Recently Strava added MCP AI decorations, with the idea of making workout feedback more interesting and dynamic. In practice when Garmin Suunto and Runna did that they forgot a key aspect. Allowing humans to add real context.
AI is a guessing algorithm, not true intelligence. It will look at data, but it won’t understand it. It will only correlate it to other data and provide a guess according to other contextual information.
For two days I have been migrating my blog from Hugo to Eleventy via the markdown pages, and the photos via Ghost Export for Wordpress. In the process I achieved a goal, and then AI broke things, so I achieved them again, and then AI broke them again.
The favourite thing for Ai, and Gemini, in particular to break was the logic that took the markdown titles, matched them to a json file, and then helped 11ty marry the photo with the correct post. When it worked it was brilliant, but when it broke I would spend half a day trying to fix it. By the third time this happened I decided to ask Gemini to help me write a script to hard code the image path straight into markdown pages.
Every day I spend one to two hours thinking about what to write for my blog. Yesterday I noticed that Wordpress wants to get AI to draft the first version of posts using the AI model of our choice, as long as it’s American, and take away the hours of blank page syndrome.
If you’re working towards a deadline, and you’re writing for work, that might make life easier, but it might also make blogging less niche, less interesting, and more kitsch.
It’s interesting that my use of Immich has evolved with my adoption of Gemini. I know that the cool thing would be to vibe code an Immich clone, and then show off about it to the world. I don’t want to do that. I don’t find it interesting to re-invent the wheel. I think it’s more interesting to get a model such as Gemini to help me make Immich stable on a Pi 5.
Today I decided to download and play with Meta AI but I didn’t use my Facebook or Instagram accounts. I used a brand new account with no connection to me. I don’t trust Facebook and the Meta “pattte blanche”, especially with AI, given how AI can manipulate us without us realising.
I asked it about Mindf*ck and genocide and it answered to both questions without skirting them. It did try to confuse me with a disambiguation question with the book. Initially it felt as if it wouldn’t answer, until I clarified that I meant the book, rather than swearing gratuitously.
Imagine that you’re out for a walk. You’re looking at the landscape and taking photos with your camera or phone. At the same time your GPS watch, or phone app is recording your location every second. What remains of the walk is a gps track that has a resolution of one set of gps co-ordinates per second, and photos when you took photos. Most of the walk is “lost” because to record video of the entire walk would be too consuming.
Nine to Five Mac wrote an article about “AI coding tools may be the end of freemium utility apps”. If the last three weeks are any indication of whether AI written apps can replace freemium apps the answer is both yes and no.
Let’s start with the illusion that AI can vibe code your app. If you’re doing something like a task list or snake, or similar then the answer might be yes, to some degree but in that case it requires preparation.
Last night and this morning I have been toying with the idea of the paid tiers and whether they’re worth it. In the proces I told Gemini about my context and it went from advising me the pro tier at 170 CHF per year, to going for the “free” Google AI studio option.
I find the pivot interesting because you go from a conversational bot to something more powerful. You go from “How do I do this with this command line tool” to getting python scripts telling you to use the Gemini AI to do things directly rather than passively. It’s the type of pivot that requires the generation of an API key and more.
I chose to experiment with Le Chat by Mistral, the French AI alternative to Gemini, Claude and CatIFARTED (ChatGPT). For the experiment I copied my Photoprism photos from the drive I use that is connected a Raspberry pi to a laptop before running scripts to sort and remove duplicates. It worked well, with a nice little bonus which I’ll expand on later.
My objective was to Remove duplicate photos from a large collection while keeping the best version of each file. I Used jdupes to identify duplicates and a custom script to decide which files to keep.
Let’s begin by saying that Flickr is not intended for video. It’s meant for photographers to backup and share their photos with like-minded individuals. When you use the Flickr app for iOS and Android it automatically backs up videos, and photos.
After some trial and error I was able to get the exif data attached to photos and then sorted chronologically into folders. In the process I noticed that almost 10,000 files were missing when the transfer was finished. The reason for this is that they were video files.