What happened this month? I went to Toronto to visit friends: mission accomplished, it was lovely spending time together. Bonus mission: snowboarding! Blue mountain is small but adequate for just one day.

These three days in Toronto were the prefix of a longer stay in North America: I had ostensibly made time for a few more days in New York as a tourist, for the first time in a bit more than ten years.

The tourist side of my stay was entertaining: I went to places I would never go to otherwise, like the Edge platform on top of the Hudson Yards tower. This is a place I will probably never go back to, but the view was breathtaking (as was the wind—and the cold!).

Other highlights included the MET, the Natural History Museum, and dumplings at Veselka.

Although I went to NYC for the sights, I stayed for the friendships. It is not just nice spending time with loved ones: it is important too.

At the time of this writing, I am spending a few days at home, resting from my previous travel and preparing for the next one. The tendinitis in my shoulders from last summer has finally abated (yay) and I was at last helped by a physiotherapist to regain some missing mobility (double yay). Sadly, I have also started to develop new pains in my forearms from excessive reading, and got myself a stomach flu from staying too close to school-age children.

Optimistic that I will soon recover, next Monday I plan to start a journey that will see me travel anew and afar to visit friends, this time more on the Western side of the USA. Hopefully, some snowboarding will happen too, although little of it has been planned yet. I will be back shortly before the end of the month, with a plan to leave one week later for more snowboarding in the Alps.

❦❦❦

I am also very well aware that either my health or the world economy may impede my ability to do this in the future. I try not to think about that future too much, but I plan to carefully remain mindful and grateful for the uniqueness of this experience and opportunity.

❦❦❦

February also saw me get back to reading.

Essentialism: The Disciplined Pursuit of Less - Greg McKeown

Ostensibly, this book guides the reader towards good use of their time—more towards things that matter, not just more stuff in less time. It’s easy to recognize in the style that the author has a business background. However, I was also positively surprised to see that the book really offers a personal philosophy, taking time to share a set of principal values and then deriving how one can behave accordingly. Many of the precepts in there are relevant to the modern times, and I easily imagine how reading this could have lowered my overall anxiety a few years ago. I’d recommend reading it to any professional.

There were a few shortcuts—the author recommending a behavior, and only afterwards saying “this is how an essentialist would do it” without much justification—but given how the ideas overlap with other things I was reading last year (e.g. Ego is the Enemy - Ryan Holiday) I’d give him a pass.

Smart Brevity: The Power of Saying More with Less - Jim Vandehei, Mike Allen, Roy Schwartz

In this book, the co-founders of the news agency Axios share their “elements of style”: how they package information and news to drive their point through to busy or inattentive readers. They also explain a bit how they use these principles to organize communication inside the Axios organization, between employees.

I’ll give them an A for consistency: the book is also structured according to their principles. Meanwhile, I found it a bit hard to read! 🤡

Joke aside, I am convinced that their ideas are relevant in a narrow set of contexts, such as when a company executive needs to introduce an organizational change, or when aiming to reach to said executive, hold their attention, and deliver important input.

For general news, I’m not so sure. Their method sterilizes the “window into the soul of the writer” from the text and I think it makes it hard to use for e.g. teaching purposes, or to define policy.

That being said, reading this was thought provoking, and I plan to keep it in my recommended toolbox for newly hired senior staff.

12 Rules for Life: An Antidote to Chaos - Jordan B. Peterson

I started reading this because someone I like recommended it to me. I really wanted to like it: I believe that we need practical guidance books, for young people especially, to help them grow a moral compass without pontificating.

Sadly, this book is not that. It’s so terribly bad. Some of the ideas behind certain pages sounded interesting, but really this is a case of “broken clocks are right twice a day”. There were at least two moments where I literally rolled my eyes and gasped aloud at how bad it was. I now believe this author is a fraud (that they are not actually qualified to work as a psychologist) and this book is actively harmful. Don’t read it.

❦❦❦

Technical things I’ve read this month.

In Files are fraught with peril, Dan Luu explains how the “file” abstraction in operating systems is not providing the guarantees that most programmers expect, and how a lot of software contains latent bugs as a result. This is “an oldie but a goldie”—I had read this twice before and enjoyed reading it for the third time.

In Alibaba/T-HEAD’s Xuantie C910, Chester Lam provides a technical analysis of the new CPU chip “Xuantie C910”, which is the (current) highest performing implementation of RISC-V. I have been a believer in RISC-V ever since the early 2010s and am always glad to see serious development with it. It was also nice to read something so close to my area of academic research; it’s not often that I feel like the ideal audience for an article, and that feels good too.

That said, there is also a political undercurrent here: Alibaba and other Chinese firms making serious progress here, as a way to become independent from previous UK (ARM) and US (x86) hegemony in high-performance computing. In my opinion, we should keep an eye on this type of work: with the role of computing in world politics nowadays, I see this as much more likely to tip the geo-political balance than other global trade shenanigans.

Meanwhile, in The PC is Dead: It’s Time to Make Computing Personal Again, Benj Edwards reminds us that open computing is an essential instrument to a well-functioning democracy. I had also made a similar point in the first chapter of my PhD thesis, so I found myself in strong agreement with the author. Even though the article is illustrated by pictures and screenshots from pre-21th century technology, do not misunderstand it: the argument is actually geared towards our current technical landscape and what we should do about it in the coming 10 years.

Another very good thing that happened is the creation of the Calm Tech Certification, an award to be granted to innovations that minimize distractions and bloat. It is inspired from Amber Case’s book Calm Technology (2015) and both the book and the award are meant to guide product design towards technology that better integrates with our lives and doesn’t incur or amplify mental health issues. I can’t wait to see a “Calm Tech” label in stores.

❦❦❦

I’ve also read a few management things.

A person called Paul Hammant has compiled a lot of technical principles around software development in this web site: Trunk-based Development. Coincidentally, I was talking to a friend last month about how we’re missing some public resource about operational best practices to help people starting tech companies or when an early-stage organization hires their first technical staff. So it appears that such resources exist, this is one and I like it a bit.

In Tech Takes the Pareto Principle Too Far, Bobby Lockhart points out the recent trend in the tech industry to only work on the 80% of a product and release that, what amounts to essentially “unfinished” products. The reason why this happens is that the venture capital (VC) investors are OK with it, and VC money is currently at the forefront of innovation, so it drives culture too. We can oppose this trend to that in game development, where it is common to implement first a “vertical slice” instead: a version of the “finished” product that only shows a slice of the full thing. For example (in the context of video games), it could show the entire user experience and visual output for just one scene or level in the game.

The main point in the article is not that it is disappointing how customers do not expect finished products anymore (although that is also true IMHO); it is that they now believe/expect that the “remainder 20%” will come later, presumably through a product update. So product publishers can “cash in” with sales even in cases when the last 20% is properly impossible to build.

Is it fraud when the customers don’t really expect results?

Meanwhile, in Quiet Quitting: Why Employees Are Demanding Fairness And Boundaries, Diane Hamilton tries to explore the causes of a recent trend in workplaces, where employees do just what their job description requires and not much more. I think the article does a good job of explaining the symptoms, but not too good of explaining the causes. Younger workers won’t ever afford a home regardless of how much overtime they make, so why bother even trying? (True in the US and the Netherlands, but also probably in many other places.)

❦❦❦

The big news in the AI/LLM world last month was probably the unveiling of DeepSeek’s models. DeepSeek is an obscure research lab and their new thing is eating everyone else’s lunch. In The Short Case for Nvidia Stock, Jeffrey Emanuel explains the big picture around this, from the angle of what we can expect from NVIDIA’s stock price.

Meanwhile, in Let’s talk about AI and end-to-end encryption Matthew Green remarks that the “AI technology” currently resides in some company’s computers far from us, and so anytime we integrate it in tools for private communication (e.g. e-mail, chat messages), we are essentially violating all kinds of privacy expectations.

The most intriguing thought came from AI Slop, Suspicion, and Writing Back from Ben Congdon. Ostensibly, this piece rants against “AI slop”, which is when a human author is lazy and uses the output of an LLM unchanged. But the key argument is this: we are expecting more and more people to use LLMs to make decisions or to develop policy. Because of this, the author suggests a moral duty to write as much stuff as possible today, as humans, to influence the training of the models that will be used to decide policy tomorrow. Voting for politicians likely won’t work anymore; it is the training input that will have most influence on society.

❦❦❦

Coincidentally, some interesting science came up last month too: The Conversation published Knowing less about AI makes people more open to having it in their lives – new research which summarizes a piece of work by Chiara Longoni, Gil Appel and Stephanie Tully. In a nutshell, the science suggests that people are less ready to accept AI technology if they are more technically literate. What angered me when reading this is how little the authors seem to care about why this is happening; for example, no comment is made about how technically literate people know more about how the technology is being misused. Worse yet, the authors suggest that “companies may benefit from shifting their marketing efforts and product development towards consumers with lower AI literacy”—which made me want to flip my table and scream into the void.

Gary Marcus also put a piece out in the New Yorker, Will a Robot Take Your Job? where the answer to the question in the title is “maybe, we don’t know yet.” I don’t believe this piece is very good, but it was recommended to me and I usually like the New Yorker opinion pieces. There’s only one argument in there which I believe bears repeating over and over:

As Oscar Wilde put it, if all-powerful machines were “the property of all, every one would benefit by it.”

I’m not keen to directly force collective ownership of the means of production, but I wouldn’t mind a much higher tax on capital, like Pikkety suggested.

❦❦❦

My apologies for the rather dry style of this edition. Writing this was excruciatingly painful, due to the flu-induced headache and muscle ache. Next month should be better hopefully.

❦❦❦

References:

• Essentialism: The Disciplined Pursuit of Less - Greg McKeown
• Smart Brevity: The Power of Saying More with Less - Jim Vandehei, Mike Allen, Roy Schwartz
• 12 Rules for Life: An Antidote to Chaos - Jordan B. Peterson
• Files are fraught with peril - Dan Luu
• Alibaba/T-HEAD’s Xuantie C910 - Chester Lam
• The PC is Dead: It’s Time to Make Computing Personal Again - Benj Edwards
• Calm Tech Certification - IEEE Spectrum
• Calm Technology - Amber Case
• Trunk-based Development - Paul Hammant
• Tech Takes the Pareto Principle Too Far - Bobby Lockhart
• Quiet Quitting: Why Employees Are Demanding Fairness And Boundaries - Diane Hamilton
• The Short Case for Nvidia Stock - Jeffrey Emanuel
• Let’s talk about AI and end-to-end encryption - Matthew Green
• AI Slop, Suspicion, and Writing Back - Ben Congdon
• Knowing less about AI makes people more open to having it in their lives – new research - The Conversation
• EXPRESS: Lower Artificial Intelligence Literacy Predicts Greater AI Receptivity - Chiara Longoni, Gil Appel and Stephanie Tully, DOI 10.1177/00222429251314491.
• Will a Robot Take Your Job? - Gary Marcus
• Capital in the 21th century - Thomas Pikkety