OpenAI Codex Review 2026 — Updated from Daily Use

kept by eddie mar 04

After a year of daily use, OpenAI Codex has become a reliably indispensable tool for production development, handling 85-90% of well-scoped maintenance tasks.

The Architecture Behind Open-Source LLMs
Open-weight LLMs now uniformly use Mixture-of-Experts transformers, with key differences in attention mechanisms, expert count, post-training via RL, and permissible licenses.
also on WorkOS, Granola, #developer-tools
Why we stopped using SDKs (via @alvinsng)
The author replaced Stripe, WorkOS, and Slack client SDKs with direct REST API calls via a thin wrapper class to reduce dependencies and complexity.
also on WorkOS, #user-experience, #software-engineering, #developer-tools
Amical launches open-source, privacy-focused AI dictation app
Amical is a local-first, MIT-licensed dictation app that uses Whisper and open-source LLMs for privacy-preserving, context-aware speech-to-text on Mac and Windows.
also on Granola, Cursor, Claude Code, #productivity, #developer-tools
Life At The Edge - Asad Khaliq
Computing advances in fractals of concentration and diffusion; AI inference is now diffusing to edge devices, but cloud models remain essential for peak capability.
also on Granola, Claude Code, #user-experience
How to Do AI-Assisted Engineering - by Gregor Ojstersek
Engineers and leaders from 15 companies share structured workflows for AI-assisted coding, emphasizing design-first approaches, rigorous review, and context discipline over ad-hoc prompting.
also on OpenAI Codex, Cursor, Claude Code, #ai-coding, #software-engineering, #productivity, #developer-tools