Efficiency

Intellectually I knew that the question I should be asking was: if I invest X hours into migrating to a new editor, will I make those X hours back in increased productivity, and over what time period Y? While a precise answer to this isn’t possible, we can often make decent guesses.

...I also think that there is a place for sharpening one’s tools just for its own sake. Especially as we get older, and our responsibilities increase, anything related to our work can start to feel like a burden. Sometimes even a small amount of time invested in sharpening our tools can bring a bit of joy and restore motivation.

Whatever the outcome we’re looking for from tool sharpening, it seems to me that the most important factor is to think explicitly about why we’re doing it, and what we hope to achieve in so doing.

When Jiro Ono (of Jiro Dreams of Sushi) weighs down the lid of his rice cooker like it’s going to blow away, or turns the fire up or down, he’s not doing it because he has an explicit understanding of the physical and chemical processes by which the rice takes on firmness, fullness, and luster, but because he trusts in his attention to detail. There’s a level on which cooking rice is a matter of ritual. Cooking as a field of knowledge doesn’t transcend human understanding, but cooking as an activity almost always transcends the understanding of the cook, and this is important.

One of the most common mistakes I see people make when looking at data is incorrectly using an overly simplified model. A specific variant of this that has derailed the majority of work roadmaps I've looked at is treating people as interchangeable, as if it doesn't matter who is doing what, as if individuals don't matter.

Individuals matter.

In 2024, I must admit I am having a bit of a personal crisis regarding continuous improvement. It doesn't matter whether we're talking about "high performing" companies and teams, or companies seeking to transform how they work—there is very little talk of continuous improvement.

The language around layoffs in tech is rife with talk of operational efficiency gains, achieving profitability targets, rightsizing, and high-interest rates. There is very little sense that employees were allowed to be part of the solution or path forward (e.g., by recommending areas of cost savings, quality circles, taking pay cuts, etc., similar to Harley-Davidson in the early 1980s, Southwest Airlines in 2001 or 2008, or countless companies mid-pandemic).

...Meanwhile, among companies less impacted by the ups and downs (real or imagined) in tech and deep in their respective transformations, we see the same old copy-paste approaches, outdated practices, and "changing but not really changing." On some fundamental level, when it comes to continuous improvement, there is no real difference between the two: in both cases, you have change imposed on teams and a lack of engagement to co-design the way forward. In both cases, existing power structures remain entrenched.

Our current definition of “productivity” is broken. It pushes us to treat busyness as a proxy for useful effort, leading to impossibly lengthy task lists and ceaseless meetings. We’re overwhelmed by all we have to do and on the edge of burnout, left to decide between giving into soul-sapping hustle culture or rejecting ambition altogether. But are these really our only choices?

On preparation, he urged his hearers to study and prepare themselves, relating the instance of the lumberjack who said that if his life depended upon his ability to cut down a tree in five minutes he would spend three minutes sharpening his axe.

People count by tens and machines count by twos—that pretty much sums up the way we do arithmetic on this planet. But there are countless other ways to count. Here I want to offer three cheers for base 3, the ternary system. The numerals in this sequence—beginning 0, 1, 2, 10, 11, 12, 20, 21, 22, 100, 101—are not as widely known or widely used as their decimal and binary cousins, but they have charms all their own. They are the Goldilocks choice among numbering systems: When base 2 is too small and base 10 is too big, base 3 is just right.

...The cultural preference for base 10 and the engineering advantages of base 2 have nothing to do with any intrinsic properties of the decimal and binary numbering systems. Base 3, on the other hand, does have a genuine mathematical distinction in its favor. By one plausible measure, it is the most efficient of all integer bases; it offers the most economical way of representing numbers.

The Mythical Man-Month: Essays on Software Engineering is a book on software engineering and project management by Fred Brooks first published in 1975, with subsequent editions in 1982 and 1995. Its central theme is that adding manpower to a software project that is behind schedule delays it even longer. This idea is known as Brooks's law, and is presented along with the second-system effect and advocacy of prototyping.

I am attempting to get my old MA field, the Digital Humanities (crudely speaking, a CS backdoor for humanities undergrads), to teach a management module that dispenses with lots of conventional management practices in favour of teaching something more akin to Munger/Buffett’s Unrecognised Simplicities of effective action. I’m positioning this as Hyper-Performance Management. I’ll refer to it henceforth in this article as HPM.

I want to use this to create a generation of evangelists for hyper-performance management standards who go into industry and revolutionise how they work. Management capabilities on the level of PARC/ARPA, the Apollo project, the best COVID vaccine rollouts; and management cultures like those at early 21st century Apple, Stripe, and in state structures like mid-century Singapore and Napoleonic Britain, are exceptions in the history of management.

I want to make them the norm by teaching them at the educational root, before students get into the workforce and are intellectually poisoned and/or hampered in their career by subpar management norms.

I want to prepare the alternative ahead of time.

And the rest is history. The “Tech Debt Friday” was born. Why Friday? I do not remember, but it had something to do with the fact that some people were off on Friday so in practice, tech debt would not “steal” 10% sharp. Still a victory!

Sometimes, when you don’t work on a feature idea right away, it loses steam and fades away. Yet, the idea of multi-edit remained surprisingly popular at Figma.

...Multi-edit is the feature which has had the longest gap between initial idea and launch. Over those years, we kept picking it up and putting it down, and we kept polishing and iterating on it. It feels like a rock that we have been tumbling in a rock tumbler until it is shiny and smooth.

One of the odd things about this process is that as we refined multi-edit, it almost started fading away. We’ve been living with multi-edit internally for months now, and we’ve kind of stopped noticing it—until we switch back to the production environment, and we realize how inefficient things were before!

“In the past the man has been first. In the future the system must be first.” — Fred W. Taylor

Taylorism was a way of thinking that came at the expense of the workers’ own knowledge of their system. Taylor summed up his philosophy thus:

“It is only through enforced standardization of methods, enforced adoption of the best implements and working conditions, and enforced cooperation that this faster work can be assured. And the duty of enforcing the adoption of standard and enforcing this cooperation rests with the management alone.”

The unscripted practices of the old offices would remain, but as a kind of subterfuge: in the future, a leisurely pace wouldn’t be the norm; time would not be given, but stolen.

“CEO-ification” refers to the trend of nonprofits and charities to increasingly mirror corporate and military structures. Often they will adopt similar language, hierarchies, and strategic approaches.

...In truth this stemmed from Taylorism, also known as Scientific Management. This theory was developed by Frederick Winslow Taylor in the late 19th century, but influenced today’s corporate language by introducing efficiency-focused terms like “time and motion studies” and “optimised workflows.”

...I spoke to Bryony Shannon about this on the Let’s Talk Ideas podcast. Bryony argues that the words we choose reflect our values and feelings, and shape how we think and act. She argues that when language focuses on processes, bureaucracy, and transactions, it can reduce people from individuals to labels – like “service user” or “case.” These words distance us from the very people we were employed to develop relationships with.

Those relationships then become mere transactions. As Rob Mitchell has said, when you’ve got a form and a process for every relationship “Love becomes relationships. Relationships become processes. Processes get processed.”

Frederick Winslow Taylor would have loved today’s world of process, customer segmentation and journey mapping. Such methodologies can approach humans lives as something that can be managed just like a car production line, or a canning factory producing baked beans.

It’s done, I’m just waiting on a review
It’s complete, I just need to do some final testing
It’s ready, I just need to merge and deploy

End users do not care. Nor should they. Nor can they. To the end user, this work provides no value. It doesn’t exist.

Work in progress has zero value. Ship!

• Before a task is shipped it provides zero value.
• Any work in progress is pure cost.
• Two tasks in progress adds cost, for no value.
• Only after shipping do you create value. Always ship.
• One task shipped is infinitely better than 4 tasks “almost done”.
• Ship something of value first. Then begin something new.

Many of us have come to expect work to involve no slack time because of the negative way we perceive it. In a world of manic efficiency, slack often comes across as laziness or a lack of initiative. Without slack time, however, we know we won’t be able to get through new tasks straight away, and if someone insists we should, we have to drop whatever we were previously doing. One way or another, something gets delayed. The increase in busyness may well be futile.

Good software comes from a vision, combined with conversations not commandments. In a craft-focused environment, care for efficiency, simplicity, and details really do matter. I didn’t leave my last job just because I wanted to make something new. I left because I wanted to make it in a way I could be proud of.

+10x engineers may be mythical, but -10x engineers exist.

To become a -10x engineer, simply waste 400 engineering hours per week.

The principle of “good enough”  is a rule in software and hardware. It indicates that consumers will use products that are good enough for their requirements, despite the availability of more advanced technology.

I've been thinking about this idea a lot lately. In fact, I've had the beginning of a blog post titled “Just Because You Can Doesn't Mean You Should” sitting in my ideas folder for several weeks now. Just because you can afford the fanciest, most powerful device doesn't mean you should buy it. I fall into this trap myself from time to time, and this is a good reminder for all of us not to get caught up in it. “Good enough” truly is good enough despite the availability of more powerful technology.

How can we make the case for slowing down to speed up when we are already going slow, and when shutting things down and starting over is not an option?

In other cases, the limitations on material reduction have more to do with building form. Most buildings, for instance, are rectangles that contain a series of rectangular rooms, because of the many benefits of this sort of arrangement. A rectangular room is easy to use (since you can push furniture flush against the walls), easy to build, packs tightly together against other rectangular rooms, and so on. Technically circle or sphere-shaped buildings would be more materially efficient (requiring less material to enclose a given amount of space), but while in some cases you see this (such as with pressurized storage tanks or the brief fad of octagonal houses in the 19th century), in general the material savings is not worth the added construction complexity, cost, and loss of convenience.

With architectural elements, we see the same general phenomenon, where further material reductions come with performance tradeoffs. In general, thick, heavy materials feel better than lighter, flimsy materials, and all else being equal people prefer the former to the latter. This is one of the reasons that drywall substitutes like vinyl on gypsum have had limited appeal, and why people still want tiled bathrooms even though fiberglass or acrylic are probably superior from a pure performance perspective.

At a former job, some colleagues would actually go stalk my GitHub to learn from my reviews! That was because those reviews were not just “LGTM” or “Please do __ instead.” They came with explanations of why something was better, or questions about someone’s goals. I often outlined tradeoffs in approach. I would leave links to other relevant materials for them to read.

If that sounds like it would take far too long, well… no, because I have practiced writing quickly and clearly for literally decades now. I could review a non-trivial change and give it non-trivial feedback in 20 – 30 minutes, and use that to level up the whole org over time. I say “the whole org” because when there was something actually substantive or important in one of those reviews, I would sometimes share the review for other people to read. (Only after getting permission from the person whose code I was reviewing, though; that’s important!)

Takeaway — engineers, do yourself and all of your teammates a huge favor and learn how to write quickly and cogently. That means practicing it! But the dividends are huge.

Last week I saw several articles about the passing of human-computer interface expert Larry Tesler. Most of the headlines mentioned his invention of cut-copy-paste, but if you read more deeply one of the things he really emphasized was eliminating modal interfaces—he even got a vanity license plate that said “NO MODES”.

I realized this morning that this is exactly my gripe with the Workday timesheet application: it’s extremely modal, where the one we used to use with our in-house developed HRMS was almost completely modeless.

...In summary, this interface is almost completely modal: each page only lets you do one thing, and you have to click through pages and pages to complete the full task.

This is why Larry Tesler fought against modal interfaces: they make it harder to get the job done. And he had this figured out back in the 1970s. Too bad big software companies like Workday still don’t understand this, or don’t care.

When we enter the world of refuse and waste, we cross over into a mirror-image economy. In the "normal" world, we pay to acquire things; on the other side of the looking glass, we pay to get rid of them. Junk isn't merely worthless; it has negative value.

A chemical engineer once told me about a recent improvement in a manufacturing process; by fine-tuning a chemical synthesis he had increased the yield of a certain commodity from 98 percent to 99 percent. I congratulated him, but I couldn't help remarking that this seemed like a rather paltry improvement. "Ah, you miss the important point," he said. "The amount of waste goes from 2 percent down to 1 percent. It's cut in half. We save tremendously on disposal costs."

The real problem is that programmers have spent far too much time worrying about efficiency in the wrong places and at the wrong times; premature optimization is the root of all evil (or at least most of it) in programming.

When you’re driving a car down a road, you might get a bit stuffy and decide to roll your windows down. The air will flow in, the wind will get louder, and the sensation of moving will intensify. Your engine will start working a bit harder–and louder–to maintain the same speed. Every sensation will tell you that you’re moving faster, but lowering the window has increased your car’s air resistance, and you’re actually going slower. Or at minimum you’re using more fuel to maintain the same speed.

There’s nothing that you didn’t already know in the first paragraph, but it remains the most common category of reasoning error that I see stressed executives make. If you’re not sure how to make progress, then emotionally it feels a lot better to substitute motion for lack of progress, but in practice you’re worse off.

The seconds you choose to use now, to lubricate your keyholes, will allow you to save seconds when you choose to do something more important than lubricating your keyholes — and that's to say nothing of the seconds you use just sitting around not lubricating anything at all.

Efficiency is produced not by the sort of movement monoculture of cars-only American cities but by a sensitively tailored combination of modes sited to exploit the particular efficiencies of each and providing useful duplication and alternative.

The reason dwelling densities can begin repressing diversity if they get too high is this: At some point, to accommodate so many dwellings on the land, standardization of the buildings must set in. This is fatal, because great diversity in age and types of buildings has a direct, explicit connection with diversity of population, diversity of enterprises and diversity of scenes.

Among all the various kinds of buildings (old or new) in a city, some kinds are always less efficient than others in adding dwellings to the land. A three-story building will get fewer dwellings onto a given number of square feet of land than a five-story building; a five-story building, fewer than a ten-story building. If you want to go up far enough, the number of dwellings that can go onto a given plot of land is stupendous—as Le Corbusier demonstrated with his schemes for a city of repetitive skyscrapers in a park.

But in this process of packing dwellings on given acreages of land, it does not do to get too efficient, and it never did. There must be leeway for variety among buildings. All those variations that are of less than maximum efficiency get crowded out. Maximum efficiency, or anything approaching it, means standardization.