it 3. if you just do that and nothing else you're gonna win

@GeoffreyHuntley yes

been playing around with back pressure recently - specifically having builds of key components: - infrastructure (nixos vm) - server, client and crates as pre-commits that run above automatically. yes? it slows down commits but it ensures that an agent can’t commit broken code the solve is more agents via parallelism and investing heavily into infrastructure that provides build caching

@Teknium @kr0der are you just using it via cc max in claude code harness?

@kr0der No, i use opus now for coding, other tasks its bad at though

My experience with both opus and Gemini. Unfortunately, OpenAI is the least jagged model now