#052 - The second time around
A look back on Complex Machinery's 2025.
You're reading Complex Machinery, a newsletter about risk, AI, and related topics. (You can also subscribe to get this newsletter in your inbox.)

Welcome to 2026!
We're starting the year off with some leftovers. Or recycling. Or maybe a hangover?
That is to say: while you dig yourself out from your post-holiday inbox, here are the best issues of Complex Machinery from 2025. (Where "best" is based on the very scientific process of "I liked it.")
Also, Complex Machinery has a new spinoff. You'll see what I mean when you get to the In Other News section.
The best of 2025
#049 - "Looking for a grenade in a haystack": I'm filing this under "best" only because I expect ("fear?") it will become painfully relevant in 2026…
#047 - "What's left after it all falls apart": Is genAI a bubble? If so, and should it pop anytime soon, this will serve as your guide.
#043 - "Taming the delightful chaos": What can the computerization of Wall Street teach us about bringing AI into the workplace? I'd written about that topic here and there over the years, but this piece gave me the opportunity to put it all in one place. (This issue also ran as an article on O'Reilly Radar.)
#037 - "Cranking out trash": Consider this my periodic reminder to Check The Damned Robot's Output Before You Make It Public.
#027 - "A difference of time": Companies selling genAI are focused on the possible-future capabilities of the technology. Too bad they keep talking about it like it's already a present-day reality.
#033 - "GenAI is an executive's best friend": Those chatbots have a lot in common with their biggest fans …
Honorable mentions from 2024
I'm dusting off these newsletter issues because they've demonstrated strong staying power.
#018 - "Robot jobs": What jobs are suitable for AI-based automation? The humble self-service kiosk will tell us a lot. (Scroll to the bottom for the graphic that I've used several times since then …)
#017 - "Stacking the deck": What happened when robots started playing poker? And what does that tell us about AI performing work?
#010 - "The Looming AI Debt Wall": I wrote this about the metaphorical debt around genAI. Since then, the major players have taken on real debt, which prompted me to write about it again in 2025…
#007 - "Asking the important questions": For those who keep asking whether genAI is in a bubble?
#002 - "It's still a wild animal": A reminder of The Random™ , the wild animal that lives inside every genAI system.
#001 - "Nothing's real but the fakery": A little something on deepfakes and robot-wrangling.
Recommended reading
In 2025 I also released two ebooks:
Black Box Tactics: Use algorithmic trading techniques to improve your AI is a tactical guide for CTOs and AI practitioners. I apply lessons from the world of algorithmic ("algo," "electronic") trading to AI-related R&D and operations.
Black Box Tactics is polished-up version of an old blog post series I ran a few years back. I was pleasantly surprised to see that most of what I wrote about ML/AI back in 2020 still applies in 2025's world of genAI.
Twin Wolves: Balancing risk and reward to make the most of AI is a higher-level, leadership-focused guide that sees a company's AI transformation through the lens of risk-taking and risk management.
While this is all new material, I've been talking about the general ideas for some years now. I was happy to be able to pull all of those thoughts into one place.
What's on deck for 2026
They say prediction is a fool's game. But as some of these are already underway, I feel confident enough to at least be vague:
I have a couple of upcoming Radar pieces in the works. Then there's a writing project that's been on my mind for a while. (I've already gathered a lot of materials on it. So now the trick is to, y'know, actually write it all down.) There are also a couple of other matters, but it’s too soon to offer even a hint about those...
In other news …
This section usually hosts a list of one-liners from recent news. A few weeks back I spun this out into its own newsletter, also called In Other News, and gave it a wider remit than Complex Machinery's "risk, AI, and related topics."
In Other News will also be more timely: it lands every Wednesday. None of this noncommittal, "a couple times a month" nonsense you get here with Complex Machinery.
I'll keep doing In Other News here, at least for a while. But if you subscribe to the newsletter version, some of the links here might seem familiar.