Currently reading: Sum: Forty Tales from the Afterlives by David Eagleman 📚
Currently reading: Sum: Forty Tales from the Afterlives by David Eagleman 📚
I kept seeing people build wild stuff with OpenClaw - agent swarms, automated research, specs that write themselves, etc. So after a lot of back and forth, I had to try it. I spun up my own setup, named it Q, and started wiring it into a real project. So here’s a quick tour of how the system actually works, why it’s useful, and how the pieces talk to each other.
First two weeks of 2026 have been crazy on the AI front. There’s been a lot of Claude Code talk (at least in my bubble) and it feels like every day you have to learn something new… and unlearn something else…. and that’s equal parts exciting and anxiety-inducing.
I’ve never been the kind of developer who treats code like a craft to perfect but more like a tool to build what I want (I’m a mediocre dev at best)… and now that AI is removing a lot of this “writing code” friction, I think what will remain valuable is how you think: problem framing, tradeoffs, and turning messy reality into something clear, elegant that’s actually providing value for the end user. That’s the part that excites me.
The anxiety comes from not knowing whether what I build today will be obsolete tomorrow, given how fast everything is moving.
Currently reading: Who Knew - Barry diller by Barry Diller📚
Starting today TWO is live on both the App Store and Google Play It’s been some back and forth with the Apple’s review process and the fifth time was the charm. Five rounds of review. Five cycles of small fixes, vague feedback, rewrites, and patience. First 2 tries were totally on me because I had a login bug on iPad I definitely missed. The next 2 were : paywall stuff where yearly tier showed a per-month price that was more prominent than the actual charge… which is against their guidelines.
Last week I had a simple idea: what if I could get weekly spending reports, with AI automatically categorizing everything and ideally with zero manual input (or as close to it as possible) So I built this Skill (or better said Claude Code build one) that’s: → Connecting to my bank via GoCardless open banking API → Pulls all transactions automatically → Categorizes spending (groceries, utilities, transport, etc.) → Calculates my actual savings rate → Shows me week-over-week comparisons
21 Lessons From 14 Years at Google
Before you build, exhaust the question: “What would happen if we just… didn’t?” Sometimes the answer is “nothing bad,” and that’s your solution.
The problem isn’t that engineers can’t write code or use AI to do so. It’s that we’re so good at writing it that we forget to ask whether we should.
Great read and a great reminder that action beats overthinking, overplanning, and overengineering. Start with something, see it through then iterate and build on top of it.
Simon Willison’s third annual series reviewing everything that happened in the LLM space in 2025
Great recap of how the AI landscape unfolded in 2025. So much has happened that already feels like ancient history (remember the o1 models?) A good reminder of just how fast this space is moving.
→ What Apple’s LLM Fumbles Say About LLMs (Rather Than About Apple)
No. Apple doesn’t offer an AI that understands you, your life, communication, location or how you use software. And no other company fully does it…yet.
That said - several companies already deliver AI experiences that are clearly better than anything Apple has shipped so far and that’s the embarrassing part…and a real sign Apple is actually falling behind.
The fact that they are the most well positioned company to actually make AI the ubiquitous yet still not seeing anything remotely comparable…one mroe sign.
Watched: Eternity 🍿
I was intrigued by the plot and it’s a fun movie, though pretty easy to guess how things unfold.
Spent the last few days going through my notes, todos, calendar entries etc to try and do a 2025 review.
Every year I’m beating myself up for not writing more.
It’s fascinating to me to see what I was thinking even a few months back.
So starting this blog is also a way to to try and write more of my thoughts, ideas, state of mind etc.
This year I managed to launch three apps. One failed. Two are still around and are taking up all my time. Heal -> tryheal.app - is a breakup recovery app built for people going through the messy aftermath of a breakup. The core idea was simple: give people structure when their thoughts are anything but structured. Journaling, reflection and an AI coach that acts more like a calm confidant than a motivational speaker.
Watched: Jay Kelly 🍿
Made me think about work-life balance I’m struggling with sometimes, and how easy it is to overlook what’s important when you’re focused on “winning”. I hope when I’m old and grey to be able to look back and have fond memories with my daughter not necessarily my work
Currently reading: Excellent Advice for Living: Wisdom I Wish I’d Known Earlier Hardcover by Kevin Kelly📚
Picked this up and it’s been such a simple, great read. Highlighted something almost every page.
As part of my 2026 goals to start a blog, I just bought a micro.one yearly subscription. The plan is to write about whatever - thoughts, ideas, apps I’m building, books I’m reading, movies I’m watching etc.