Anyone that has spoken to me recently knows I've become a little evangelical about using AI tools to automate processes - for my personal and work life.
I'm often having similar conversations - themes of "I should do more", "I'm using it, but really only to replace google", "I know it can be useful at work, but I don't have the time to learn how to do that" or "I tried it and it didn't work, didn't sound like me".
These aren't people who lack capability - they're smart, experienced professionals who are perfectly able to & are using these tools. But when it comes to moving past asking it what to have for dinner - they haven't worked out how to get started using some of the other baked in capabilities.
Resistance or apathy feels like it is coming from two places:
However when I demonstrate something I've done - and folks can see how easy it is to get started, occasionally something unlocks. It's like when someone showed me a Pivot Table in Excel for the first time, I could see how I could replicate that to solve other problems. The trick is making it relatable and specific, so that it doesn't get overwhelming.
I was discussing with a network buddy, why when folks have access and ability to use AI tools - there is still resistance within teams, departments etc.
If the starting point at work feels too high stakes - there is a risk of getting it wrong or too embarrassed to ask for help.
Of course Claude can help streamline the creation of the board pack - but that feels too big. We discussed what would a low-risk, low-stakes, doesn't matter if it went wrong activity would look like - something like creating an app that counts my use of the word boondoggle (with a leader board & reward badges, of course):
Boondoggle is an actual word - it means a pointless, futile activity - a waste of time and money.
Inspired by that conversation, I started documenting my own experimentation with AI tools under the theme of "boondoggle." It's taken on a bit of a life of its own.
I've used it to test image creation with Gemini, build pointless apps with Claude, connect tools together using Canva and Excel add-ins, pasting HMRC links into NotebookLM to answer my hairdresser's question about Making Tax Digital.
While on a bus heading home from a ski trip, just using my phone - just for the joy of it, I created this app using Claude Code, got it to push it to github and publish on Netlify for me: boondoggleski.netlify.app
Previously, I'd get frustrated with something - I'd have an idea about how to fix it - but give up on it because without coding skills it was too hard to do anything about it. The crazy thing is that "app" was created end of February - things have advanced since then, the interpretation of requirements just using natural language and MCPs connecting different tools is mind blowing how much simpler it is to do things just a few weeks on - even with free versions of tools.
It's very unlikely that I will properly publish a version of my Ski Tracking app - I'm just chuffed that I worked out how to get Claude to sort out Github and Netlify for me just using my phone and voice commands.
Showing people examples like this - even without developer/coding skills - it is possible to create things. Encouraging low-stakes experiments end up progressing much faster than those who try to jump straight into work. In going through the process, you're learning vocabulary, working out how to connect stuff - once sorted, it isn't such a leap to start applying it to work life.
In this video I walk through some of my boondoggle experiments - image creation, app building, brand development across multiple tools, and more:
Follow me on LinkedIn, where I first published this blog.
I've been running workshops covering Claude basics - projects, skills, connectors, connecting to your files, and how to build your own training plan. The full write-up with demo videos, copy-paste prompts, and step-by-step guides is here: Introducing Claude: a no-jargon tour of menus and features
For the hands-on version: Introduction to Claude & Cowork
