
When I’m not rummaging around in production I spend my energy making it easier to understand, and eventually more reliable. Member of Technical Staff @OpenAI
Copenhagen, Denmark
Joined April 2008
- Tweets11.8K
- Following788
- Followers1.5K
- Likes4.5K
Pinned Tweet
Yay 🍾 the Reliability issue of @IncrementMag is out today! 🥳 I’m really proud to have contributed an article to the issue - it was an absolutely joy working with their amazing editors @rebeccahiscott and @ohreallysid increment.com/reliability/ob… – at Odsherred, Danmark
Mads ♥mann retweeted
DM if you want to help scale dots to 1B users and have done similar things in the past
Join this team so I can figure out what animal is taking a dump in the middle of my garden multiple times a week
Do you know an embedded systems engineer? We’re looking for someone to join us at @beholdHQ.
A good chance for your work life to involve Linux and badgers, too.
Shares appreciated.
Drop me a DM.
Mads ♥mann retweeted
Do you know an embedded systems engineer? We’re looking for someone to join us at @beholdHQ.
A good chance for your work life to involve Linux and badgers, too.
Shares appreciated.
Drop me a DM.
Working when SF sleeps means I get to talk to a lot of co-workers and teams agents, to the point where I know their agents better than the people behind them 😆
My brother needs help. He just can’t. stop. shipping. 😂🚀
Minesweeper, but it’s a real-time strategy game.
Read the clues. Claim the land. Build your kingdom.
Keepsweeper is out now on iOS. Free to try.
apps.apple.com/app/id6814857…
Really nice that your Dot (Blobby) gets its own Slack handle 👌🏻 makes it super convenient to delegate arbitrary tasks instead of tagging Codex, Linear, or any specialized agent directly. My Blobby knows my preferences best anyway
This is what I’ve been working on the last couple of weeks ✨ give it a try and ping me if you hit any bugs I should fix 🙏
Codex Security Cloud is getting a major upgrade, with access to cyber-capable models through Daybreak Blue included by default.
It scans entire GitHub repos, continuously reviews new commits, investigates and deduplicates findings, and prepares fixes for review – even when your laptop is closed.
Available as a plugin in Codex desktop and web.
Usually demos like these have someone back stage doing the exact same thing, ready for instant cut over. If that also fails, a pre-recorded video is the 2nd cut over.
This made me like OpenAI even more. Live demos are so hard, the internet in this venue can be flakey, you are super nervous, etc..
This stuff takes weeks to prepare and I doubt these presenters had that much time
🤯 turns out Codex (Astra, Medium) is pretty damn good at creating patterns for leather working. This is just based off a casual chat about a leather journal I'm making for my sister
I really enjoy using Voice to talk through technical topics. Find some topic you don’t understand, ask it to explain it to you, and keep asking questions until you feel that you’ve learned something
Reading up on Interpretability over the weekend. I think I’m drawn to the topic for the same reason I find observability interesting and important. I want to be able to peek under the hood when things are off
I'm having an unreasonable amount of fun building a site for my 2009 Mac Pro that now work as a home lab computer running NixOS. It's OCaml using SSR, web components, and SSE.
Finally got a replacement PSU and found time to install it.
Still doesn’t turn on 😂
Good news is that ChatGPT has gotten substantially better at debugging it it would seem
Taking my IELTS exam today. Ended up having to fly to London to take it as the test center in Copenhagen only runs tests at most once a month so couldn’t wait.
My flight was delayed so arrived 2am, slept 4 hours, and then commuted to the test center 🫠
This was my experience as well. Crazy how quickly you can get up to speed when you can easily find answers to
- What projects are in flight that might be interesting to me
- Had anyone ever thought about XYZ. If so, what was the conclusion
- I want to do XYZ, what’s the process
Onboarding at OpenAI is one of the fastest I have ever experienced.
Step 1) login to Slack
Step 2) download and login to ChatGPT.
That’s basically it. Anything in this company can be driven from here, and it comes readily configured. You can figure out how to do anything or find the right person for anything. All workflows and tools integrate well into Slack and so you don’t have to leave these two surfaces.
This is the future of onboarding and enables engineers to ship on day one at scale.
I didn’t think it was possible to keep this velocity at this scale, but it is.
Yes, we’ll experience more shitty software due to LLMs but we’ll also get a ton of super high quality software experiences from a bunch of opinionated, detail oriented people who were never able to build software before.
Mads ♥mann retweeted
An open letter for a global surge in cyber defense, signed by over 100 organizations including Anthropic, AWS, Google, Microsoft, OpenAI, and Oracle.
















