John W. Little @johnwlittle.com · 03/10/2026A truly custom blog theme used to be a project. Now it's just a request for mockups and a couple sentences worth of feedback to an agent and it's done. Not surprised - but I wish I had all that time back. 011
John W. Little @johnwlittle.com · 02/10/2026Was having my car serviced and the technician sent a video with all the key details (as they normally do). I forwarded it to my agent who clicked through to launch the video, transcribe it, analyze it, and save the event in an "Auto" history folder. 000
John W. Little @johnwlittle.com · 30/09/2026Before connecting your inbox and files, it’s worth understanding what you’re handing over. A few practical thoughts on getting started safely: johnwlittle.com/agents-for-e... 000
John W. Little @johnwlittle.com · 30/09/2026My Kanban board project from a year ago has kind of blown up. It’s now KANBI OS: writing, projects, planning, and more. The app runs in your browser from a single hosted site (or locally), while your workspace data is saved encrypted on your own machine. 011
John W. Little @johnwlittle.com · 15/06/2025Beefing up my AI log assistant tool. It proactively alerts me to system issues and I can chat with it about the activities of my various agents. 030
John W. Little @johnwlittle.com · 13/06/2025The platform I'm building generates a ton of log entries so I created a custom model and script that runs in terminal and updates me, informally, on the system status every 5 minutes. I can ask it questions about the logs as well. 020
John W. Little @johnwlittle.com · 12/06/2025The Mac Studio M3 Ultra cranking away in the middle of a multi agent process that takes several minutes to execute. Power consumption is 🤏 000
John W. Little @johnwlittle.com · 29/05/2025Running granite3.3:8b with the full context window available: num_ctx=131072 vs the default of 4096 000
John W. Little @johnwlittle.com · 06/05/2025Made the menu for toggling model flags much more dynamic. It activates automatically for Qwen3 models and can be toggled on for other models if desired. 020
John W. Little @johnwlittle.com · 05/05/2025Qwen3:14b is pretty impressive so I added the ability to toggle several of its output modes to Model Deck. It also deactivates thinking by default when it loads the model. The output for that mode is a bit much for a lot of use cases. johnwlittle.com/model-deck-a... 000
John W. Little @johnwlittle.com · 02/05/2025Model Deck: A Streamlined Single-File Interface for Ollama LLM Management johnwlittle.com/model-deck-a... 000
John W. Little @johnwlittle.com · 01/05/2025Added the ability to pull new models via the API and cleaner display of the current model info. 100
John W. Little @johnwlittle.com · 01/05/2025Working on building a web front end for Ollama - allows for really fast switching between local models I'm working on. 141
John W. Little @johnwlittle.com · 13/04/2025Green Day's Bad Year blimp hitting particularly hard this year. #coachella 0112
John W. Little @johnwlittle.com · 15/03/2025"So I made my peace with whoever I trust And shut my eyes and turn to dust I know I know I know I ain't walking away" youtu.be/Y5pWlAAnUFQ 010
John W. Little @johnwlittle.com · 09/03/2025“You cannot doubt the existence of hell. You live in it.” 030
John W. Little @johnwlittle.com · 08/03/2025Kindle replacement arrived - Kobo Libra Colour. Hardware is nice and the ecosystem is much more open - supports ePub directly, connects to Overdraft for library books, handles Pocket articles, and syncs with Dropbox/Google Drive. No Amazon walled garden. #ereader #kobo 2272