My last post led to a thread on Mastodon about the merits of LLM-assisted writing in one’s second language. There were some interesting replies, some of which are unfortunately behind the 🔒, so I can’t share them here. After some consideration, I think I can do better than my last post by not letting the… Continue reading LLMs and librarianship in one’s second language
Category: ai
Yegge
I had a couple of hours to kill recently so I installed Steve Yegge’s Beads. I am hoping this solves a problem for me, which is that storing state for Claude Code in markdown files is not optimal. Beads addresses this by handling state more robustly with git. Yegge casts a rather long shadow over… Continue reading Yegge
Vibe, part 2
So now that I have a subscription to Claude Code and have poked at it a bit, I felt it was time for a fully vibe-coded project. This is in advance of the librarians’ vibe workshop, which starts on Monday. I considered this for a few days and ultimately settled on building a Mastodon bot.… Continue reading Vibe, part 2
Sandboxing
In my last post, I talked about how I only trust Claude Code about as far as I can throw it. Beyond what I linked in my previous post, Simon Willison continues to post stories about coding LLMs exfiltrating data. This is not my area of expertise, but the problems seem significant, and are being… Continue reading Sandboxing
Claude code
I’m looking forward to this semester’s vibe coding workshop for librarians. I was notified that I was accepted into the program yesterday. It should be fun! I’m told we’ll be using Claude Code, so I set this up on my laptop last night. First of all, I don’t particularly trust Claude Code. Allowing an agent… Continue reading Claude code
Vibe
I have applied to what I’ve been calling the “vibe coding workshop,” which is more officially known as the Agentic AI Peer-Mentoring Program. It’s an interesting initiative from CUNY’s Office of Library Services to teach CUNY librarians how to code with agentic AI. To be fair, I suspect that most people who learn to code… Continue reading Vibe
LLM malaise
You may have seen my previous posts about my initial attempts to code with LLMs. At the time, they seemed promising. But in the past few months, my mood has shifted over into a sort of depression. Below are a few of my complaints. [1] Using LLMs takes a significant amount of the fun out… Continue reading LLM malaise
Web accessibility: a supplementary approach
You may have heard about the requirement in the U.S. for most state and local governments’ web content to be WCAG 2.1 AA compliant by April 26th. This is a big lift, and our librarians have already been fretting about it for months now. But as I go about poking at Aider and o3, I… Continue reading Web accessibility: a supplementary approach
Getting started with Aider
Last night, I got Aider properly set up on my laptop, both in the terminal and in neovim. It is remarkable. I honestly see how people get into a lot of trouble with vibe coding using tools like this. It’s a world of difference from chatting with Claude.ai via the web, or using GitHub Copilot… Continue reading Getting started with Aider
Curious about Aider
Someone I rather respect (both as a programmer and as a librarian) recently described the period of time after he got properly set up with LLM-enhanced coding tools as “the most productive two weeks of my life.” In a fit of enthusiasm, he completed a ton of previously unfinished projects in very quick succession. Despite… Continue reading Curious about Aider
