haruki zaemon

I am reminded day after day how important it is to remain ruthlessly curious. Smart, capable people will either solve problems for themselves, realise they don’t know how, or realise they do know how but the environment is causing them to not do anything about it. This could be because they lack time, or they’re focused on other things, or any number of reasons. The trap is to think that I have the answers.

Here’s an example where I asked what I thought was a reasonable question but was in hindsight loaded with bias:

Q: Why aren’t we doing {A, B, C}?

A: Because it’s not important.

Q: It’s important to me. We should be doing it and we’re not. Why?

A: This has never been a problem. Why would it be a problem now?

That conversation went around and around for a while getting nowhere. So, I abandoned the approach and reflected on what was going wrong. Why couldn’t they see what I thought I could see? Why weren’t they worried about this the way I was? Were they naive?

At some point I realised I’d become stuck in a cycle of trying to be preacher, politician, and prosecutor. What I needed to do was flip things around, stop assuming the premise, and start acting like a scientist:

Q: What things would, could, or should we be doing but aren’t because we lack the necessary resources and time, or because we have skills and knowledge gaps?

A: {A, B, C, …, X, Y, Z}.

There it was. Smart, capable people already knew the answer. They weren’t naive. They knew what needed to be done, and they knew they should be doing it.

So what was stopping them? Why had they even gone so far as to say it wasn’t important to them?

I’ve observed that people will resolve dissonance in their favour, rather than have to sit with it. I’d mostly thought of it as choosing the “easier” option. What I had never really thought about (until this podcast episode) was that dissonance is a mismatch between belief and reality: You can choose to update your beliefs; or you challenge the reality.

Which in hindsight explains why, when I asked another question out of genuine curiosity, it became apparent:

Q: And, assuming you had the necessary time, resources, skills, and knowledge, how important do you think those things are?

A: Yeah, important. We’ve been wanting to do some of them for a while now but, didn’t feel we could …

They faced challenges they didn’t feel able to overcome, and that meant they experienced discomfort, i.e. dissonance. In order to resolve the dissonance, they had updated their beliefs: those things had never been a problem, and are therefore not important. The resolution of that dissonance looked like resistance to change.

Proselytising, making stump speeches, and reading facts into evidence rarely changes peoples minds. If anything, it’s likely to make them defensive, search for even more ways to rationalise their behaviour, and ultimately reinforce their position.

Instead, maintaining a mindset of curiosity and setting aside your own biases can reveal underlying conflict and help people motivate themselves into action.

I read Do Androids Dream of Electric Sheep? when I was a kid (c. 1982) and loved it. Not long after, I watched Blade Runner in the theatre (as was the fashion at the time) and loved that too. More recently we watched Blade Runner 2049 as a family which I also enjoyed, but certainly not as much as my memory of the original.

Re-reading the book for the first time in over 40 years, I realised my memory of the book was actually a memory of the original movie. It was clearly written in a different period and to a different set of expectations and norms. The stereotypes, the gendered character tropes, etc. all felt a bit stale. It also reminded me a bit of Man in the High Castle in that it felt like it went nowhere in the end.

All that said, I enjoyed it, and think I’d happily read it again, no doubt in another 40 years when they release Blade Runner 2077.

I’d been searching for a new sci-fi book to read and, having really enjoyed The Wool series, I thought I’d give Machine Learning by Hugh Howey a go.

Sadly, I did not enjoy the book overall. On a positive note, I did appreciate the author notes at the end of each story.

I’m coming to appreciate that I am not a fan of first-person narratives, nor short stories. Definitely not first-person short stories.

Maintaining the iambic pentameter of the original, as well as the number of lines, this English translation totally worked for me.

What more can I say? I’m certainly not about to review the story itself!

About as compelling as a horoscope. I never made it past chapter 2.

I Tried Simulating The Entire Ocean. I’m pretty much never going to need to do this (and let’s face it, the math is beyond me anyway) but I love watching the way people solve these kinds of scaling problems. I think it’s also super important to show new developers (like my kids) how the creative process is iterative, with many “failures” along the way.

A virtuous and systematic approach to sustainable organizational change. If authority relies too heavily on factors related to identity, the result will be something like an extremist cult. On the other hand, if the basis of authority focuses entirely on methodological rigor, we end up with a technocratic system that alienates the people it is meant to represent.

The “Hero’s Journey” Is Nonsense. My kids are taught Campbell’s “hero’s journey” in school. Turns out, it’s vague, ethnocentric, sexist, hetero-normative, cis-normative, and ignores the ways stories are shaped by the cultures and time periods in which they are produced.

Voyager 1 survives clogged thruster issue billions of miles away. Both spacecraft are unimaginably far away, and require ongoing manoeuvring so they can explore interstellar space. Something no one expected when they first launched nearly 50 years ago.

Why “AI” projects fail. Companies are spending a lot of time and money integrating AI for the sake of getting onboard the AI train, and not a lot of time understanding and solving real problems.

Technical Leadership. Instead of resenting non-technical people for occupying leadership positions, technical experts can (and should) develop business acumen, and interpersonal and communication skills, in order to scale their influence.

We Spent $20 To Achieve RCE And Accidentally Became The Admins Of .MOBI. The protocols underpinning the internet are old, and venerable, and riddled with holes.

I watched the Marlon Brando and Val Kilmer movie in 1996 and remember almost nothing of it. I’m not sure if that helped or hindered my enjoyment of the book.

I enjoyed the book for what it is. “A good example of that sort of thing” as they say.

It’s a relatively quick read with what I imagine were shocking science fiction themes in 1896.

If late 19th century colonialism is your thing, I would recommend Heart of Darkness instead.

Who knew they had another book series? Apparently, everyone but me.

The Mercy of Gods is the first book of The Captive’s War trilogy and brings with it a whole new world building journey. I’ll admit it took me a few chapters to get over this realisation but by the end, I was absolutely hooked, and a little sad that I’ll now have to wait for the next installment.

I can only imagine the expectation that any new series live up to The Expanse, and they nailed it. If you’re looking for a new sci-fi series to get into, and you love Daniel and Ty’s writing, get into this.

Transformed is the book I wish Marty et al. had written first. At the very least, the book I wish I’d read first. It’s definitely the book I think most executives and leaders should read first if they are wanting to switch to the product operating model.

Marty explains what it really means to make the shift away from the IT operating model, the challenges, the mistakes, the benefits. I predict that having read the book, many who thought they wanted the product operating model will realise they actually don’t, or will never get there.

For many organisations, the product operating model with outcome-focused, problem-solving, continuous-delivery, cross-functional teams is where they began as “plucky startup”. The IT operating model is where they are moving to in the name of “maturing” and “scaling” as a business, emphasising project-based, output-focused, requirements-driven, siloed teams, with formal change management.

If (like me) you’re already singing from the Marty Cagan song sheet, Transformed is probably more bias-confirming than eye-opening. If you’re product operating model curious, in a position of authority, and have a mandate to actually make the change, this is probably the book for you.

Managing work in progress (WIP) is not just about optimising operational efficiency; it’s also about fostering a healthy and empowering work environment:

• When there is too much WIP, people feel overwhelmed. This constant state of feeling overwhelmed leads to a sense of lack of control over one’s work. This lack of control combined with constant task switching and changing priorities leads to disengagement and helplessness.
• The more tasks a team is responsible for, the less time they tend to dedicate to collaboration. Divide-and-conquer is seen as the right approach, while working with others becomes synonymous with inefficiency. This reduction in teamwork leads to a sense of isolation.
• Increased cycle times and reduced efficiency means it takes longer for people to see the results of their work. The harder it is for employees to understand their impact, the more likely they are to feel ineffective.
• With too many tasks in progress, feedback loops become longer and slower. This delay in receiving feedback leaves employees feeling uncertain about their performance, and unable to know how best to make improvements.
• Executives over-emphasis on utilisation and keeping people busy leads to a lack of focus on bigger picture and strategic outcomes. This focus on day-to-day operations leaves little time for alignment and planning, communication of vision and direction, tools and training, or resourcing.

To counter this we need to: Constantly communicate vision and principles so those closest to the work can make better, more informed, more aligned decisions; reassure people that long-term, strategic outcomes are just as important as “quick wins”; make explicit all the things people believe they’re expected to do, discuss what is a true priority vs nice-to-have, and proactively plan for tasks we might need to “drop on the floor”; reward people for reducing WIP rather than taking on more; and keep asking people what they need to be and feel more empowered.

Too much work in progress is one of the more insidious and perhaps inadvertent means by which we disempower people. If we genuinely want to foster an environment where employees feel in control, engaged, and connected, it’s critical we ruthlessly prioritise our own workloads, and support others to ruthlessly prioritise theirs.

Departure Mono. A monospaced pixel font created by Helena Zhang, and inspired by early command-line and graphical user interfaces.

Newly Discovered Antibody Protects Against All COVID-19 Variants. A study led by The University of Texas at Austin has identified SC27, an antibody capable of neutralizing all known variants of the COVID-19 virus, opening prospects for universal vaccine development and improved treatments.

DOJ, Nvidia, and why we restrict monopolies. We shouldn’t allow past success to enable unfettered future profits when it stifles further innovation and competition, and definitely not at the expense of society or humanity as a whole.

Facilitated Workshops Create the Problems They Try to Solve. In my experience, facilitated workshops can bring about change and make a meamningful difference to the way groups of people work. At the same time, I agree they are a form of social engineering, and can be abused to reinforce existing structures and forms of control (via Chris Corrigan).

OpenAI Pleads That It Can’t Make Money Without Using Copyrighted Materials for Free. Another from the annals of “they nearly got there…”

Why Login Security Sucks. Login security is a mess. Username/password, TOTP, HOTP, and passkeys all have their issues and aren’t universally supported. Implementing solutions that balance security and usability can become complex and costly even for simple use cases.

Is my blue your blue? Test your colour perception. My blue/green boundary is at hue 174, the population median at the time I submitted my results, apparently.

What’s the Point of Bay Leaves? I mentioned to Benji that I’d voted “does nothing” in a Mastodon poll about the use of bay leaves in cooking. I was wrong.

Are you trying to watch Rings of Power season 2 and it is way too obscure? Let Glenn Fleishman give you a pathway. No spoilers. FTR, I enjoyed Season 1, and so far Season 2 as well.

Capt. Grace Hopper on Future Possibilities: Data, Hardware, Software, and People. 1982! I was 10 and Grace Hopper was more insightful about information technology in general than most people today. Smart, funny, and can talk coherently for hours without notes. Part one (of two).

The secret inside One Million Checkboxes. Someone made a website game with a million checkboxes open to the public to check on and off, prompting others to secretly encode easter eggs from simple messages to images, GIFs, and more.

Bypassing airport security via SQL injection. Look, we’re all going to end up on the wrong end of some security or privacy mistake at some point in our careers–sometimes more than once–but this is seriously amateur hour stuff (via Jon Eaves).

Misogyny Makes You Stupid. Misogyny leads people down a path of increasingly irrational and extreme beliefs, making them detached from reality. We all need to question our biases and the structural forces that produce them to avoid the trap of misogyny-fueled stupidity. As Brooke says: “You don’t want to be stupid, right? Question that shit.”

Students Find New Evidence of the Impossibility of Complete Disorder. This stuff blows my mind. I can’t even begin to think in a way that would conceive of asking the right question, let alone discovering the answer.

Silicon Valley’s Very Online Ideologues are in Model Collapse. I absolutely love the idea that humans are susceptible to a kind of (mental) model collapse, analogous to the way large language models are (via Baldur Bjarnason).

No one’s ready for this. From here on in, all photos are fake unless proven otherwise.

On body count. People in medieval times had and enjoyed sex, and discussed and found the topic as challenging as we do today.

Exceptions & Rules. I thought this nicely encapsulated some of the tension I’ve experienced and observed in various organisations, between those who strive for reliability (consistent and replicable outcomes) and those who seek validity (exploiting nuance and complexity).

Quantitative Criteria for Defining Planets. Sorry all you true believers, even the latest proposed definition means Pluto misses out on being a planet (via Deborah Pickett).

The Truth Behind Rising House Prices. The shift from US houses being treated like a commodity (1890-1972) to an asset-like investment (1972-) was not driven by supply and demand factors, but rather the energy-intensive nature of housing construction, and the growth of household debt (via Steve Randy Waldman).

Have CEOs Changed? The average interviewed CEO candidate has lower overall ability, is more execution-oriented/less interpersonal, less charismatic, and less creative/strategic compared to pre-GFC.

The 4.2 billion-year-old ancestor of all life on Earth today. I find it equal parts mind-blowingly cool and unsettling to know I’m related to my cats and the trees in our yard.

Why Gov. Tim Walz Drives Them Crazy. The Democratic VP candidate shows that it’s not just possible but natural and normal to be a middle-aged white guy and not a complete wierdo obsessed with controlling other peoples lives and bodies.

The quintessential Cantonese condiment - Fried Dace with Black Bean. I absolutely love this stuff with fried rice (seriously, I want some now just writing about it), and I found the history of it fascinating (via Adrianna Tan).

From Bike Rides to Boardrooms: The Evolution of Managing Up. It’s easy to be cynical but managing up is not sucking up, it’s about building mutually beneficial relationships, and shifting those relationships from transactional, subordinate, and tactical to long-term sustainable, collaborative, and strategic partnerships.

Magic moment: Sydney aquarium filled with song after sea birds mourn death of gay penguin Sphen. I’m not crying, you’re crying.

Fighting Through Mental Struggles: Incredible Renderings Created with Only a Pencil. 10/10 no notes.

A former security architect demonstrates 15 different ways to break Copilot. Surprise! Microsoft 365 Copilot can be made to reveal sensitive information.

Namibia: Live stream in the Namib Desert. I’m yet to see any animals (that could be the times I look) however I saw a Wart Hog and Oryx drinking. Others report seeing Zebra, Wrt Hogs, Oryx, Spotted Hyena, and more (via Emma Evans).

Digital Fibre Optic Gyroscope (DFOG). I don’t claim to really understand beyond it being a form of Inertial Navigation System (INS), using coiled optical fibre that you know, senses the Earth’s rotation.

Chris Patterson on Mastodon:

The advent of LLMs masquerading as artificial intelligence has made the notion of an absurdly powerful computer, constructed at great expense, and given unseemly resources to answer a meaningful question, only to return the answer “42”, feel more and more prophetic.

How cascades of rapid change routinely sweep across families, institutions, and nations. Transformational change comes about when small, loosely connected groups, are driven by a shared purpose. As leaders, we should connect those groups, empower them to succeed, and provide them with that sense of purpose.

The AI Supply Chain Tug of War. Suppliers up the AI chain are offloading their demand risk by taking profits now while Big Tech continue to pour in capital. This is a fragile equilibrium, and Big Tech could find itself in trouble if the music stops playing.

AI Unicorns Are Running Amok. Om with a similar take, comparing GPU investment to investments made in Sun Microsystems and Cisco at the start of the dot-com boom and eventual bust.

Consenting to decisions. There is a difference (albeit a fine line) between wanting people to accept something you already thought of, and wanting to generate and buy-in to something more. That’s one of the things I love about consent-based decision-making, and structured participatory practices in general.

How the First Sports Bra Got Its Stabilizing Start. In the 1970s, sports-bras weren’t a thing (because of course they weren’t). Then three women turned a jock-strap upside down and the first prototype was born.

What Are JPL’s Lucky Peanuts? I thought this was a joke when my kids first told me but nope, since July 1964, NASA launches have gone hand-in-hand with eating good-luck peanuts.

Someone asked a question about job levels. I’m paraphrasing here for the sake of brevity and anonymity:

As an early-stage startup, we made a deliberate decision to hire an experienced engineering team. As such, all our engineers have the same title. We’re now exploring the introduction of different levels and would love to hear your thoughts.

I certainly don’t claim to have all the answers however, I have been a part of implementing various approaches (flat, multi-track, sub-levels, etc.) with mixed results.

I’m currently of the view that unless you are a mega-corp (and perhaps even then) or Government:

• External title != internal role. Externally, people can call themselves whatever they want, within reason.
• Don’t use roles/titles and promotion to reward and recognise people. Have a smaller set of roles with large overlapping pay bands and use expectations to provide additional compensation.
• Accept that some people will want to focus on deep coding and that’s ok. Some folks will want to extend into other duties, and that’s ok too.
• Use roles as a way to provide clarity and focus on what’s expected. Expectations are the things people can meaningfully compare, hold themselves accountable to, and be assessed on.
• Look for ways to provide people with opportunities to extend themselves in their current role. If in doing so, they periodically fulfill the expectations of other roles, recognise and reward them for that. Don’t demand they do it all the time, and don’t delay that recognition and reward until they’re “officially” in another role.
• Even within a role, there are likely fundamental things we expect of everyone, and then specialisation that allows people to contribute in different but nevertheless meaningful ways.
• Avoid creating roles based on “jobs to be done.” The classic example is tech or delivery lead. Depending on the nature and scope of the work, it could be performed by a junior engineer, or a principal.
• People management really is a different skill set. No really. You may or may not need specific roles but whatever you go with, give people explicit and ongoing support and training.
• Minimise the number of vacancy-based roles. Make them accessible to internal staff (as well as external candidates) and invest in opportunities and training to maximise their chance of success.
• Don’t create “unicorn” roles by back-solving for people you want to retain. They lack clarity of expectations, lead to the perception of (if not actual) favouritism, and easily become misaligned with real business needs.
• In a small startup, roles are far less clear. If you’re vetting people sufficiently in the hiring, you’re likely to bias towards people who will pretty much do anything, so the above matters far less.

Companies want to attract and retain people, and to set pay and benefits. Individuals want to feel a sense of progression and achievement, and to signal this to others.

A mistake I’ve made repeatedly in the past is trying to capture everyone’s needs in some form of job level structure where the only way is up, and in which people are inadvertently incentivised to compete and outshine one another.

These days, I advocate for a simple structure with multiple overlapping levers. One in which there is more clarity in what’s expected, more opportunity to be recognised and rewarded, and more flexibility in how organisations can support people to grow and perform at their best.

Edgar Schein’s Humble Enquiry. Perhaps unsurprisingly, I love this so much. I have come to really appreciate how important relationship building is, and how much more effective we can be when we prioritise genuine curiosity and humility over efficient knowledge transfer and decision-making.

Crash Course Pods: The Universe. I can’t get enough of this podcast series. The most awesome Dr. Katie Mack (aka AstroKatie on Mastodon) chats with author John Green about the history and mysteries of the universe, and explores some of John’s existential crises along the way.

We unleashed Facebook and Instagram’s algorithms on blank accounts. They served up sexism and misogyny. The algorithms of Facebook and Instagram serve up sexist and misogynistic content even without any user interaction or input. The platforms’ algorithms make assumptions about young men’s interests and serve them content related to “Manosphere” influencers and misogynistic memes. The findings align with similar research on YouTube and TikTok.

When ChatGPT summarises, it actually does nothing of the kind. ChatGPT produces something more akin to an extreme abridgment (my words, not the author’s) than an actual summary. In doing so, it can omit important context and facts, and make stuff up. This matches my experience, and why I only use automated summaries to get a rough idea of the content before diving in deeper.

How Does OpenAI Survive? OpenAI’s costs are estimated to be ~US$5 billion annually, and revenue is not keeping pace. Given the current lack of clear mass-market utility for the technology, OpenAI’s business model seems unsustainable over the long-term.

Deep frying coffee beans: yay or nay? The result, according to James Hoffmann, was “surprisingly good,” and the flavour “really interesting.” I would happily pay a hipster Melbourne barista to make this for me, once (via Benji and Warren).

How Physicists FINALLY Solved the Feynman Sprinkler Problem. Dr. Ben Miles explains the solution to a problem I never knew I needed to know. Associated with physicist Richard Feynman, it actually dates back to Ernst Mach’s textbook The Science of Mechanics, first published in 1883.

Marsh Family parody adaptation of “Dancing Queen” by ABBA, on JD Vance. 10/10 use of “Tangerine.”

List of eponymous laws. I’m up to “Newton’s flaming laser sword, also known as Alder’s razor: What cannot be settled by experiment is not worth debating” (via David Lee).

Iceberger. Draw an iceberg and see how it will float.

I love books that give me a coherent model and vocabulary for things I’ve observed but not necessarily put together myself.

We often describe people as having a growth mindset and the book covers a lot of ground on how to foster that. As the title suggests, it also demonstrates that mindsets are cultural, and that the collective mindset is a predictor of the individual mindset.

Growth mindset cultures inspire learning, foster collaboration and inclusion, spur innovation, and build trust. Leaders in growth mindset cultures believe that great ideas come from everyone. People in growth mindset cultures tend to celebrate the success of others.

So, what’s the opposite of growth mindset culture? It’s genius mindset culture.

Genius mindset cultures undermine learning, stifle innovation, and create mistrust and inter-personal competition. They’re based on the premise that, as the name suggests, some people possess inherent genius and should be rewarded as such.

Leaders in genius mindset cultures favour people who think the same way as them, look fondly upon those who unquestioningly execute their plans, and reward people for their overt individual contribution.

Genius mindset cultures have in-groups of people deemed brilliant or talented, and out-groups seen as incompetent. In-groups get all sorts of leeway to make mistakes, whereas failure in out-groups is further evidence they cannot be trusted.

People in genius mindset cultures are incentivised to move from the out-group to the in-group, and remain there at all cost. They are more likely to cut corners, own other’s ideas, and resent others for their accomplishments.

The book is based on rigorous scientific research and provides many resources to promote growth mindset cultures. For me, the standout concepts are that mindsets are cultural, and that thing I’ve observed, genius mindset, now has a name.

4 lessons for high-quality software from a surprising place. Compliance can seem like a necessary evil for many developers, especially those coming from startups. Having worked in a number of regulated industries, my experience (and that of the author) is that tackling the reality of compliance head on actually “establishes a foundation for success for high-quality code and minimized risk that save time and money in the long run.”

Embracing clarity: Shifting away from a culture of certainty. A culture of demanding certainty leads to presenting assumptions and partial knowledge as facts, stifling innovation and deeper problem-solving. Focusing on understanding root causes and context, rather than rushing to solutions, leads to more impactful and sustainable outcomes.

The Iceberg Model: towards unraveling our patriarchal legacy. The software development industry has a gender diversity problem rooted in historical patterns and societal norms. Software development should be viewed more as a collaborative, interdependent sociotechnical system, rather than a purely technical engineering discipline. A culture of high standards, kindness, and openness to diverse perspectives can foster a creative and productive work environment.

How Slavery Inspired Modern Business Management. Growth and profit-seeking have been built upon violence, injustice, and the exploitation of human lives. The failure to fully reckon with the connections between slavery and modern business practices amounts to denial (via Warren).

LGBT and Marginalized Voices Are Not Welcome on Threads. As if I needed more validation that deleting all my Meta accounts (Facebook, Instagram, Threads) was the right thing to do.

The Data Is In: Return-to-Office Mandates Aren’t Worth the Talent Risks. Speaking of diversity (or lack thereof) return-to-office mandates particularly impact women and minorities. When Gartner says it, maybe companies will listen?

Opinion | Donald Trump Is Unfit to Lead. Did the NYT finally grow a spine?

I had Expanse withdrawal, and Memory’s Legion proved an effective remedy.

I really enjoyed this collection of short stories, each intended to fill in some part of the Expanse universe that would have been overkill for the main storyline. As a nice bonus, the audiobook also includes author’s notes narrated by Daniel Abraham and Ty Franck themselves.

I liked that each story goes deep on one aspect of one character (often secondary characters at that) and left me with no doubt as to the depth of world building the authors have gone to.

I did find the one first-person story less enjoyable than the rest. I don’t think it was objectively bad, but the perspective and the cadence were so different to everything else in the series that I found it a little jarring.

If they wrote more, I’d be in. At the same time, Memory’s Legion gave me some closure, as it were. Well worth the read.