Kai Rollmann

@bcherny quick bug report: when i'm in ghostty, in a tmux session, and i have multiple panes running claude, and i expand one pane, then any prev text output remains truncated to the width the non-expanded pane had before. zooming causes re-render and fixes it, but is cumbersome.

@bcherny Wish there was a better tutorial! Feels powerful, but odd: crontab -l doesn't show loops, unclear how claude keeps the crontab / server-side? It checks every 1 sec if my /loop every day at 10pm should run? What if I have no / multiple sessions running / or only in other folder?

3/ Two of the most powerful features in Claude Code: /loop and /schedule Use these to schedule Claude to run automatically at a set interval, for up to a week at a time. I have a bunch of loops running locally: - /loop 5m /babysit, to auto-address code review, auto-rebase, and shepherd my PRs to production - /loop 30m /slack-feedback, to automatically put up PRs for Slack feedback every 30 mins - /loop /post-merge-sweeper to put up PRs to address code review comments I missed - /loop 1h /pr-pruner to close out stale and no longer necessary PRs - lots more!.. Experiment with turning workflows into skills + loops. It's powerful. code.claude.com/docs/en/schedu…

I wanted to share a bunch of my favorite hidden and under-utilized features in Claude Code. I'll focus on the ones I use the most. Here goes.

So much fun to finish half-done side projects with AI & learn something along the way. Check out my little git branch switcher CLI, re-written in Go, published to brew + arch github.com/Kadrian/gbr Ignore ofc that switching branches by hand is somewhat of an ancient workflow

@TweetsOfSumit To fix all the tiny design system inconsistencies ✌️🤓

What would you join Parqet for to fix?

@TweetsOfSumit I can confirm :)

AI sucks so hard at frontend and CSS that I think frontend devs have absolutely no problem finding a job in the next years 😂

Hey @bcherny , since a couple days, here and there Claude Code seems to say "the user" instead of "you", as in: "the user's suggestion is ...", instead of "your suggestion is ...". Feels odd, is it a bug?

@karpathy Agreed to a lot here. One more thing I can't fully grasp: the work mode with agents feels more like "iterative render passes" at a location + surface area, with frustratingly limited depth, when actual programming feels like uncovering perfect pixels, but on a narrow surface

A few random notes from claude coding quite a bit last few weeks. Coding workflow. Given the latest lift in LLM coding capability, like many others I rapidly went from about 80% manual+autocomplete coding and 20% agents in November to 80% agent coding and 20% edits+touchups in December. i.e. I really am mostly programming in English now, a bit sheepishly telling the LLM what code to write... in words. It hurts the ego a bit but the power to operate over software in large "code actions" is just too net useful, especially once you adapt to it, configure it, learn to use it, and wrap your head around what it can and cannot do. This is easily the biggest change to my basic coding workflow in ~2 decades of programming and it happened over the course of a few weeks. I'd expect something similar to be happening to well into double digit percent of engineers out there, while the awareness of it in the general population feels well into low single digit percent. IDEs/agent swarms/fallability. Both the "no need for IDE anymore" hype and the "agent swarm" hype is imo too much for right now. The models definitely still make mistakes and if you have any code you actually care about I would watch them like a hawk, in a nice large IDE on the side. The mistakes have changed a lot - they are not simple syntax errors anymore, they are subtle conceptual errors that a slightly sloppy, hasty junior dev might do. The most common category is that the models make wrong assumptions on your behalf and just run along with them without checking. They also don't manage their confusion, they don't seek clarifications, they don't surface inconsistencies, they don't present tradeoffs, they don't push back when they should, and they are still a little too sycophantic. Things get better in plan mode, but there is some need for a lightweight inline plan mode. They also really like to overcomplicate code and APIs, they bloat abstractions, they don't clean up dead code after themselves, etc. They will implement an inefficient, bloated, brittle construction over 1000 lines of code and it's up to you to be like "umm couldn't you just do this instead?" and they will be like "of course!" and immediately cut it down to 100 lines. They still sometimes change/remove comments and code they don't like or don't sufficiently understand as side effects, even if it is orthogonal to the task at hand. All of this happens despite a few simple attempts to fix it via instructions in CLAUDE . md. Despite all these issues, it is still a net huge improvement and it's very difficult to imagine going back to manual coding. TLDR everyone has their developing flow, my current is a small few CC sessions on the left in ghostty windows/tabs and an IDE on the right for viewing the code + manual edits. Tenacity. It's so interesting to watch an agent relentlessly work at something. They never get tired, they never get demoralized, they just keep going and trying things where a person would have given up long ago to fight another day. It's a "feel the AGI" moment to watch it struggle with something for a long time just to come out victorious 30 minutes later. You realize that stamina is a core bottleneck to work and that with LLMs in hand it has been dramatically increased. Speedups. It's not clear how to measure the "speedup" of LLM assistance. Certainly I feel net way faster at what I was going to do, but the main effect is that I do a lot more than I was going to do because 1) I can code up all kinds of things that just wouldn't have been worth coding before and 2) I can approach code that I couldn't work on before because of knowledge/skill issue. So certainly it's speedup, but it's possibly a lot more an expansion. Leverage. LLMs are exceptionally good at looping until they meet specific goals and this is where most of the "feel the AGI" magic is to be found. Don't tell it what to do, give it success criteria and watch it go. Get it to write tests first and then pass them. Put it in the loop with a browser MCP. Write the naive algorithm that is very likely correct first, then ask it to optimize it while preserving correctness. Change your approach from imperative to declarative to get the agents looping longer and gain leverage. Fun. I didn't anticipate that with agents programming feels *more* fun because a lot of the fill in the blanks drudgery is removed and what remains is the creative part. I also feel less blocked/stuck (which is not fun) and I experience a lot more courage because there's almost always a way to work hand in hand with it to make some positive progress. I have seen the opposite sentiment from other people too; LLM coding will split up engineers based on those who primarily liked coding and those who primarily liked building. Atrophy. I've already noticed that I am slowly starting to atrophy my ability to write code manually. Generation (writing code) and discrimination (reading code) are different capabilities in the brain. Largely due to all the little mostly syntactic details involved in programming, you can review code just fine even if you struggle to write it. Slopacolypse. I am bracing for 2026 as the year of the slopacolypse across all of github, substack, arxiv, X/instagram, and generally all digital media. We're also going to see a lot more AI hype productivity theater (is that even possible?), on the side of actual, real improvements. Questions. A few of the questions on my mind: - What happens to the "10X engineer" - the ratio of productivity between the mean and the max engineer? It's quite possible that this grows *a lot*. - Armed with LLMs, do generalists increasingly outperform specialists? LLMs are a lot better at fill in the blanks (the micro) than grand strategy (the macro). - What does LLM coding feel like in the future? Is it like playing StarCraft? Playing Factorio? Playing music? - How much of society is bottlenecked by digital knowledge work? TLDR Where does this leave us? LLM agent capabilities (Claude & Codex especially) have crossed some kind of threshold of coherence around December 2025 and caused a phase shift in software engineering and closely related. The intelligence part suddenly feels quite a bit ahead of all the rest of it - integrations (tools, knowledge), the necessity for new organizational workflows, processes, diffusion more generally. 2026 is going to be a high energy year as the industry metabolizes the new capability.

Wouldn't it be amazing to have on-device speech recognition, available via a web API, for browsers with a locally installed speech-to-text model? Fill all forms via voice! Speech -> send text + zod rules to external LLM -> get JSON with form values back Help! @jaffathecake

@dhh @dhh are you sure you've listened to what Pistorious said? Have you lived in Germany before? I was looking up to you, makes me quite sad to see you so easily persuaded by that Vance speech.

As on cue, the Germans, who outlaw insults of elected officials and have explored banning their political opponents, are very disappointed about getting called out on it. x.com/AlexandruC4/st…

Europe is in a bad way with freedom of speech. No doubt that even these chilling examples will be dismissed or even excused by most European elites. Incredible to see Vance call this out to clearly. May the seed planted eventually bloom.

@cursor_ai Dear team, I'd love to have a voice-enabled composer mobile app that connects with the repository I've got open on the laptop and allows me to go for a walk, speak the composer commands and review & accept the diffs by swiping left / right. Thank you the amazing work!

As a weekend project I visualized the German federal budget. Fun fact: all ministry of health expenses are covered entirely by the earnings from the tobacco tax alone! kairollmann.de/bundeshaushalt

@dhh Kindof looking forward to an awair element level in-depth review of your findings on what's the gold standard for the crispest font rendering out there

Fuck it. What's the point of being rich if you can't order an 8K monitor on a whim. Arriving on Tuesday! 😄

After realizing that 80% of my dislike of Windows font rendering was down to using non-integer scaling, I'm seriously tempted to give it another go. Maybe even get that crazy @dell 8K monitor, and really go to town! techradar.com/reviews/dell-u…

Say you have an @Apple #iMessage family chat, 10 years of text, img + videos, total of 40GB. You'll want to upgrade to $2.99/month for 200GB #iCloud soon. Say you're 6 family members, you're together paying 6 * 3 * 12 = $216/year, just for family chat? And NO data export option?

@braun_hartmut @LageNation Genau, einfach 3 verschiedene Zoom-Stufen drauf. Finde irgendwie wenn man den Standard-Chart zur höhe des Steuersatzes selbst anschaut, dann versteht man nicht so richtig, dass am Ende genau diese Kurve dabei rauskommt.

@kairollmann @LageNation Warum 3 Grafiken? Eine hätte doch gereicht. Wegen der Details?

@LageNation Zur aktuellen Debatte: hier die effektive Einkommensteuer einmal geplottet

@ashtom What would you recommend German software engineers to do? Start or join European AI startups? Organise local AI meetups? Switch career focus on ML? I’m curious over here :)

As Germany stares economic stagnation, industrial slumps, and worker shortages, it’s critical that we boon our economy with the generational productivity gains AI will offer. I love my home, I deeply want us to succeed. ❤️(3/3)

I am encouraged by this investment, but it’s not enough. Germany is the largest economy in Europe—if we cannot compete on the global stage with AI, Europe itself will fall behind. (1/3) reuters.com/technology/ger…

@dhh Using TS full stack allows us to be end to end type safe. Someone alters a DB column in a migration and you'll catch a resulting frontend issue at compile time. I'm curious: in the RoR ecosystem and now without TS, what's your approach to getting similar benefits?

"So farewell, TypeScript. May you bring much rigor and satisfaction to your tribe while letting the rest of us enjoy JavaScript in the glorious spirit it was originally designed: Free of strong typing." world.hey.com/dhh/turbo-8-is…

@YuriSulyma You're right, that's so interesting. Yet even for formulas that express relationships, couldn't we generate code that illustrates a minimal example use as in: "this is how it plays out when applied"? Maybe to *augment* the formula reading experience instead of *replacing* it?

Could a browser extension be built that replaces formulas in math notation on Wikipedia with code? To make the logic easier to read?

@naval couldn't you say: *all* jobs make you do the same thing over and over – only careers don't?

The worst paying jobs are the ones that make you do the same thing over and over.