The short version: I just got made redundant, and I’m going to use this as a forcing function to finally go deep on AI.
This blog exists to make that commitment public and to keep me honest.
Why AI, and why now?
AI has quietly moved from a “nice-to-know” to a fundamental capability for builders. I want to:
- understand the foundations (probability, optimization, representation learning),
- build practical systems end-to-end (from data to deployed models),
- and stay close to the rapidly shifting tooling ecosystem (LLMs, vector DBs, agents, orchestration frameworks, etc.).
Writing regularly is my way of slowing the firehose down enough to actually understand what I’m learning.
What I’ll write about here
Expect a mix of:
- Foundational notes – things like cross-entropy, gradient descent, attention, and how they show up in real systems.
- Project logs – small end-to-end builds: from toy models to “real” prototypes.
- Reading notes – papers, courses, and books distilled into practical takeaways.
- Meta posts – like this one, reflecting on how the learning process is going.
Nothing here is meant to be authoritative. Think of it as a lab notebook that just happens to be public.
My constraints and strategy
Being made redundant means I have:
- more time (for now),
- finite runway,
- and strong motivation not to waste either.
My strategy is simple:
- Ship something small every week (code, a notebook, or a post).
- Bias toward implementing ideas, not just reading about them.
- Use this blog as a place to explain ideas to my future self.
If you’re reading this and on a similar path, I’d love for this to be useful to you too.