Proton

I expect pricing will normalize around transparent usage tiers instead of opaque “Pro+” bundles. As well as breakout into the smaller pots of coding platforms as “unlimited” era is now over for cloud #LLM services

@TimJayas I don’t see any livestream

We getting GPT-5.6 today? OpenAI only livestreams when they have something big to drop This event doesn’t feel like one of those... but maybe we still get GPT-5.5 Spark at least What are you expecting from OpenAI today?

@sflorimm It is a perfectly valid development method. Not everyone needs a deep agentic, automatic flow. Sometimes you just need some quick snippets or surgical fixes.

met a dude, he still copy-pasting the code from chatgpt

@icanvardar Harness engineering is the winner.

context engineering beats prompt engineering and almost nobody has caught up yet

I've been learning Zig for a day, and honestly, if you look past its flaws, it's completely flawless

@ProgenticEngine hard to disrupt search when you own search

Google has the data. Google has the talent. Google has the infrastructure. So why aren’t they dominating AI? 🤔

can't believe i'm sitting on a train working on my project with #Codex on my PHONE. how is this possible? honestly generational stuff.

@Rachit6783531 Git for sure… AI will come back around

Which one is more important? - AI - git You can only choose one

The problem with the upcoming #Dell XPS 13 is not the hardware—its @Windows 11. #Microsoft platform isn’t tightly integrated and optimized for a “single platform” the way macOS is, they own both software & hardware so excel at optimized performance that @Dell can’t match.

@AishwaryaDevv The point is they are dedicated to maintaining their existing products and services as their core business... AI is just to supplement them, so they won't devote billions and billions towards AI that would take away focus from Ads or Search.

Customer call today. New product workflow tomorrow. Proaction is a 5-person team building fleet management software with Codex across sales demos, support follow-ups, marketing assets, and engineering work.

@ChatGPTapp Long overdue. Finally.

Long-form writing works better with a little more space. Now you can edit longer pieces in full-screen and save them to your Library to come back to later.

JUST IN: Anthropic, makers of #Claude and Claude Code, have filed for IPO Expect the price to increase and the rate limits to be harsher, as it has to show increased profits each quarter.

@rustaceans_rs githubpr.progentic.org

JUST IN: Rust-lang/rust is adding an official LLM policy: require disclosure, human originality, and understanding of changes to curb low-effort AI-generated PRs. #rustlang #rust

@UltraLinx Google has existing products like Ads and Search… OpenAI and Anthropic only have to focus on AI… they don’t have to make sure existing products/services aren’t broken or properly integrated.

Not sure how Google of all companies hasn't kept up with or beat OpenAI and Anthropic. They literally have everyone's data.

@op7418 Its a middle layer issue within Codex itself, nothing you as a user can do. They are testing GPT 5.6

妈的，终于搞定我的 Codex 断联和推理慢的问题了！ 我有一台电脑上的 Codex 推理速度巨慢，即使开了快速模式也很慢，还有一个问题就是经常连不上。 我一直以为是自己的网络有问题，但它和我的 Mac Studio 处于同一个网络环境，Mac Studio 运行就非常快。 结果今天发现，是我的 config 配置文件有问题。里面写死了两个参数，还写死了两个必须加载的 MCP，导致速度巨慢。 如果你也有类似问题的话，我推荐让你的 Codex 自己检查一下配置文件。你只需要跟它描述具体的表现就行。

@ZaynHao Apple doesn’t move to developers speed, but #WWDC should reveal some more informations from #Xcode

现在 Claude Code / Codex 的 AI Coding 能力这么强，为什么 Apple 还没有将 SwiftUI 的性能优化好？ 导致开发者频频吐槽且换成 AppKit 🥲