Erik Dao
@erikdao
Machine Learning Engineer
@bachable- Stockholm, Sweden
- erikdao.com/about/
- github.com/erikdao
- Joined October 2026
Commit activity
50 commits in the last year
longest streak 4 days
1,175 / 2,500 XP1,325 XP to root
Tue, Nov 3
replying to @sulimanbenhalim
Right now it doesn't support real-time collaboration yet. But soon will be!
Basically, it'll work just like Google Docs or Overleaf, conflicts would be resolved from the server side and all users who edit the same score would received synchronized, latest updates.
replying to @nanomaduro
Hey @nanomaduro Sorry for the late response, it's been quite well. I've only tested with single-staff music though. The quality of the scan, expectedly, has a huge impact in transcription accuracy.
Tue, Oct 27
replying to @kojoarhinful
Right now @superlogical is inviting for private Beta by batch. So. yeah, it's not public yet.
The clone looks sleek. Not sure how much it can match the features, and also cross platform (mac, windows, linux, iOS)
Mon, Oct 26
Thu, Oct 22
replying to @younes-io
thanks! i frequently follow Andrej Karpathy's resources. they are great. but i don't think he has written anything about LLM inference engine.
Wed, Oct 21
Tue, Oct 20
replying to @benediktschoeffmann
Hack is a nice font.
And wow, I've never heard of Tilix and Zellij. Thanks for introducing them.
replying to @borkasm
who would have thought.
But for real, this has been fun so far. I hope @/dev/null can stay clean of bots
replying to @bharatanaik
I can definitely share this feeling with you.
When I first used Claude Code and it wrote a lot of complete, decent programs, I was amazed. And I've been building with Claude very much every day over the past year. It's been great, so far, overall.
But I do miss the old-day…