Full Width 
Shortcuts 
Sign Up 
Log In 
Articles 
Reading List More from Daniel Marino
Now that I use AI regularly to help write code and complete tasks, I find myself in a position where I’m not learning as much as I used to. That feels a bit crummy. Fairly often, Claude writes code I don’t fully understand. For example: I know enough TypeScript to be dangerous, but I’m definitely not an expert. Recently I had Claude take some typing I thought was overcomplicated and drastically simplify it. The new version was clearly better — I just couldn’t tell you why. So I asked Claude to explain it in simple terms. Sure enough, it did. I guess that’s a silver lining to using AI: it can at least explain why it did what it did. A coworker mentioned he has Claude write one-off explanations with examples whenever this happens, and I liked that idea. So I built a skill around it. The skill looks at my uncommitted code, breaks it down, and writes a short blog article for me to read. I’m a fan of Josh Comeau’s blog and the way he uses interactive examples to teach a concept, so I made that part of the skill too — Claude generates interactive examples alongside the explanation, then packages everything up in a small Vite environment. The first iteration was expensive to run. At least 3 minutes, and roughly 10K tokens. I’m fine burning tokens, but I’m also thrifty 😬. The second iteration kept the Vite environment and styles already wired up, so Claude only had to write the article and the examples. That brought it down to about a minute and 7K tokens. Better — though I’d still like to get the token usage lower. The blog format is great, but it’s a lot of work (for Claude) to generate all of this just to delete it when I’m done reading. It’d be nice to have these articles persist somewhere I could revisit later, or share with other people. That brought me to my current iteration. It’s an Astro project that lives on GitHub. The skill works roughly the same way — review uncommitted code, generate a post complete with interactive examples — but now it publishes to a TIL subdomain where I can read the articles at my leisure or pass them along. It’s all vibe coded, and that feels appropriate. I’m up front about it on the site itself: Things Claude changed that I didn’t understand — explained… by Claude. I’ll probably keep tinkering to bring the token usage down further, but it was a fun project to put together, and I’m happy to let Claude teach me what it’s doing.
This is an updated list of what I’m using in 2026. I’m only sharing what’s changed since last year’s list. Applications Chrome — Chrome was on my list last year, but I’m mentioning it again because I actually used Brave for most of 2025. I like Brave, and honestly I can’t even remember why I switched back to Chrome 🤷🏻. ChatGPT — I hate that I’m listing AI tools, but if I’m going to be transparent, it has to be here. I use it a bit for code, but mostly for rewording messages or rubber-ducking ideas. Claude Code — I wrote about my thoughts and reluctance to embrace AI for coding. My company pays for it, so I use it. Endel — I’m neurodiverse and have been trying to be more intentional about using tools that help me stay focused. Endel is super cool and has a lot of genuinely helpful features. Hyper — I was previously using iTerm2, but switched to Hyper mostly because it feels prettier. That’s probably a dumb reason since I could have customized iTerm2, but I installed Hyper on my new machine and never felt the need to switch back. Retcon — My team is pretty opinionated about Git history, and rewriting history via the command line or VS Code started to feel tedious. Several coworkers use LazyGit, but I didn’t want another terminal-based tool. Retcon does exactly what it promises and lets me handle more advanced Git workflows without having to be a Git super-nerd. Notion — I was using Bear for notes, but didn’t feel like I was getting enough value out of paying for it. That’s not a knock against Bear—it’s a great product. Notion’s free plan works just fine for me. I also tried Obsidian this year and liked it, but parts of it felt quirky. It also seemed like a bigger time investment to make it feel as polished as Notion or Bear. VS Code — I tried Zed for a bit, and there are things I really like about it. That said, I ran into a few weird and buggy behaviors that I just couldn’t get past. Equipment I got a 2025 M4 MacBook Pro when I started my new job this year. It’s more than powerful enough for anything I need to do.
When I first got into hand-lettering, I had a hard time finding people who shared their full process—especially the digitizing phase. So here’s a look at how I typically work. Everyone’s process is different, and a lot depends on your style and how you plan to present the piece. For me, since I don’t have the steadiest hand, I tend to embrace rough, scrappy, distressed looks. Most of my work ends up digital anyway, so mistakes can be fixed later. Here’s a short process video I made a while back. Thumbnails & Sketching Start with quick thumbnail ideas on scrap paper. Refine the chosen direction with a pencil on Bristol Paper. If possible, I use a drawing pencil (softer, lighter lead makes cleanup easier). After inking, I clean up with an art gum eraser. Inking My go-tos are Microns in different thicknesses. In the video I sharedI used a .08 for outlines and Graphic 1 for filling letters. I used a Gelly Roll 10 to add cut-outs, create depth, and fix small mistakes 😬. Digitizing I scan the inked artwork using Scanner Pro on iOS. Clean up in Pixelmator Pro. Add textures for distress and wear (I’ve collected a bunch of texture packs over the years). Where I’ve Used This This is basically the same process I followed for my State Motto series, and most of my other lettering projects. Don’t be afraid to lean into your natural tendencies (like a scrappier style if your hand isn’t steady) and use the digital phase to enhance or correct. Trial and error is a huge part of it.
Does anyone else hate the term vibe coding? I’ve been pretty resistant to incorporating AI directly into my development workflow. I use it all the time to clean up emails, Slack messages, and blog posts (including this one). It’s helped me plan vacations, write my cover letter for Planning Center, and even gain insights from journaling. But coding? That felt like crossing a line. At first, I confined AI to repetitive tasks—formatting long word lists into arrays, deciphering cryptic console errors, or exploring how to integrate Vite into a Rails codebase still using Sprockets. All helpful, but I steered clear of anything that touched “real” engineering. Still, AI isn’t going anywhere. And as much as I hate admitting how much I rely on it, it genuinely makes my life easier. So I gave vibe coding a shot. Spoiler: I’m sold—with a few caveats. Vibe Coding an Alfred Workflow I’d played with basic Alfred workflows before, using the built-in UI or tools like Alfy. But I had an idea for something more complex—and neither the time nor the motivation to learn how to build it from scratch. The Problem At Planning Center, our design system has 300+ tokens (and counting). To grab one, I’d open our Storybook instance, scroll to the search bar, type in a few characters, click to copy, then return to my editor to paste it. Repeat that a few dozen times a day, and the friction starts to add up. The Solution An Alfred workflow that lets me search tokens via fuzzy matching, press Enter to copy, and paste directly—eliminating multiple steps and speeding up my flow. Using Claude Code I’ve tried ChatGPT and GitHub Copilot. They both work well (though I find Copilot’s integration with VS Code a bit invasive). I’m not ready to pay out of pocket, so I started using Claude Code through Planning Center’s access. Setup took less than 10 minutes. Anthropic’s documentation is clear, and their tips on effective usage are actually helpful. Building the Workflow with Claude Code I kicked things off with a prompt: I want an Alfred extension where I can enter “tt” followed by characters, and tokens from URL redacted will be shown using fuzzy search. Claude generated a Python script… that didn’t work. It produced a workflow Alfred couldn’t import. So I responded with: The Alfred extension fails to import. Claude replied something like, “You’re right, let me fix that,” and, impressively, it did. After about three hours of iterative prompts and debugging, I had a fully functional Alfred workflow that: Stores design tokens as a JSON database Caches data locally Refreshes if unused for over an hour Fuzzy-finds tokens on input Displays color swatches for color-based tokens I also used Claude to: Generate 100+ color swatches (since Alfred can’t use data URIs) Write a Bash script that builds the workflow with semantic versioning I don’t know much Python, but I know enough to skim the output and feel confident in the structure. And because this was a small, siloed project, I wasn’t concerned about maintainability or codebase conventions. The Cost Aside from the monthly AI access fee, the entire three-hour build process cost around $10 in usage tokens. That’s ~$3.33/hour—far below the value of my time. And now, every use of the workflow saves me 10+ seconds. That adds up quickly. If I’d tried to build this from scratch, it easily could’ve taken 10+ hours and been half as effective. That’s not even accounting for generating the color swatch images. I did spend about an hour trying to get Claude to scrape the tokens directly from the design system site, but that turned out to be unreliable. Still, the final solution works and required zero extra effort from the rest of my team. The Pros and Cons of AI in the Workflow Pros Efficiency Boost: Great at handling menial but necessary tasks—JSON formatting, token lookups, error decoding, etc. Rapid Prototyping: I built a working Alfred workflow in a single evening. Even if it were only a prototype, it would’ve been valuable. Creative Leverage: Designers and PMs can use AI to sketch out ideas and flows, helping engineers jump into development faster. Cons Skill Stagnation: Over-reliance on AI could erode problem-solving ability or deeper understanding of frameworks and languages. Job Displacement: AI is replacing some roles. At Planning Center, leadership has been clear AI won’t replace people—only support them. But the broader industry picture is less certain. Environmental Cost: Running large AI models consumes a lot of energy. Price Tag: Even if it saves time, AI access and usage can get expensive—especially if you’re footing the bill. So Why Do I Still Feel Icky? It took me a while to name the feeling, but I got there: I’m grieving the loss of what it means to be an engineer. I’ve spent over 20 years solving problems, building UIs, and turning ideas into code. I’ve always taken pride in being able to take a design and bring it to life. But now, what it means to be an engineer is changing—and that’s uncomfortable. I don’t want to become just a prompt wrangler. But I also don’t want to be the person who gets left behind because they refused to adapt. The logic is clear: why pay one engineer for 10 hours of work when another can produce the same result in two with AI? Finding the Sweet Spot I think the answer lies somewhere in the middle. Engineers should still understand their craft—be able to reason through problems, write maintainable code, and build features from scratch when needed. But AI should be a tool we use intentionally to reduce toil and accelerate progress. We don’t call it cheating when someone uses code completion. Why should this be any different?
More in programming
I listen to a lot of podcasts, and I like how they fit around other tasks. I press play, lock my phone, and put it down. I’m free to wash the dishes, fold the laundry, or shop for groceries. Unfortunately, more and more information is only published as a video. Technical talks, conference sessions, video essays – they don’t work in an audio-only podcast app. I could convert these videos to MP3 files, but that breaks down the moment a video isn’t pure spoken word. If a speaker says, “Look at this slide” or holds up a diagram, an audio-only file leaves me stranded. I don’t want to give up the podcast player I like, nor stare at a screen for an hour – but I do want the information in these videos. To solve this, I’m abusing my podcast player’s chapter support. This gives me the best of both worlds: I can listen to a video as audio-first, and glance at my lock screen if I need a moment of visual context. The idea: Chapters every few seconds MP3 files can have ID3 metadata, and ID3 metadata can include chapters. A chapter covers a particular time range, and it can have an associated title, description, and cover art. My podcast app of choice is Overcast, which can’t play videos, but it does have robust chapter support. I can jump between chapters, navigate a table of contents, and see per-chapter cover art. To get videos into Overcast, I’m creating MP3 files with a new chapter every few seconds, and the per-chapter cover art is a corresponding frame from the video. As I play the file, I get a slow, stop-motion-like rendition of the original video. If my phone is locked, I can glance at my lock screen and see the current frame in the Now Playing screen. Overcast is developed by Marco Arment, and I got this idea from Forecast, his app for adding chapters to podcasts. In particular, I was struck by its ability to create chapters that don’t display in the chapter list – ideal if I don’t want a table of contents with hundreds of entries. As I was developing my script, I compared my output to the output from Forecast to ensure I was creating the chapters correctly. The code: FFmpeg and Mutagen There are three steps in this process: Convert a video file to an MP3 Extract images from the video at a fixed interval Insert the images as hidden chapters in the MP3 file Let’s go through each in turn. 1. Convert a video file to an MP3 Converting a video file to an MP3 is a single FFmpeg command: ffmpeg -i video.mp4 audio.mp3 This is consistently the slowest step of the process, and I do wonder if I could use different settings or an alternative encoder to make it go faster – but it’s not slow enough to be worth further investigation. 2. Extract images from the video at a fixed interval Extracting images from a video needs a more complicated FFmpeg command: ffmpeg -i video.mp4 \ -vf 'fps=1/5,scale=iw*sar:ih,scale=min(iw\,945):min(ih\,945):force_original_aspect_ratio=decrease' \ thumbnail_%04d.jpg This extracts an image every 5 seconds, downscales any image larger than 945 pixels square (while preserving the original aspect ratio), and saves the results as sequentially numbered JPEG images (thumbnail_0001.png, thumbnail_0002.png, and so on). The key is the -vf flag, which defines two FFmpeg filters: The fps filter selects one frame every 5 seconds (fps=1/5). The first scale filter scales the width based on the sample aspect ratio (scale=iw*sar:ih). Without this filter, frames can be stretched and distorted. The second scale filter scales the input video, preserving the original aspect ratio (force_original_aspect_ratio=decrease), and ensuring the output images fit within 945×945px or the size of the input video, whichever is smaller. My limit is 945 pixels because that’s the largest size that cover art is shown on my iPhone. This filter still isn’t completely correct – it sometimes creates images from portrait videos that are smaller than I’m expecting – but it’s good enough. These are only thumbnails for glancing at, and if I want to change it later, I can always do the image resizing outside FFmpeg. 3. Insert the images as hidden chapters in the MP3 file Inserting the chapters into the MP3 file is more complicated. Although FFmpeg has basic support for ID3 metadata, as far as I know, it can’t insert chapters with per-chapter artwork. Instead, I’m going to reach for Python and the Mutagen library. Here’s the code to add a chapter to an MP3 file: from mutagen.id3 import APIC, CHAP, ID3, PictureType audio = ID3("audio.mp3") with open("thumbnail_0001.jpg", "rb") as f: img_data = f.read() image_frame = APIC(mime="image/jpeg", type=PictureType.OTHER, data=img_data) chapter_frame = CHAP( element_id="chp1", start_time=0, end_time=5 * 1000, sub_frames=[image_frame] ) audio.add(chapter_frame) audio.save() This creates a single chapter that lasts the first 5 seconds (0 to 5000 milliseconds), and the per-chapter cover art is thumbnail_0001.jpg. If we ran this in a loop, we could add images for every 5 second slice of the original video. This code is inserting two frames into the ID3 metadata: The CHAP (chapter) frame contains the timing information, and it can have subframes for metadata like title, chapter art, or associated URL. The APIC (attached picture) subframe contains information about a picture, which can either be a blob of image data or a URL to an image on the web. Normally, you’d also insert a CTOC frame which defines a table of contents, but I don’t want a TOC with hundreds of 5-second chapters, so I’m deliberately not doing this here. This is allowed by the ID3 spec – you’re not required to insert a CTOC frame if you’re using chapters, and you can have chapters that aren’t listed in your table of contents. To work out which frames I needed, I used Forecast to create some chapters by hand, and I inspected their frames. In particular, loading an MP3 and calling Mutagen’s pprint() method shows a human-readable list of frames, and then I could drill into the individual fields: from mutagen.id3 import ID3 audio = ID3("audio.mp3") print(audio.pprint()) I wrapped all this code in a project called glancecast, which allows you to convert a video file with a single command, with optional flags to set the frame length and chapter art size: $ python3 glancecast.py interesting_talk.mp4 interesting_talk.mp3 The process takes a minute or so to complete, most of which is spent transcoding the video file to MP3. The resulting MP3s are usually 40 to 50 MB in size, which is very reasonable. The outcome: How it looks in practice Here’s what one of these “glanceable” podcasts looks like in Overcast and on my lock screen: Maggie Appleton presented this talk over two years ago and it’s been on my “talks to watch” list ever since. Once I put it in Overcast? I listened to it in less than a day. It’s not a lot of extra information, but enough that I can quickly glance down and get the gist of what a speaker is saying. Both views update with a new frame every few seconds, or I can put my phone in my pocket and ignore the screen. I’ve used this approach for half a dozen videos so far, and I’m happy with the results. I expect to keep using it, because I have a long queue of videos I’ve been meaning to watch. If you’d like to try this, check out glancecast for the full code and instructions. [If the formatting of this post looks odd in your feed reader, visit the original article]
Andrew Baker, the current Group CIO at Capitec Bank wrote an interesting piece on AI and open source, and how these tools that generate code according to one’s specification may replace the general reliance on open source implementations done by contributors around the world. I’d really recommend reading it. I have great admiration and respectContinue reading "AI Isn’t Replacing Open Source"
A framework for thinking about when AI involvement is additive or a violation
Why we need richer, thicker interfaces and better boundary objects for collaborative planning with agents
A look at 10 foundational pillars that enable agents to operate more competently and more efficiently in any codebase.
