661–670 of 1001 posts
The balancing markets are used to keep the power grid in good shape, by smoothing out any last minute mismatches between energy production and consumption.
The project started out of frustration of not being able to get this information without friction.
- Scriptless AI web interface in TS
- Custom static site generator in TS
- Local app-less notification server for iOS
- Minimal websocket-based daily note taking app
The process itself is extremely time consuming, when done manually. My application speeds up the process by a factor of 50 up to 150, depending on how you measure it.
Finally it allows "everybody", to find the 0DTE trades, that are really profitable - something, that currently only the "Pros" can do.
I am experimenting with the current SOTA multimodal LLMs, but performance is still not yet there, they still hallucinate non-existent teeth. (As an aside, I have found a simple but very telling test, I have an image with only 4 teeth visible up and 10 down, so I prompt the modal to count, non have been able to, but Gemini 2.5 pro is the closest of the lot, performance is worse in the description when the counting test fails).
I am going to try segmenting the image to see if I will have better results by prompting to describe segment by segment.
- LegalJoe: AI-powered contract reviews for startups, at the "tech demo" phase right now: https://www.legaljoe.ai/
- ClipMommy: A macOS tool to help (professionals who record a lot of videos | influencers) organize their raw video clips. Simply drag a folder of "disorganized" videos onto ClipMommy, and ClipMommy organizes the videos into folders / subfolders, adding tags, based upon some special statements that you can make at either the start or the end of your video (think audio-based "clapboard"). I'm expecting to release this within a week or two on the Mac App Store (Apple allowing...).
As an aside, I've been very impressed with Claude Code, it's (for me at least!) leading the way for how the next generation of business software might leverage AI. I plan to iterate on LegalJoe to make more "agentic" as a result of what I've seen is possible in Claude Code.
Currently a one-man side project: https://laboratory.love Last year PlasticList discovered that 86% of food products they tested contain plastic chemicals—including 100% of baby food tested. The EU just lowered their "safe" BPA limit by 20,000x. Meanwhile, the FDA allows levels 100x higher than what Europe considers safe. This seemed like a solvable problem. Laboratory.love lets you crowdfund independent testing of s…
https://github.com/rishighan/threetwo
Think of it as a Plex for the digital copies of comics. Point it to a folder full of comics, and it will infer metadata, and present your collection in a Plex-like manner.
ThreeTwo supports Comicinfo.xml, Metron's format. Generally there is no universally agreed-upon metadata format for comic books, comic book archives are essentially .zip or .rar files with images with a fragmented naming convention. ThreeTwo itself uses regexes to parse filenames and match that against ComicVine to extract metadata from there. This is currently the problem I am trying to attack.
Other than that, it integrates with DC++ via AirDC++, and also incorporates an OPDS server.
An AI native issue tracker without manual task management