2025 was not a great year for me. My company (InVision) closed after 13 years. My little Lucy Goosey - the love of my life - died at age 13. I had a midlife crisis. I fell into a depression (which I’m still dealing with). I’ve had nagging tendonitis which is impacting my health goals. And I feel like I’ve barely had a chance to demonstrate my value at PAI (my new place of employment).

Add to that this AI (Artificial Intelligence) revolution / bubble / reality that the tech industry is navigating, and I feel like the solid foundation beneath my feet has turned to sand.

I used to feel almost embarrassed about how happy my life made me. And now, I feel ashamed that I can’t seem to hold onto that happiness; despite the fact that, on balance, I have so very much to be thankful for.

I’m not here to complain - I’m just processing my feelings. And trying to look at what steps I can take to improve my outlook in 2026. This whole post is really just me thinking out loud. I know that I have within me the tools necessary to succeed - I just need to figure out how to move forward in a professional capacity.

"Just In Case" vs "Just In Time" Learning

On the final episode of People I Mostly Admire (podcast), Steve Levitt clearly articulated the two types of learning: "Just in Case" and "Just in Time". Just in Case learning is proactive learning you do on a given subject in case you need that information one day. Just in Time learning is the reactive learning you do in response to a specific problem that you’re experiencing.

I didn’t know that these two strategies had a name. But, both of these learning patterns are helpful for me in different ways. When learning a new framework (for example), I will often try to read a book or the complete suite of documentation on a subject in order to give me a lay of the land. And while I might forget many of the "just in case" details, the breadth of the possibilities will stick with me.

That said, the vast majority of my learning comes from "just in time" scenarios, primarily relating to work and side-projects. Most of the blog posts on this site are written the day after something stumped me and I had to figure out what to do about it.

Going forward into 2026, I want to continue using both of these learning strategies. I haven’t been as good about the "Just in Case" learning these past few years. And while it’s not as effective a teacher, I could certainly do with more of it in my life.

Just In Case Front-End

At InVision, I had to support IE (Internet Explorer) 11. In some ways, this was very comforting because it meant that the goal posts for certain aspects of the product were fairly static. I was able to focus on the product itself; and could put the platform mechanics on the back-burner.

The web (platform) has moved forward quite a bit in the last few years and I need to catch up. But I still operate best within constraints. So my plan is to look at the Google Baseline project and read-up on the widely available JavaScript and CSS features that I’ve missed (those that have been available in all major browsers for at least 30-months).

Apparently I can finally start using at the <dialog> element. Woot woot!

Just In Case Back-End

I need to continue to flesh-out my ColdFusion knowledge. At work (PAI), we use Lucee CFML and are aggressive about upgrading it. In my personal projects, I’m using Adobe ColdFusion 2025 (recently upgraded from 2021). In both cases, my working knowledge of CFML is a bit behind the curve and could be freshened-up.

Additionally, we use Microsoft SQL Server (MSSQL) at work. I’ve been a MySQL person for the last 20 years; and I get by on the T-SQL commonalities; but, I need to learn-up on what makes SQL Server unique.

I also feel like there’s a whole universe of Cloudflare products that would be useful to at least know about. I already use their CDN and Domain/DNS products; but that’s just the tip of the iceberg.

And, of course, AI coding agents. But, that topic is so vast and amorphous that I think it can really only be learned through "Just in Time" learning strategies. Which is where my just in time projects will come to bear.

My hope is that I can pick a just in time project and use it as a context in which to start applying something like Claude Code. I think it would make sense to start with a smaller, more focused greenfield project before I try to apply agentic coding to a large legacy codebase that’s been evolving over many years.

Just In Time Projects

One-page experiments and stand-alone demos have their place. But, my most effective learning comes from building and maintaining hands-on projects. I like that building a project forces me to think holistically about a problem; and I especially like that maintaining a project over time teaches me about the impact of my choices (both positive and negative).

The following are possible areas of focus. These aren’t in any particular order; nor is this an exhaustive list. This is just the bag of ideas that’ve been floating around in my head.

Big Sexy Poems

For the last few months, I’ve been building Big Sexy Poems in public. It’s a small ColdFusion / MySQL / htmx / Alpine.js application that helps me author poems by providing tools for synonyms, rhymes, and syllable counts. It’s been a great playground in which to try out new architectural and code collocation ideas.

Personal RSS Reader

It’s been a really long time since I’ve been good about passively reading other people’s content. When I find articles in Google searches, I read those (yes, I’m one of those weirdos that still uses Google). But, I’d like to try and get back into having some sort of finger on the pulse of the community.

I think it would be fun to build a personal RSS feed reader. Some place that I could store links for future consumption. And then, perhaps on a Sunday, just sit down and consume all the links I’ve stored over the week.

Clean Up My Blog Code

This blog is quite old and crufty. It’s gone through some significant clean-up and refactoring over the years. But, there’s still so much left to do. Plus, I’ve started using a lot of patterns in Big Sexy Poems that I’d like to bring over to the blog. This would make it easier to maintain since all the newest patterns would be front-of-mind.

Plus, I’d really like to replace my use of Hotwire with htmx. Hotwire had too many rough edges; and the Basecamp team primarily build technologies for themselves. htmx feels like a better solution.

Create Syntax Highlighting API

The one part of this blog that I don’t wholly own is the syntax highlighting. When I author a post, I still make an API call out to GitHub to create a gist; and then transclude that gist content back into my blog content.

This has worked well over the years. But I’d love to create my own syntax highlighting mechanics. Either something that works internally to the blog itself. Or exists as a separate API that I can invoke. I just want it to be something that I own and operate so that my success doesn’t rest on the shoulders of another company.

Aside: this also feels like a perfect example of where Agent Coding could be a big help. I don’t have the first clue as to how to deal with syntax highlighting. I’ve looked into it a few times but quickly get overwhelmed.

Clean Up Dig Deep Fitness Code

Much like my blog, the code for Dig Deep Fitness was hacked together over time and is in need of much love. As with my blog, I’d like to take the patterns that I’ve discovered in Big Sexy Poems and apply them here as well, making my three main properties (blog, BSP, and DDF) easier to maintain.

I might also considering making the source code of this project public source (like Big Sexy Poems on GitHub) since that would force me to be more careful with the code. Though, it would have to undergo a thorough investigation to make sure I don’t expose anything sensitive.

Clean Up Incident Commander Code

As with the blog and Dig Deep Fitness, I’d like to get my Incident Commander application code on the same set of patterns. But, this is low down on my list since I don’t deal with any incidents these days.

Write ColdFusion Book

For a while, I’ve been collecting chapter ideas for a "Things Every ColdFusion Programmer Should Know" book. This is a particularly contentious idea since the value-add of reading books seems extremely small (and diminishing) in the age of AI. If I were to try and write a book, it would have to be purely for the fun of stretching my brain’s ability to focus and execute; and, as a context in which to use AI / agentic coding for the mechanics of publishing / constructing the book (ie, generating the ePub and PDF assets or perhaps the "git book" style web experience).

Plus, maybe it’s something that would have a community building aspect. Instead of writing the whole book myself, it could make more sense to have other people write chapters on topics that they were interested in. Then it could be more like an anthology as opposed to just my book.

Micro Podcast Using Voice Notes

I’ve been curious if there’s something I could do in the podcasting space that uses a micro format. I’m especially keen to see if there’s a way that this could be conversational - like an aggregation of thoughts and reactions that all get merged together. I don’t know what this looks like or whether or not it would even be a good consumer experience. But, I think it would be something I could get AI to help me with.

Continue To Generate Non-AI Content

As much as I’m excited / pressured into learning more about AI / Agentic coding, I want to continue to use this blog as a place to generate non-AI content. This blog is me. The words that I write here are the words that are rattling around in my head. I don’t ever want this blog to be the "AI version" of me. There’s plenty of places on the web that I can strike a pose and pretend that it’s me. But this blog will always be the genuine article. What you see is what you get.

With that said, I’m wishing all of you beautiful people a happy new year. May 2026 being filled with great things!

https://bennadel.com/4858