Nail the Message, revisited

In last week's weeknote i talked about nailing the message. This week i finished a peice of work where i needed to reduce a whole of complexity and analysis into a clear set of message. I find this always takes a lot more time than i think its going to.

How i think my time will be taken up:

A pretty even distribution.

How my time is actually taken up:

Lots more time than i ever estimate spent trying to boil the complexity down to some really key messages.

Vibe Coding impasse

Off and on over the last few weeks i've been using V0 to 'vibe code' a tiny little app to solve a problem for me. The problem i was trying to solve is, whenever i'm delivering on a video call, either a presentation or some sort of speech, or have a need to deliver a clear set of messages, in addition to spending time thinking about what i want to say, and i need be, rehearsing, i also make sure i've got some form of script or bullet points to hand on screen. The problem is i end up juggling windows, or flick between windows, or i don't have the script top middle so its a near to the eye line with my webcam as possible, or i end up not paying attention to my audience's reaction because i've got a big window over the top of the Team's/Zoom/etc app.

So, i thought to myself, as an exercise in trying out some of the new AI coding tools, i'd build a really simple autocue/teleprompter app with the following features:

• App always sits front top on my monitor so the eye line is as close to my webcam as possible
• The transparency of the window can be controlled by a slider, so if i want i can see the call in the window beneath the app
• An adjustable autoscroll feature, so as i move through my talk the next line is always at the top of the screen
• a bunch of other basic stuff like loading/saving script files, adjusting the text size and colour etc.

(TBH, i don't know why this isn't just a feature in Teams/Zoom/etc)

Over the course of maybe 4-5 hours of 'vibe coding' over the period of a 3-4 weeks, I ended up building a working Electron app (so i could build a version for Mac and Windows). And i figured, if i've solved a problem for me, maybe i've solved a problem for other people, so i looked into how i'd distribute it, and maybe charge a couple of quid for it. Unfortunately, I soon realised that i'd need to spend a couple of hundred quid to be able to build a properly signed and distributable app for Mac and Windows. I've got no real desire to market or sell an app that may not even meet anyone's need but my own, so i can't really justify spending the money, so i'm now at an impasse. I wanted to be able to go from vague idea through to finished app, but i guess i'll just have to settle for a) meeting my own need and b) learnt a lot about the current vibe coding landscape through using the tools.

BTW, let me know if you do want a really basic autocue app, if enough people express interest then maybe i'll package it up :)

Pre-holiday stress compression

I'm on holiday (well, staycation) this coming week, which is nice. unfortunaterly it concided with quite a bit of work. It made me think how what i've probably done is taken the volume of stress i would normally feel if i were working next week, and squished it into this week just gone. So i've basically pre-paid my stress debt in the week just gone in order to have a stress payment holiday this coming week. There must be a better way.

You cannot respond to or address unexpressed concerns

You (and I) cannot respond to or address unexpressed concerns. Trying to is a fool's errand. I think this needs to become a mantra for me.