Christian Bager Bach Houmann

1.1K posts

Christian Bager Bach Houmann banner
Christian Bager Bach Houmann

Christian Bager Bach Houmann

@chrisbbh

AI/ML. Software Engineer. Optimist. Endlessly curious. Always learning and building.

Denmark Beigetreten Mayıs 2015
624 Folgt1.2K Follower
Angehefteter Tweet
Christian Bager Bach Houmann
In the last few years, I've been reading books and meticulously taking notes on them. I'm sharing my notes for free. If you want to learn something useful, head on over: bagerbach.com/books
English
5
4
33
0
Felipe Coury 🦀
Felipe Coury 🦀@fcoury·
Codex CLI 0.119.0 is out and I'm excited about two features I worked on: 📋 Ctrl+O now copies the last agent response as markdown - works everywhere, even over SSH. I've been using it nonstop to grab plans for review. 🐛 Zellij support is fixed — no more scrollback bug. 👇
English
44
38
899
86.9K
Christian Bager Bach Houmann retweetet
Jack Lindsey
Jack Lindsey@Jack_W_Lindsey·
Before limited-releasing Claude Mythos Preview, we investigated its internal mechanisms with interpretability techniques. We found it exhibited notably sophisticated (and often unspoken) strategic thinking and situational awareness, at times in service of unwanted actions. (1/14)
Jack Lindsey tweet media
English
151
772
6.8K
961K
Christian Bager Bach Houmann retweetet
Alex Albert
Alex Albert@alexalbert__·
We released Claude Opus 4.6 just two months ago. Today we're sharing some info on our new model, Claude Mythos Preview.
Alex Albert tweet mediaAlex Albert tweet media
English
857
1.2K
17.7K
3M
Christian Bager Bach Houmann retweetet
Obsidian
Obsidian@obsdmd·
Obsidian Sync now has a headless client, so you can sync vaults to a server without using the desktop app. Try the open beta:
Obsidian tweet media
English
114
319
4.1K
888.9K
Christian Bager Bach Houmann
I have been working on a version for Pi. They’re super cool - I’ve seen some really impressive trajectories so far. My initial simple version just worked on a single large context source, but I needed a multi-source implementation. So that’s what I’m working on. Seems like some big labs are investing here also? At least signs they are implementing some of the principles of RLMs. Interesting tech for sure.
English
0
0
1
164
SIGKITTEN
SIGKITTEN@SIGKITTEN·
has anyone tried RLM on one of the existing open-source clankers yet? seems like a pretty easy thing to try
English
12
1
56
11.7K
Christian Bager Bach Houmann retweetet
OpenAI
OpenAI@OpenAI·
GPT-5.3-Codex-Spark is now in research preview. You can just build things—faster.
English
597
640
5.7K
1.5M
Christian Bager Bach Houmann
@tunahorse21 You can literally just build while getting your degree. The bar is so incredibly low and having less time forces focus. Constraints are good.
English
0
0
1
136
Christian Bager Bach Houmann
This might actually finally give me the full agentic coding feedback loop when building plugins Agent can: - Auto build main.js to vault - Trigger plugin reload via commands - Run code in the Obsidian environment via eval to test the plugin - Repeat But I do wonder if the new plugins are just agent workflows. Getting pretty close for me at this point.
English
0
0
4
2.2K
Christian Bager Bach Houmann retweetet
Obsidian
Obsidian@obsdmd·
Anything you can do in Obsidian you can do from the command line. Obsidian CLI is now available in 1.12 (early access).
English
486
1.6K
18.5K
4M
Tibo
Tibo@thsottiaux·
What could we do better on Codex? App, model, strategy and features… what’s wrong in how we approach things that we should improve immediately?
English
1.2K
11
942
101.3K
Christian Bager Bach Houmann
Ed Thorp is a mathematician who figured out how to beat casinos, then figured out how to beat Wall Street. I loved this book. The stories are incredible and the investing chapters are some of the best I’ve read. bagerbach.com/books/a-man-fo…
English
0
0
0
109
Christian Bager Bach Houmann
Just off the top of my head: - worktrees: copy files to worktree dir, like .env - multi-directory / repository projects - thread search across projects - the commit and PR creation features sometime feel entirely detached from today where it created a PR saying it couldn’t solve the problem due to something with the sandbox, while the agent actually already has solved the issue - automations seem broken? They instantly archive for me, without doing anything, while the same prompt works fine in the chat window - subagents aren’t visible in the chat view, which is a bit confusing - want to emphasize performance; for many simultaneous sessions/agents, large git diffs (seems to be affected heavily at over 100k lines in index) - clear context when plan is accepted? - forking
English
0
0
2
481
Andrej Karpathy
Andrej Karpathy@karpathy·
Finding myself going back to RSS/Atom feeds a lot more recently. There's a lot more higher quality longform and a lot less slop intended to provoke. Any product that happens to look a bit different today but that has fundamentally the same incentive structures will eventually converge to the same black hole at the center of gravity well. We should bring back RSS - it's open, pervasive, hackable. Download a client, e.g. NetNewsWire (or vibe code one) Cold start: example of getting off the ground, here is a list of 92 RSS feeds of blogs that were most popular on HN in 2025: gist.github.com/emschwartz/e6d… Works great and you will lose a lot fewer brain cells. I don't know, something has to change.
English
546
914
9.2K
1.3M
Christian Bager Bach Houmann
@iannuttall Its super promising! I ended up turning it off after some days of testing. It’s doesn’t feel quite ready, and I’m not yet getting better performance that my existing workflow. The prompts are pretty interesting and useful for inspiration: github.com/openai/codex/t…
English
0
0
1
537
Ian Nuttall
Ian Nuttall@iannuttall·
codex cli pro tip: get early access to plan mode by adding collaboration modes under features in ~/.codex/config.toml: [features] collaboration_modes = true it's a bit rough around the edges right now, but asks good questions!
English
21
22
376
28.4K