📣Calling all app developers! Starting today, you can submit your ChatGPT app for review. Approved apps will be listed in the app directory, a new surface for users to search for apps directly in ChatGPT. openai.com/index/develope…

The Hidden Cost of Unlimited Agent Scope The fastest way to break trust in an agentic system: Unlimited scope. If an agent can: - Touch everything - Change everything - Decide everything Users will either: - Disable it - Stop trusting it - Micromanage it Safe agentic UX

Stop collecting AI coding hacks. Start building a workflow. As AI coding tools keep growing, I see many developers chasing awesome prompts, secret hacks, and the next shiny tool. But what works for someone else may not work for you. You work differently. You think differently.

x.com/i/article/2010…

I used to fight with AI agents constantly - bad context, hallucinations, endless back-and-forth. Then I changed 5 simple habits and my output went through the roof. Quick list of what actually works: - Plan first, code second - Let the agent find files itself (no more manual

Chrome DevTools now supports *individual* network request throttling! Folks have been asking for this for years! DevTools now allows developers to simulate slow network conditions for specific requests rather than the entire page. This helps in testing how a web application

The role of code review is evolving, and it's fascinating to watch. I've been reviewing a lot of AI-assisted code lately. Here's what I've noticed: - The code looks clean - Follows best practices - Usually works as intended But here's the thing, that doesn't automatically

Remotion now has Agent Skills - make videos just with Claude Code! $ npx skills add remotion-dev/skills This animation was created just by prompting 👇

The quiet shift in engineering that nobody's talking about. AI isn't replacing us. It's changing where our value lives. I've noticed something in my day-to-day work lately. I'm writing less code, but I'm making more decisions. And those decisions matter more than ever. Here's

Remotion just dropped Agent Skills, and this honestly changes how I think about making technical videos. I used Remotion Agent Skills + Claude Code to generate a full code-driven video explaining the React rendering process, no timeline editors, no manual keyframes, no After

I’ve been thinking about how many people talk about "vibe coding" and why that framing feels incomplete. Yes, AI lets you move fast. Yes, you can prompt your way to working code. But shipping software isn’t about vibes. It’s about responsibility. What I’m seeing is a quiet

We recently talked about the "Team of Rivals" architecture which we bring to bear aidnn to scale sophisticated AI Agents, here is our paper on arXiv.org: arxiv.org/abs/2601.14351 Discuss: news.ycombinator.com/item?id=468023…

AI is making me code faster. But is it making me a better engineer? Here's what I've learned after years of using AI tools: AI doesn't make you better or worse. It amplifies who you already are. When I use AI the wrong way: - Skip the thinking process - Copy-paste without

Using AI has changed how I debug, but not in the way I first expected. AI doesn’t magically find bugs. What it really does is speed up thinking, if I’m already thinking clearly. When I’m stuck, the most useful thing AI does for me is: help generate possible explanations, not

Getting phenomenal feedback, and resonance, from the Isotopes AI "Team of Rivals" technical paper published in arXiv. Here is the summary from DAIR.AI: "The paper introduces the idea of organizational intelligence for AI: → 50+ specialized agents organized

updated.

ML Engineering ≠ AI Engineering I've noticed a lot of confusion around these two roles lately. Let me break it down: ML Engineering = Training models from scratch AI Engineering = Building reliable systems around existing foundation models Here's the thing: You don't need to