antho1404

@theo I noticed they tend to touch more things that they shouldn’t than before it’s a bit frustrating, “fixing” some random types that have nothing to do with the feature being built for example, every days it happens at least once for me

Serious question. Has anyone ever noticed meaningful regressions in Codex/OpenAI models? I feel like we talk about this a lot w/ Anthropic but I've never seen a similar discussion with OAI.

@prateekjannu @lexrus Really cool app I will definitely put in my list of app to check

@lexrus Yup we have it already at coasty.ai it’s also open source with similar UI here github.com/coasty-ai/open…

We need an open-source project for this awesome permissions flow

Literally a few keystrokes away, really cool video and love the soundtrack

Now I have to try with my dog 🐕

I was wondering why my Mac was feeling so slow, thanks @nextjs... I would love some ways to reduce that, using monorepo with turbo with basically one app as a landing page and another app with about 20 pages only

@theo @jdecool Can’t wait to see what they do more at this point… ban if they see any mention of Claude in a codebase… or tweet…

Claude Code now throws an error if you use it to try and analyze the Claude Code source

@thekitze @ericzakariasson Yes completely agree 4.5 outside of codex is not great but really amazing inside codex

@ericzakariasson i feel like gpt 5.4 was way smarter in codex than in cursor. i think the harness makes a huge difference.

are u using codex 5.3 or gpt 5.4 for coding

@naaiyy_ That’s nice it’s a good way I think to rethink the dev experience I was always curious to have my browser with my terminal when working it feels less wrong than the terminal with the ide

Introducing the new Glass website ✨ See it here : glassapp.dev

@benitoz @iruletheworldmo That would be a nice birthday gift 🎁

GPT-6 rumored for April 14 from @iruletheworldmo Unverified leak but the details are worth watching through a hardware lens 2M token context window Natively multimodal $2.50/$12 per million tokens “Superapp” merging ChatGPT, Codex, and Atlas browser into one agent Now connect this to what just happened with @AnthropicAI @bcherny confirmed Anthropic is deprioritizing third-party tool access because agentic workloads don’t fit their capacity model. OpenClaw got cut from subscriptions. Token rationing is here Two of the biggest AI labs are telling you the same thing at the same time: Agentic compute demand is outrunning infrastructure Anthropic can’t efficiently serve it on TPUs. OpenAI is allegedly throwing every GPU at their next model and killing Sora to free up capacity. Both are rationing access This is the Co-Design Thesis The models are ready. The software is ready. The hardware isn’t keeping up. And the hardware that handles agentic workloads best (dynamic, unpredictable, CPU+GPU coherent, memory bandwidth dense) is NVIDIA 2M token context? That’s a Memory Wars problem. KV cache at that scale requires HBM4-class bandwidth. There is no workaround Every frontier lab is converging on the same bottleneck. And Jensen has been building for it for three years The race isn’t to build the best model anymore anon The race is to serve it

@theo @ThePrimeagen One net benefits of that would also be a much healthier competition on the tools and general ux. If everyone has the same limits/cost then users will choose the best product

@ThePrimeagen I also want the subsidies to end. I just wish they'd end them for Claude Code users as well 🙃

i am happy that anthropic is doing what they are doing with subscriptions the real cost of ai is really high and this subsidized life coming to an end more and more is a good thing.

@pukerrainbrow That’s why I don’t use chats like chatGPT and use agents instead as they have much better context management and can do a lot more

You have to start a new chat because ChatGPT doesn’t understand anymore

@pcshipp If you want to manage them easily have a look at teamy.run 😉

I’ve hired 5 people for my startup 🤝 Co-founder: OpenClaw 👨‍💻 Coding: Opus 4.6 🐞 Debugging: GPT-5.2 🔬 Research: Grok 4.1 ✍️ Writing: GPT-5 Mini Work anytime, no leave, no drama

@shadcn Maybe you could try teamy.run heavily inspired by codex in terms of UX, of course based on shadcn and not focus exclusively on code, it’s more your agentic team

I need Chat in Codex. Codex UI + ChatGPT.

@Rasmic Are you talking about the undo on top of the summary of files changed? I never had any issue with that and I do it often

Unless I'm missing this, codex not having a simple revert / rewind button is wild... Does it exist? I tried clicking undo and it half undid lol

1/ I stopped managing my own productivity system. My COO agent, Alex, does it now. It's the first system that's actually stuck for me in 10 years. I built this with @TeamyRun Here's how it works 🧵

@sickdotdev Codex is open source since day one

Anthropic is now more open than OpenAI

@ctatedev Him I might have an issue somewhere, probably an old version. I don’t have the —name. Thanks I will try again, that was the only issue I had, otherwise amazing work 😀🙏

@antho1404 Yes - in a linked worktree on branch `xxx`: portless run --name app.teamy next dev Produces: xxx.app.teamy.localhost

Portless v0.9: new defaults based on your feedback HTTPS on 443, HTTP-to-HTTPS on 80, clean URLs, worktree prefixes. All out of the box. Before: http://my-app.localhost:1355 Now: https://my-app.localhost Zero config. The way it should've shipped from day one.

People often argue that open source is useless because people never review the code anyway… this is a good example that proves otherwise, hundreds probably thousands of people spent their day today reviewing the code

@altryne The leak might have come from bun x.com/mrkaran_/statu…

If you, like me, just woke up, let me catch you up on the Claude Code Leak (I know nothing, all conjecture): > Someone inside Anthropic, got switched to Adaptive reasoning mode > Their Claude Code switched to Sonnet > Committed the .map file of Claude Code > Effectively leaking the ENTIRE CC Source Code > @realsigridjin was tired after running 2 south korean hackathons in SF, saw the leak > Rules in Korea are different, he cloned the repo, went to sleep > Wakes up to 25K stars, and his GF begging him to take it down (she's a copyright lawyer) > Their team decided - how about we have agents rewrite this in Python!? Surely... this is more legal > Rewrite in Py > Board a plane to SK🇰🇷 > One of the guys decides python is slow, is now rewriting ALL OF CLAUDE CODE into Rust. > Anthropic cannot take down, cannot sue > Is this "fair use?" > TL;DR - we're about to have open source Claude Code in Rust