We need a new type of frontend stack/platform that can be used both by engineers doing context engineering for fine tuned changes but also lovable style pure vibe coding for non technical audiences. Does this exist?

What are folks doing about using AI in coding interviews? We’re actively encouraging it but candidates seem hesitant to use it. By all means if you can one shot our coding challenge with AI and explain the code please do. Also if this sounds like you DM me

Worst part about the cloudflare outage is I couldn't use X to bitch about the cloudflare outage

My new pre-meeting ritual is dispatching as many AI coding agents as I can before I join

Cursor 2.0 has been pretty sweet so far. Very nice managing multiple agents in flight. However once I make a PR it becomes a little more clunky. Really need a way for the agents to operate on their own branches so I can more quickly respond to AI code review

Starting off this monday with fixing timezone date logic tests

Nothing says "Monday" quite like an AWS outage. Shout out to all my fellow engineers pinged at 12am PT this morning

Can you imagine if Costco made zyns? Kirkland signature nicotine would slap

Who’s got a sora invite code? I need to make some fire memes

If you're absolutely cracked at AI and can crush data eng/sql/analyst etc you should apply for this role. You can literally save people's lives by enabling the organization to utilize our data effectively. We have a solid foundation, but need someone with a strong technical

@_opencv_

Claude would have me blacked out in 10 minutes of code review easy

Take a shot every time the LLM says "You're absolutely right!", how many minutes you lasting?

Squid game season 3 is so awfully bad I’m questioning how I watched this show at all

Going to try this prompt with claude -> gemini mixed with the custom claude planning commands from @kieranklaassen 's workflow this week. Will report back how it works

Add this (or a modified version) to your CLAUDE​.md and tell Claude to use gemini manually or it will do it on its own as needed. Prompt on Reddit: https://t.co/xUFMxIqDuY

Before I join any meeting I make sure to kick off at least two AI agents on coding tasks. Always be prompting

here's a practical case of how i use LLMs to code day to day - different than the "shit out a whole app" goal everyone is chasing

We really need to develop a standard for coding LLM rules files. I know we’re capable of developing a standard because somehow all the LLM providers do great at ignoring mine.