cloudhead @cloudhead.io · 12/07/2026I think Fable 5 / GPT 5.6 write better code than the average open source developer now. My reaction to seeing an AI-generated OSS repo is starting to go from negative to neutral. 140
cloudhead @cloudhead.io · 25/05/2026This. It’s just as hard to write high quality code with agents, if not harder. 020
cloudhead @cloudhead.io · 22/05/2026yeah it seems like it's not an issue to pull-off technically, but haven't seen a UI that offers this yet. 000
cloudhead @cloudhead.io · 22/05/2026It would be cool if you could rebase a conversation with an LLM after editing one of the earlier prompts. Just like `git-rebase`, but over a chat log. This would allow edits to trickle down to all subsequent responses. 120
cloudhead @cloudhead.io · 21/05/2026Vibed a little TUI to configure my Obsbot webcam under Linux, since there is no official software. Works great! 010
cloudhead @cloudhead.io · 18/05/2026Isn’t it the other way around? Secure code will finally be possible because we’ll squash all the bugs 120
cloudhead @cloudhead.io · 14/05/2026If you're building something to last, it's probably not a good idea to vibe code it, unless you plan on rewriting it in the future. 020
cloudhead @cloudhead.io · 14/05/2026More code incidently also means 4) higher cost to audit, 5) more bugs and 6) poorer cpu and memory performance. Still, lots of hand-written software is slop and people still pay for it. 120
cloudhead @cloudhead.io · 14/05/2026The major downside of vibe coding is that you end up with 1) much more code, and 2) badly factored code, which leads to 3) very poor token efficiency when making changes. It probably still all works, but the cost of maintenance and change grows much faster than for manually written/reviewed code. 220
cloudhead @cloudhead.io · 14/05/2026So it's really about building the best possible tool that answers the question "what changed?", and "what does that mean?" 010
cloudhead @cloudhead.io · 14/05/2026I think there's a lot of low hanging fruit already just in pure "old school" review workflows. That's where I'm starting: first building a better diff viewer, then optimizing for code review. When I was spending 90% of my day in code editors, all the small details there mattered, now it's diffs. 110
cloudhead @cloudhead.io · 14/05/2026LLMs are incredibly useful despite writing poor quality code, but we need better code review tools. 270
cloudhead @cloudhead.io · 30/04/2026I know plenty of very talented engineers who use AI assistance because they are able to produce better work with it. It's just a tool. Moderate people, not tools. 000
cloudhead @cloudhead.io · 30/04/2026This is the right policy. The problem is not AI in itself, it's low quality contributions, which there are more of now. That is the root issue and that's what should be addressed, not the tools used by the contributor. github.com/rust-lang/le...github.comPolicy proposal: No low-effort contributions · Issue #273 · rust-lang/leadership-councilNote EDIT(@jieyouxu): Please see #273 (comment) for the finalized wording and scope for the policy being proposed for disposition-merge, which is different from the text in this starting issue desc... 120
cloudhead @cloudhead.io · 24/04/2026Ah, that's a nice change. This means it should pick up this new tool, and we're back full circle! 030
cloudhead @cloudhead.io · 24/04/2026This is what the output of `git show` should look like. I've taken some of the work I did on terminal-based diff rendering for @radicle.dev and made it work as a general git-diff tool. 2130
cloudhead @cloudhead.io · 04/04/2026Good engineering + clever use of LLMs will become unbeatable. We're just getting started. 0130
cloudhead @cloudhead.io · 28/03/2026Trying out autoresearch to reduce compiler code size. Not bad, let's see how much of it I'll keep. 010
cloudhead @cloudhead.io · 07/03/2026An AST editor is always how I imagined the future of code editing -- text is just the visual representation of it, but the editor should operate directly on the AST. This also makes things like code formatters redundant. This is directionally right: ki-editor.orgki-editor.orgKi Editor | Ki EditorMulti-cursor structural editor 040
Reposted by cloudheadRadiant Computer @radiant.computer · 02/03/2026Radiant is open-sourcing its compiler toolchain and launching code.radiant.computer today. 1305
cloudhead @cloudhead.io · 28/02/2026Some more details about compiler bootstrapping, fixed points and trust. Though originally the fixed point was reached after 3 stages, it is now reached in 2! 010
cloudhead @cloudhead.io · 25/02/2026The problem with agents is they don’t know what they don’t know. We humans do have an intuition for it. 000
Reposted by cloudheadRadiant Computer @radiant.computer · 23/02/2026The Radiance compiler has reached a fixed point. This means it can now compile itself and generate identical output to itself. 1322
cloudhead @cloudhead.io · 17/02/2026user> just fix all the bugs I'm tired and going to bed. llm> ok, I'll fix all the bugs. ... 8 hours later ... llm> Wait, the issue might be... Actually... blah blah llm> But wait! Let me just.. blah blah blah user> I'm going back to bed. 020
Reposted by cloudheadRadiant Computer @radiant.computer · 16/02/2026🪵 A new log entry was posted: "A.I. and the Future of Computing" radiant.computer/notes/ai-and...radiant.computerA.I. and the Future of ComputingA new kind of personal computer 072
cloudhead @cloudhead.io · 15/02/2026The bootstrapping stage is a bit of a mindfuck. R0 = C implementation compiled with clang. R1 = Radiance implementation compiled with R0. R2 = Radiance implementation compiled with R1. ⬅️ We're here. 010
cloudhead @cloudhead.io · 15/02/20261. Writing a compiler in C to compile Radiance to RV64 ✅ 2. Porting the C compiler in (1) to Radiance ✅ 3. Compiling the ported compiler in (2) with the compiler in (1) ✅ 4. Compiling the self-hosting Radiance compiler (3) with itself 💥😵💫 120
cloudhead @cloudhead.io · 13/02/2026I can't think of anything more soul crushing in the UX space than trying to make a keyboard work on a smart phone's touchscreen. It simply is the wrong interface. In fact touchscreens are the wrong interface for most things. ios-countdown.winios-countdown.winFix the iOS KeyboardA countdown for Apple to fix the iOS keyboard or lose a customer. 000
cloudhead @cloudhead.io · 12/02/2026Yes, this is a web-based Git repository browser created from scratch in a couple of hours using an LLM. 260
cloudhead @cloudhead.io · 11/02/2026Great read, thanks! You might find @radiant.computer interesting, it is very Wirth-inspired. 000
cloudhead @cloudhead.io · 06/02/2026Great talk about hardware/software co-design and why serious software developers should think about hardware. This is one of the core principles of @radiant.computer h/t @lorenz.leutgeb.xyz www.youtube.com/watch?v=v0Jj...youtube.comBryan Cantrill: Andreessen’s Folly - The False Dichotomy of Software and HardwareYouTube video by Jane Street 040
Reposted by cloudheadRadiant Computer @radiant.computer · 27/01/2026Incompatibility allows true progress. 171
Reposted by cloudheadRadiant Computer @radiant.computer · 25/01/2026🪵 A new log entry was posted: "Radiance Intermediate Language" radiant.computer/log/011-radi...radiant.computerRadiant Log #011A new kind of personal computer 022
cloudhead @cloudhead.io · 21/01/2026Agreed. Having used both extensively I think the reason is simply that the Claude Code CLI is much better, and Claude is faster at coding. 000
cloudhead @cloudhead.io · 16/01/2026"On Being a Computer Scientist in the Time of Collapse" is a really excellent and thought provoking read. I'm one of those optimists that is heavily criticized in this essay. web.cs.ucdavis.edu/~rogaway/pap...web.cs.ucdavis.edu 010
cloudhead @cloudhead.io · 15/01/2026‘What Remains of Edith Finch’ puts every other game I played recently to shame. What a crazy experience. 010
cloudhead @cloudhead.io · 13/01/2026It's a bit like power tools, they are faster but less precise, and may not give the same results in the end, due to the process being different. 010
cloudhead @cloudhead.io · 13/01/2026It's generally still quicker to do certain kinds of edits by hand, if you want something very specific. There's also the fact that writing code is a way to form thoughts and ideas that can be superior to prompting, ie. as purely a thinking tool to explore a design space. 110
cloudhead @cloudhead.io · 13/01/2026And "by hand" doesn't mean typing every character manually, but using Claude in a piecemeal fashion, ie. telling it specifically what functions to write, vs. telling it what the end state should be. 110
cloudhead @cloudhead.io · 13/01/2026The problem is identifying how Claude will perform early enough in the process, and that's hard, even with experience using LLMs. I think in the future I will limit this kind of workflow to maximum 2K LOC, anything over that should be written by hand or broken up in pieces somehow. 110
cloudhead @cloudhead.io · 13/01/2026It does seem like focusing specific code leads to better results, eg. if I ask it to simplify the function `lowerFieldRef`, it would find opportunities to simplify the code which it wouldn't if I asked it to do that for a set of functions which includes that one. 010
