Two people have mentioned my blog in the past 24 hours, and that's infinitely more attention than I ever wanted. You are looking at this on a nice website, but I'm typing this into a file on GitHub. There isn't some rich text interface here. If I want to bold phrases, I need to squint at the "Markdown Cheat Sheet" taped behind my computer and scan for How to bold text.
I do think about this blog here and there. I wanted it to be a Town Square where I'd post updates about my life and create accountability for myself by shouting my intentions into the void. Unfortunately, the situation is closer to that of a barking dog who gets distracted by a butterfly, and then he runs off into the woods, and now he's back, but you haven't forgotten and he probably hasn't changed.
Today I'm preparing material for a virtual assistant. Have them monitor facebook groups, post ads, edit videos, blah blah blah. The more interesting (and ironic) part is: I'm working on a tool to analyze my tutoring videos. I've periodically asked students to record lessons, and students have generally acquiesed (there's no spell check so have ~faith~). Consequently, I have ~60 hours of recorded material; I swear there are a few parts that are endearing and fun and rich. It can't ALL be me, slouched in my chair, eating hard-boiled eggs off my bulbous stomach while my students toil and slave away at their math problems. I mean, it COULD be all that, but that's what we're gonna find out with this cool new tool.
The idea is that I'll feed it a video and it'll find the juicy bits. It should transcribe the videos locally, and then analyze sentiment across the video to find parts where students laugh (or cry). I can then funnel the transcript into an LLM to find patterns that might indicate juicy bitz, which the assistant can inspect with human eyes and decide if there's any juice in those JUICY BITZ. Take those, chop em up, make social media. Also though: analyze video and provide feedback for the tutor. There are no "Tutor Training" programs, and if they exist, they probably suck. A few years ago, I thought it would be prudent to develop a training program (education, I had hoped, would become more decentralized as educational resources passed from the hands of teachers to students directly); however, every person I talked to seemed to agree on one thing: you can't teach people soft skills. You can't take someone who is, for lack of better words, boring and humorless, and teach them how ton be cool and funny and inspirational.
Despite the admonitions, people can still improve. I mean, I just used italics, and I didn't even look at the cheat sheet!
Regardless, the video analysis project could very easily be extended out to be a general first-pass training program for teachers or speakers more broadly. This should exist, and maybe it does, but it probably costs money and it'd be easier to release it as a free executable to run on your own computer.
Video analysis project aside, I have to organize all the left over ideas from past advertising/marketing projects, and distill them into actionable blocks. Once I have dozens of these ideas ironed out (and serialized), it should be easy to assign and execute them.