Jacob Brooks

4/12/2026 Weekly what's up

Sun, 12 Apr 2026 00:00:00 -0400

What I’m reading

I finished the Book of Dust, Philip Pullman’s followup to the “His Dark Materials” trilogy. One part prequel, two parts sequel, it follows Lyra (the main character from the original trilogy) in the 10 years before and after the original setting. It didn’t quite stick the landing, but I still enjoyed it and recommend reading both series to anyone who hasn’t (or hasn’t read them since they were a child, like me). I think that the first book, the prequel, was my favorite - just a great, gripping self-contained story with fun connections to the original books. The story in the sequels is also good (action/thriller/intrigue type stuff) but I think Pullman stumbles a bit when trying to elevate and wrap up the philosophical side of the series. Throughout the sequels there is a promise of learning the true nature of some of the fundamental forces in this world but he doesn’t stick the landing. The last 20% of the book is a lot of characters monologueing about their interpretation of what they’re learning as a means of exposition, with the exposition itself not being all that compelling. Still, though, I’m glad I read it all!

What I’m playing

Got a 3-player game of Slay the Spire 2 in with Dad and Matthew - very fun! We managed to beat the 3rd act boss, my first time doing so after ~8 rounds of solo play. Maybe I’m bad and they carried me (likely), or maybe the game is currently more tuned for multiplayer. Very fun, want to play more and I hope they add more multiplayer-specific cards and interactions.

4/7/2026 Weekly what's up

Tue, 07 Apr 2026 00:00:00 -0400

What I’m reading

It’s 17776 day! Seeing this upcoming I re-read this wonderful… alt-media piece?.. from Jon Bois. Still a delight years later, and thrilled to see that there is a book in the works that will hopefully conclude the story from 20026. If you haven’t read before - open and read on a computer, not your phone!

What I’m watching

Finished up Jury Duty season 2. Still feel that overall it’s not as strong as the first season, but got better towards the end and they got the payoff for the big reveal. There’s no way they can do it a third time, right??

3/29/2026 Weekly what's up

Sun, 29 Mar 2026 00:00:00 -0400

What I’m playing

A little bit of Slay the Spire 2. It’s good… very much “more slay the spire”. All of the new character are great, need to try co-op mode in the next week or two. If it wasn’t for economics, bringing in multiplayer is probably the driving reason to release this as a sequel instead of a DLC?

What I’m watching

Jury Duty season 2: Company retreat is pretty good so far, not quite hooking us in the same was as the first season did. Can’t believe they managed to pull it off again!
Not watching baseball but… on day 1 the new ABS stuff is absolutely electric. “He’s arguing with the robots, you can’t defeat the robots!” is immediately all-time, calling for two strikes in a row, it’s so good. Is this the best long-term rule? Feels like we should just use the robots, but for now this is seriously entertaining.

3/15/2026 Weekly what's up

Sun, 15 Mar 2026 00:00:00 -0400

What I’m reading

I finished His Dark Materials and am halfway through The Secret Commonwealth, the second book in Philip Pullman’s The Book of Dust follow-up trilogy. The original books were a joy to re-read, and so far I am loving the followups. La Belle Sauvage was probably my favorite book in the series so far - gripping and absolutely lovely characters.

3/11/2026 Weekly what's up

Wed, 11 Mar 2026 00:00:00 -0400

What I’m reading

The 2028 Global Intelligence Crisis from Citrini caused quite the tizzy online and in the markets. Take it for what it is - a thought experiment for one possible scenario. Is this specific path probable? I don’t think so. But there are elements here worth considering!
The Anthropic/DoW saga has been quite something. I want to write more about it in a dedicated post, but here I’ll share a couple of good pieces that resonated:
- Ben Thompson’s Anthropic and Alignment
- Dean W Ball’s Clawed

What I’m watching

A Knight in the Making was a fun behind-the-scenes look at the show. Set design was incredible - they built an entire bridge!

What I’m up to

Busy couple of weekends - we visited our friends Mitch and Alli in Phoenix, and Kaitlin was away in Joshua Tree for Kiran’s bachelorette party. Phoenix was lovely - nice to get some sun and see friends, and we had a lot of fun at the zoo (though the heat got to us in the end). But good to be all home together now.

What Rory is doing

All he wants to do is “swing” (he repeats, again and again) his golf club. Big meltdown last night - he was not satisfied swinging his club in our backyard, he needed to get in the car and go to the driving range. The only way to calm him down? We put on Every shot from Rory’s final round at the 2025 Masters. He is my son!

2/22/2026 Weekly what's up

Sun, 22 Feb 2026 00:00:00 -0500

What I’m reading

Child’s Play was a fun read. Watching the Donald Boat saga unfold in real time was surreal and this was a good retelling. Author is pessimistic on AI (which authors in this milieu aren’t?), but overall a good examination of viral marketing culture.

In a way, Donald Boat had achieved the dream of every desperate startup founder in the Bay Area. He had propelled himself to online fame, and used it to relieve major investors of their money. But somehow he’d managed to do it without ever once having to create a B2B app. He was a kind of pure viral phenomenon.

What I’m playing

Still playing Cairn, though it can be throw-your-controller frustrating. Lost ~30 minutes of progress last night, brutal! Writing that, though, makes me think I’ve gone soft - 20 years ago I think that’d be a common occasion?

What I’m writing

I wrote up our ski trip: Idaho ski trip.

Idaho ski trip - McCall

Sat, 21 Feb 2026 00:00:00 -0500

We took a short trip to McCall, Idaho last weekend in search of snow with our friends Kincaid and Ellie. Why McCall? Mostly convenience - a cheap flight for us, a six hour drive from Bend for them, and two resorts on Kincaid’s Indie Pass. It ended up being a great call though as we had a lovely stay in a small town and two great ski days at Brundage and Tamarack.

We all got in late Friday night and after a leisurely start drove 15 minutes up the hill to Brundage on Saturday morning. After some snow earlier in the week the resort was 100% open with decent coverage everywhere except for off-run on the lower parts of the mountain. We all explored the skier’s right side and Lakeview backside in the morning, finding pockets of decent snow but mostly ripping the groomers. The long lifts let you get in some long runs, comparable to Northstar or any other big resort.

After lunch Kaitlin tapped out and Kincaid, Ellie, and I went up to the hidden valley side to explore some of the more interesting terrain. This was a super fun area, but tough to explore on a first visit and with moderate-low coverage. Rocks and cliffy rollovers were all over, making it tough to commit to any line from the top down.

Cliffy, fun rocks

On our run out there we looked up to the ridgeline and saw what looked like some promising snow further afield, so at the bottom swapped skis and grabbed our backcountry gear for a bit of exploration.

Kincaid and Ellie on the ridge traverse

A lot of traversing on the ridge led to… mostly disappointment. Clouds came in and visibility was pretty low, so again we were a bit too nervous to commit to any of the protected lines in the cliff bands not really knowing what was below us. We traversed out until we found more open terrain.. and pretty mediocre windblown snow. Oh well! Still fun to explore.

Ellie dropping in

On Sunday we drove south to Tamarack, a much more developed resort and ski hill. It felt like a Northstar or Olympic village - big hotel/village area with retail and a spa. It was a great day to ski, but unfortunately Tamarack needed another couple feet of snow to open the whole mountain. On the first lift up we could clearly see all of the summer mountain bike park features, yikes! In practice, only the summit lift was open meaning everyone was skiing a single lift. 15-30 minute lift lines for most of the day put a damper on what was otherwise an awesome hill and great day.

Looking down at the lake

Once again we were tempted by lands unknown and did a bit of hiking out the backcountry gate in search of better snow. Mixed results again - we found three or four turns of nice powder but above that was too blown and it quickly got warm/heavy below.

Looking at the next ridge over

Kincaid dropping in

We stayed in the resort for the rest of the day, finding good snow in pockets and finally getting some hot laps in after people started to clear out around 2:30.

Kincaid skiing

That evening we ventured to the Shore Lodge in search of a pre-dinner cocktail with a view but struck out - busy weekend!

Shore Lodge

Mountain reflections

After a bit of wandering we ended up at the locals pub Foresters. Great pizza, good beer, and the Olympics on TV was a great way to cap off a lovely weekend! McCall was a great little town to visit, I’d certainly go back. Small, independent ski resorts like these are becoming my favorite way to ski inbounds.

2/18/2026 Weekly what's up

Wed, 18 Feb 2026 00:00:00 -0500

What I’m reading

Finished The Adolescence of Technology. There are many possible readings of a piece like this from someone in Dario’s position, but my belief (70%?) is that it is genuine. Naive, maybe - it does feel good to believe that someone in power is thinking seriously about the risks of what is coming.

What I’m playing

Started Cairn on the way to Idaho, very impressive. Probably the best climbing mechanics I’ve played with, it is challenging and feels realistic but is fairly natural to control.

What I’m watching

We blazed through the first five episodes of A Knight of the Seven Kingdoms in one evening. Great stuff, reminds me how good Game of Thrones can be. Maybe I should watch House of the Dragon? Whenever GRRM gets the next book out I’m looking forward to a re-read of the series.

What I’m up to

We went to McCall, Idaho for the weekend to ski with Kincaid and Ellie at a couple of small resorts. Fun small town, nice to have a couple of days to ski and relax but we missed Rory! Won’t leave for that long again for a while.

2/8/2026 Weekly what's up

Sun, 08 Feb 2026 00:00:00 -0500

What I’m reading

I hate fish from Rand’s resonated deeply in my bones. I am feeling in a rut with my Omnifocus-driven productivity system, and likely going to try and re-think from (my own) first principles and vibe-code something else in the next few months.
How Food Delivery Is Reshaping Mealtime was painful to read. I thankfully still have a visceral distaste for DoorDash/etc. and try to avoid it. We usually use it when hosting friends/family which feels more economical.

What I’m watching

Catching up on season 2 of Cody Townsend’s Fifty+ series to build the stoke for some upcoming skiing. Best adventure-skiing content out there!

2/1/2026 Weekly what's up

Sun, 01 Feb 2026 00:00:00 -0500

What I’m reading

Working through Machines of Loving Grace and the follow-up The Adolescence of Technology. It’s rare to get this kind of writing from the leader of one of the (5? 3?) most important companies on Earth and I deeply respect it.

What I’m watching

Nothing! I don’t remember the last TV show I watched all the way through, probably Andor season 2? I’m not unhappy about it, but would like to watch more movies. Need to make time for this.

What I’m up to

I got Specmark to a place that I’m happy with and wrote about it here: Specmark: annotating markdown specs for AI feedback. Fun process, learned a lot about end-to-end AI-driven development.

Specmark: annotating markdown specs for AI feedback

Sun, 01 Feb 2026 00:00:00 -0500

The idea came to me in the AI coding frenzy this holiday season as I was burning through the extra Anthropic API limits and experimenting with development via my iPhone and Termius. For bigger features, I try to follow a spec-driven development model: Ask the agent to write a spec and iterate on it before actually writing any code. This is what Claude Code’s Plan mode does, and you can get pretty far with that and just interactively asking questions.

There comes a point, though, when you really should actually read the whole spec. At least, if you care a bit about the code or want to give the implementing agent the best chance at one-shotting the feature. I’d do this in the editor, which was fine for making changes but if I wanted to give the planning agent broader feedback a couple of things annoyed me:

It’s hard to reference specific lines. You could type out “in the auth section…” but that’s extra typing, exactly what I’m trying to avoid!
It was practically impossible in Termius, and any extra typing there is more painful.

I thought of how I do this professionally - we have a pretty strong writing culture, and I am always reviewing documents and adding inline comments in Word or Confluence. Why not bring that same experience to editing markdown?

What is Specmark?

Specmark.dev is a lightweight tool that allows the user to:

Paste markdown content
1. Or, “teleport” it to the site from anywhere using short-lived share codes
Annotate the rendered markdown content with comments, a la’ Readwise Reader or any review mode like in Word or Confluence
Copy the comments made with context like line numbers and the quoted content to the clipboard in order to be fed back to an LLM.

Everything except for the share-code feature is local in the browser. There is also a small cli tool to make the share code easier to use and import markdown files directly.

Here’s the code: github.com/jlbrooks/specmark. This is also an experiment in fully AI-driven development — 99% of the code was generated without detailed review.

Challenges building it

The highlighting interaction took way more iteration than I expected. I had a kinda-crappy version working within a couple of hours — maybe 25% of the total time I spent on the project. Getting to “done” was the real work. This is an area I have no familiarity with, and is not a super common thing - maybe not “in distribution” for LLMs compared to other interactions. I spent too much time in some form of “make it better!!” loop which was not fun and not efficient. I eventually got to an ok place by giving it browser access and pointing it at the Readwise Reader interface for inspiration.

The other area where I would have struggled was design. The first iteration website was… fine… but not great. Thankfully I am married to a fantastic visual designer, and Kaitlin whipped up the outline of what you see on the site now in an hour or so in Figma. I pointed the LLM at the Figma MCP and the look and feel of the site increased tenfold.

What I learned

Human expertise is still critical. The highlighting experience was instructive for me. Many of my LLM interactions on software until now have been on topics that I have a decent understanding of, know the pitfall, know what decent approaches look like, etc. When that’s true, it’s very easy for me to:

Set the agent on the right course from the start
Course-correct when things are getting off the rails
Throw away bad implementations and try again

I had none of this expertise (or “taste”) on highlighting in the browser. This made development so much harder as I was flying blind. In hindsight, a better path may have been something like:

Use the agent to try and learn the space. Have it go read APIs, blog posts, etc. and come back with reports.
Actually read and internalize this myself. I think it’d be easy to skip this step and just feed the report into the implementation planning agent, and that might work pretty well, especially for a small side project. But for bigger important features and long-running projects that will need to evolve over time I believe that it is critical for the human operator to internalize these details.
Take those learnings and start the implementation as usual - write a spec, review, iterate, and then execute.

Maybe another year of model development will make this irrelevant, and I’m sure that there will be many successful projects where the operator has little to no understanding of the internals or interfaces used. Hell, this is kind of true in general of software - we build on abstractions, most web devs do not know and do not usually need to know the details of TCP networking or TLS. But those that do are more successful and more valuable, and I believe that will continue to be the case.

What’s next?

I’m pretty happy with the tool as-is and don’t have any plans for the web version. An obvious next step is some sort of native experience - why can’t I do this in my editor or terminal? Could a history of these comments be persisted somehow into the project history so that future agents can internalize the feedback?

I think that this broadly falls in a similar space as “local code review of agent work”, a problem that I know many people are working on. Excited to see how the space develops, maybe I’ll try to make my own contribution.

1/25/2026 Weekly what's up

Sun, 25 Jan 2026 00:00:00 -0500

What I’m reading

Re-reading The Golden Compass for the first time since I was a child. Spurred by learning (h/t Triple Click) that Pullman finished the follow-up trilogy The Book of Dust in October of last year. It’s just as weird as I remember it, looking forward to digging back in.

What I’m up to

I hope to have solved my ski boot problem in the most brute-force way: buying new boots. Had a great experience at Olympic Boot Works this morning - the fitter immediately recognized my issue with forefoot pain in the current boots and put me in a new pair (Technica Cochise 120) that felt better with a stock fit than I’ve felt in boots in years. Fingers crossed, will find out next weekend.

What I’m writing

I wrote up my 2026-01-25 2026 theme, objectives, and key results

What Rory is doing

He is more fun to interact with than ever, but starting to be more challenging than ever. Making his will known; little breakdowns are regular every mealtime when he doesn’t get exactly what he wants. Hardest part: Not laughing when he scoops food off his plate and dumps it on the ground while looking you in the eye.

2026 theme, objectives, and key results

Sun, 25 Jan 2026 00:00:00 -0500

For the past few years I’ve tried to set some year-long goals according to a theme. This year: Go deeper and finish things. I’m drawn to this theme because I tend to start a lot of projects, dabble in hobbies, and leave things 80% done. This year I want to resist the urge to chase novelty and instead commit to finishing what I start - especially important for tech projects in this age of rapid experimentation and prototyping!

Did I structure them in the form of OKRs, because that’s what I know best? Yes. Is that extremely dorky? Also yes. I’ll be doing monthly check-ins (monthly personal reports? MPRs?) to myself and sharing with Kaitlin to look at progress and identify changes to routine if things are off-track.

Fitness

Objective: Get back to running consistency and build full-body strength

2024 was the year I became a runner, ending at around 750 miles total driven by a half marathon trail run in April. In 2025 I fell off, not able to build a good routine with more work around caring for Rory. This year, I want to return to basics: Run consistently, lift regularly, and keep the foundation strong. Key results:

Run 500 miles
Lift 2x/week (104 sessions)

Specific action to help drive this: Put workouts on my work calendar at the start of the week.

Software projects

Objective: Finish and polish software projects to the point where I want to share them with the public

I have a graveyard of half-built projects. This year I want to push through the uncomfortable “last 20%” and actually ship things. Key results:

Launch 4 projects on my blog and Hacker News
Launch 1 app that has billing

Board games

Objective: Play more board games and ensure my collection is intentional

I’ve accumulated a shelf of games over the years, but have lost the habit of regular play. Going through the collection is a fun exercise and will help me prune those that I don’t want to play again. Key result:

Play every game in my collection (55 games)

Golf

Objective: Develop my golf game with intentional practice and professional feedback

Golf has become one of my favorite activities, but I’ve plateaued. Time to get serious about improvement. Key results:

Get coaching feedback 2x/month during the golf season (Q2/Q3)
Practice 1 hour for every round I play

Writing

Objective: Expand my writing habit using my personal blog, writing more and about more in-depth topics

Last year I managed 28 weekly updates (54% of weeks) and 6 longer posts. I want to be more consistent and push myself to write more than just my weekly updates. Key results:

Write weekly updates for 90% of weeks (46/52)
Write one non-weekly-update blog post per month (12 total)

1/18/2026 Weekly what's up

Sun, 18 Jan 2026 00:00:00 -0500

What I’m up to

Early stages of setting up a personal assistant via Clawdis. Goal is to have something connected to calendars, tasks, email, Dropbox/Obsidian, with some web search capability for Kaitlin and I to use together. I’m sure that there will be products (maybe from the major model providers) like this soon, but I’m a big fan of self-hosting something like this.

What Rory is doing

Getting much more dextrous, going very fast on his bike and tossing a ball back and forth for ~5 turns. Super fun, glimpse of what is to come!

1/11/2026 Weekly what's up

Sun, 11 Jan 2026 00:00:00 -0500

What I’m reading

How I Stopped Being Sure LLMs Are Just Making Up Their Internal Experience was a thought-provoking read. On a spectrum from “They’re just next token predictors” to “They are alien intelligences that should be treated with respect and have rights”, I’ve been slowly moving towards the alien interpretation. I’m not fully there yet, but I think that within a generation or two of models and with some soon-to-come capabilities (online learning) we should be treating them as sentient life.

Will we? Knowing humans, probably not. I am certainly starting to treat them with respect, and will implore others to do the same.

What I’m playing

A quick game of Sky Team up at the Tahoe Cabin, need to play through more this year to get to some of the more difficult landings.

What I’m up to

We got up to Tahoe this weekend and I got my first ski day of the season in at Alpine Meadows, a nice 17k foot day with blue skies and warm temperatures. Felt pretty good, though I seriously need to get my boot situation figured out - can only last for 3-4 hours while buckled in before crippling pain and needing to get out of the shells.

What Rory is doing

Had fun in the snow at Tahoe! He was trepidatious at first but warmed (heh) up to the experience. We got him going in the sled and he had a great time. Looking forward to getting him on skis for real next year.

1/4/2026 Weekly what's up

Sun, 04 Jan 2026 00:00:00 -0500

What I’m reading

Unhinged stuff from the creator of Beads: Welcome to Gas Town. I’ve been using beads and I like it, though it feels a bit excessive, token heavy, and over-engineered at times. Love that it’s just local and uses git, nice for syncing between agents and having them create issues themselves.

Gas town builds on beads and takes it to another level: a full-on agent orchestration system. He analogizes to Kubernetes, a comparison that resonates with me. Explaining the core of what Kubernetes is was a favorite interview question of mine for prospective engineers with infra focus: it’s a state reconciliation machine. You say what you want, that is stored in etcd, and it has many controllers that try to make that state a reality. Gas Town does the same but for… actual feature development with multiple agents coordinating? It’s a lot. I mean..

If I tell the Mayor, “Our tmux sessions are showing the wrong number of rigs in the status bar — file it and sling it”, the Mayor will file a bead for the problem, and then gt sling it to a polecat, dog, or crew, depending on how it feels as a factory chimpanzee that day.

He describes 8 stages of AI development:

  Stage 1: **Zero or Near-Zero AI:** maybe code completions, sometimes ask Chat questions
  Stage 2: **Coding agent in IDE**, permissions turned on. A narrow coding agent in a sidebar asks your permission to run tools.
  Stage 3: **Agent in IDE, YOLO mode:** Trust goes up. You turn off permissions, agent gets wider.
  Stage 4: **In IDE, wide agent**: Your agent gradually grows to fill the screen. Code is just for diffs.
  Stage 5: **CLI, single agent. YOLO**. Diffs scroll by. You may or may not look at them.
  Stage 6: **CLI, multi-agent, YOLO**. You regularly use 3 to 5 parallel instances. You are very fast.
  Stage 7: **10+ agents**, **hand-managed**. You are starting to push the limits of hand-management.
  Stage 8: **Building your own orchestrator**. You are on the frontier, automating your workflow.

This past week I’ve been between stage 5/6 - I guess I need to up my game before trying out something like Gas Town ;) And also up my Claude subscription…

What I’m playing

I have a 2026 goal to play every game in my collection (burn down?). This past week: Two plays of Patchwork. Such a tight little two-player, I love it!

What I’m up to

Writing (telling agents to write) lots of code! This has been a very fun two weeks, enabled by the enhanced Claude usage limits and my Termius->Claude-box connection. So many things seem possible now, and I feel very much like the bottleneck to getting stuff done - agents can write code faster than I can give them instructions! I’ll be continuing to invest in improvements to this workflow and raising my level of abstraction throughout the year.

But - I want to actually ship too, and not get nerd-sniped every other day by a little side idea that I spend 2 hours on then leave in the dust. Another 2026 goal is to actually finish and ship stuff. My thinking right now is 4 projects (1x/quarter) that are polished enough to write up on the blog, permanently post to my projects page, and post on HN/equivalent. Plus a stretch - one project that I hook up some form of monetization/payment. Gotta pay for those tokens somehow!

What I’m writing

I wrote up my favorite video games of 2025. Next up - an outline of my 2026 theme/goals, and a full writeup of the Specmark app.

My favorite video games of 2025

Fri, 02 Jan 2026 00:00:00 -0500

Just like in 2024 I want to recap what stuck with me this year. A great year for games!

GOTOY (games of the other year) honorable mentions that I played this year

Ridge Racers 2 (2006)

What a treat! Inspired by Into the Aether gushing about this one on their PSP episode, I started playing on PS5 and was hooked. It’s simple but it feels perfect - constantly drifting, you learn the tracks and find the zone.

Death Stranding Director’s Cut (2022)

Knowing that the sequel was releasing this year made me re-visit. I had tried to play the original in… 2020? And bounced off. But the first needle drop this time (Don’t be so serious, my goodness what a track) had me hooked, and the touching story propelled me through the long cutscenes to the end in just under 30 hours. It probably resonated with me more now that I’m a father! (Lou!)

Astro Bot (2024)

After this won all sorts of awards last year I had to check it out, such a joy. Pure platforming bliss through the whole thing, constantly changing things up Mario Odyssey style to keep it fresh.

1000xResist (2024)

Probably my favorite game this year, a masterpiece of Sci-fi storytelling and soul. Is “are videogames art” still a conversation? This is art.

Games Released in 2025

(6) Avowed

Enjoyed this, enjoyed being back in the world of Pillars even though I did not finish (got to the 3rd big zone, I think?) Probably the best 1st-person magical fantasy combat I’ve played, this sets the bar for upcoming games like Elder Scrolls 6.

(5) Kingdom Come: Deliverance II

A late start for me, and I’m still early on, but this is right up my alley and I’ll be picking away at it for a while. I love the setting, I think there is so much opportunity for historical RPGs like this that aren’t fantasy or science fiction. Take me to the Mughals in India! Or ancient Egypt! Yes, assassins creed exists, but what makes this great is how grounded it is - truly a “role-playing” game.

(4) Death Stranding 2: On The Beach

Finished the first one just in time to start the sequel. It builds on it in a lot of great ways - the combat is better, the menus are easier to navigate, the world is more pleasant to navigate, and it is stunning to look at. And there’s an adorable baby. I did like this a lot, but tapered off maybe 75% of the way through. I think that sanding off some of the rough edges didn’t help me connect in the same way? I was not a completionist in either, I don’t want to grind for the pizza guy, I wanted to persevere through adversity to see the story. While it’s quite pleasant to play I didn’t feel that challenge here.

(3) Blue Prince

This was an incredible surprise to start the year. Such a cool twist on the roguelike genre, every puzzle and mystery solved is as satisfying as any other progression hook. This is one that I need to get back to - I rolled credits but did not go further, and I know there’s so much more to see.

(2) Hollow Knight: Silksong

It’s Hollow Knight, but feels great from the very start and has a spunky protagonist. What more could you want?

This game is quite a challenge and I’m not sure I’ll get to the end - I beat act 1 but took a break due to our move/life stuff, so another one I need to get back into. What I did see was stunning though, it’s amazing what Team Cherry is able to do with light-touch environmental storytelling and minimal dialog.

(1) Sektori

And, at number one, a December surprise to finish the year! This is the kind of game that puts me in a trance (but it’s not trance music - it’s techno) as I try to chase perfection and another high score. It fills the screen with light and motion but somehow you are able to parse it all, or at least you think you can - or you surely could on the next run.

This didn’t grab me until a few runs in, but once it did I couldn’t think about anything else. It’s the reason why I haven’t returned to anything else on this list in the past month, every time I open Steam it calls to me. This is a masterpiece of an arcade game, every choice is so intentional and everything is so tightly integrated to make a beautiful whole. Go play this game!

12/29/2025 Weekly what's up

Mon, 29 Dec 2025 00:00:00 -0500

What I’m reading

A couple of good articles on agentic coding:

My experience with Claude code 2.0. Pretty good overview, nothing too revolutionary here but a great starting point if you haven’t been following Claude Code throughout the year.
Shipping at Inference-Speed. One for the Codex-heads! I like the emphasis on simplicity, I do feel validated not going all-in on the skills/plugins/etc. hype (“30 sub-agents to run your business”). The model capabilities and harness catch up quick! Will be working Codex more into my personal projects.

What I’m playing

Some boardgames! Need to get back into this habit and make more time regularly. Burndown through my collection in 2026 is a goal. Recent plays of Patchwork and Tapestry with Kaitlin, and Terraforming Mars with my dad and Greg.

What I’m up to

All-in on agentic coding over the holiday break. Currently working on a portfolio tracker, as nothing I’ve found quite meets my needs. It’s the era of personal software, baby!

Particularly fun has been coding from my phone. I set up an Ubuntu server box on my home lab, connected it and my phone to tailscale, and talk to Claude Code via Termius on my phone. I can ask it to make changes and see updates in real time to the app running on my local network.

Semi-related, I was frustrated trying to develop on my Windows machine with WSL - too many networking annoyances and little gotchas. I was debating buying a Mac Studio but had the thought “why not linux?”. One afternoon later and my daily driver is Pop!_OS and I’m not looking back. Gaming was the main thing keeping me on Windows, but with Proton/Steam Deck development over the past few years Linux support has gotten extremely good. Is 2026 the year of the Linux Desktop?!?

12/21/2025 Weekly what's up

Sun, 21 Dec 2025 00:00:00 -0500

Wow, 6 weeks since the last update! Not quite a weekly update, hope to get back into a rhythm now that we’re settled into our new (old) house.

What I’m reading

Started Too Like the Lightning - very unique so far. Have heard it compared to Neal Stephenson’s Baroque cycle so excited to get further in.

What I’m playing

Sektori has its claws in me, such a fun loop but so hard! Recommendation: Don’t look anything up, it intentionally does not tell you very much about the mechanics of scoring and that discovery is a part of the fun.

What I’m watching

We continued our Paul Thomas Anderson kick with Boogie Nights. So good, probably my favorite so far. Read the Grantland Oral History after watching!

What I’m up to

Moving back in means lots of projects! Heavy-duty floating shelves in my office vexed me for a bit, learnings: Don’t use the drywall anchors that come with shelves, and really don’t use drywall anchors. Just find the studs.

What Rory is doing

Talking, moving, talking. Feels like he’s getting a new word every day, super exciting to see this progress after not having much to say.

11/10/2025 Weekly what's up

Mon, 10 Nov 2025 00:00:00 -0500

What I’m reading

Reading more of Rand’s leadership blog, Become the Consequence was really good. Need to practice more of this! Picked up his “Art of Leadership” book, mostly a collection of essays from the blog but it seems like a good highlight reel.
The Hatred of Podcasting was pretty good.
- “Of course, most aren’t actively listening to Fridman and Huberman for two hours at a time. The episodes are a layer of white noise, a way to blot out thoughts. Podcasts came of age amid the growing absence of meaningful contact in the average person’s day, in a time when silence is hated.”

What I’m playing

More Ark Nova! Want to break out Terraforming Mars again to compare how they feel.

What I’m watching

A long flight top India was good movie time.

Napoleon (2023) was… just ok. Not a fan of using Josephine as the sole lens to view the man, it felt pretty reductive.
Tenet was incredible!? I don’t know how I didn’t see this until now. Need to watch again.

Also watched Poor Things which was, uh, quite something. Interesting film about women’s liberation, I think there’s an essay to be written comparing to Barbie?

What I’m up to

I traveled to India for work last week. Great to visit and meet people in person for the first time, cool to visit a country (and continent!) that I’d never been to before. Highlights: The food, Mughal forts in Jaipur, and the food.