Sleep More

100-hour weeks are the dumbest thing about otherwise-brilliant Silicon Valley culture.  They result in suboptimal quality, but worse, they waste human capital.

A recent true-life post, The Truth About What It’s Like Working For Uber, sounds like every job I had in the Valley — except the company I ran, where I could  encourage rest.

In the San Francisco Bay Area, the engineering culture for at least 40 years has promoted self-flagellating sleep deprivation.  For example, at a VC-funded software company I wrote code for in Menlo Park, our head of engineering held a contest to see which coder could sleep the least during the final weeks to our initial release.  The “winners” were basket cases for years afterward.  I took last place, sleeping an average of 5 hours a night — yet I am the only guy on that team to finish another programming project in the next 5 years.  The rest were wiped out, ruined in their mid-twenties.  This is not unusual.

Later, I ran a software company in Mountain View.  With the previous company’s death marches fresh in my mind, I gently asked everyone to leave the office by 6 or 7pm, to get some sleep.  With no death marches, we still hit deadlines, either by narrowing product scope or through clever workarounds.  We released groundbreaking mobile apps a decade before everyone else, but more importantly, we wasted no human capital:  our engineers generally went on to productive jobs elsewhere.

Programmer David Heinemeier Hansson concurs in his post, Sleep Deprivation is Not a Badge of Honor.  DHH created the Web development framework Ruby on Rails, co-founded cloud-based project management SaaS firm BaseCamp, and won the 2014 Le Mans auto race (how’s that for polymath?).  He does it on a reasonably consistent 9 hours of sleep.

Just sleep more.

Minimalist geopolitics

Global influence is won by a strong financial position. Weapons are merely a cart, pulled by horses that include tax base, borrowing capacity and GDP. These are the things that buy influence, whether through guns, elections or direct investment. (Conversely, empires collapse for lack of money.)

Nations seeking influence focus all their energy on raising their own people’s aggregate income, which increases private-sector financial strength, as well as the tax base.

This means investing mainly in education, infrastructure, public safety and rule of law, while avoiding resource drains like civil strife and, especially, external wars.

Countries that did this single-mindedly in recent decades — China, South Korea, Japan, Singapore — gained significant influence in a short time. Countries distracted by internal strife — Thailand, Malaysia — gained less. Countries that didn’t prioritize at all — the United States — lost relative influence.

Any country that is not primarily focused on increasing sustainable national income is off its game. It will lose relative influence over time. If your country under-educates its people, encourages consumer debt, and prioritizes spending over investment, it will decline. It really is that simple. We just get lost in the details.

Everything is like Jenga

The game Jenga is actually a contest of minimalism.  You start with a stack of wooden blocks as shown, and take turns removing blocks, until the tower collapses.

To a certain degree, everything is like Jenga.  The most efficient solution to just about anything involves removing as many extraneous pieces as possible without destabilizing the original aim.

Fewer pieces means less time, less curation, fewer errors, fewer losses, lower maintenance, higher return, higher efficiency.

Few people realize that Occam, who passed away in 1347, was actually talking about Jenga.  In a way.

Amazon fumbles simplicity

Years ago, I had a FewerNicer revelation:  buy everything with Amazon Prime.  It did not save money, but economized on something much more valuable:  time and attention.

Prime was not the cheapest, but ended shopping as we know it.  You think, “I need sunscreen and shaving cream” — one minute later, the order is in, the need struck from your to-do list, and you can get back to work, with barely any loss of focus.

Yet today, for the first time in years, I bought sunscreen and shaving cream from different sites.  Why?  As before, the change had nothing to do with price, and everything to do with time and attention.  Amazon is getting complicated, ruining their primary advantage.

My sunscreen brand is now sold only as an “add-on,” meaning I need to come up with $25 of other Amazon purchases before they will send me what I actually want.  Or I can wander  the site looking for another vendor of the same item at a higher price.  But this demands time and attention — exactly the cost I’m trying to minimize.

The shaving cream, a high-margin item, is now overrun with fakes and “new old stock,” sold by dubious third-party vendors that use the Amazon distribution network.  Again, I have to wander around the site, judging the quality and quantity of customer reviews, in hopes of finding the real item.  Decisions, decisions.  More time.  More attention.

Or, I can just go to the manufacturer’s website, buy direct, know that I have exactly what I want, and be done.  For today’s purchase, Amazon Prime lost the simplicity contest, which was my only reason for using it.

Manage to a single number

For a long time, I had various business goals — revenue per year, cost per month, widgets per microsecond, and so on — but business only took off when I boiled it all down to a single number I could measure daily.

Essentially, that number was “net daily change in gross profit.”  What was today’s revenue from new clients, minus revenue lost from any departing clients, minus the cost to acquire new clients?

Because my other costs were relatively fixed, that one number turned out to explain most of any change in fortunes for the whole business.  This, in turn, told me exactly how I was doing every day — a highly motivating and focusing flow of information.

Imagine how effective US economic policy might be, if managed to a single number.

The Federal Reserve once managed to a single number — inflation — but this was abandoned  in recent years, in favor of two numbers, namely inflation and unemployment.  There are legitimate reasons to want to reduce unemployment, but the two are often irreconcilable, and at minimum are a dilution of focus.  So this was a step backward from a management perspective.

Still, at least the Fed uses only two metrics.  Congress and the White House use an unlimited number, ranked in no particular order.  Cut inflation?  Control unemployment?  Flatten the income distribution?  Yes, yes, and yes.  Everything is important, so nothing stands out.  And then, nothing ever happens.

US economic policy could be hugely effective if we developed a single number that captured all the other goals.  The best metric, in my opinion, would be one that doesn’t exist yet:  median real disposable income per work hour.

What a mouthful!  But every word matters.  “Median” forces government to act in the interest of the middle class, because the median won’t go up much unless the masses earn more.  “Real” forces inflation management.  “Disposable” holds policy makers accountable to uncontrolled expenses like healthcare, education and housing, which for a generation have eaten up more than 100% of the real increase in middle-class income.  “Per work hour” acknowledges that workers’ time is not unlimited:  working twice as many hours doesn’t make them twice as well off.

If that single number goes up, Americans are generally getting better off, in a way that they will actually feel.  If it does not, then they are not.  Simple.

“I wanna boat.”

My friend bought a Catalina 36, a sturdy sailboat exactly the right size for cruising with family.  We sailed it this weekend, had a great time, and, as usual, my internal rogue compulsion to own things (see previous posts) began to cry, like a frustrated 2-year-old, “I want one!”

And, as usual, that urge made no sense.   There are sailing clubs in southern California that give you access to dozens of boats for $30 a month, plus $300 per day of actual use.  I’m in such a club, yet rarely find the time to use it.  Owning a boat, then, is a 100% guaranteed way to spend a ton of money with no concomitant increase in quality of life.

In general, renting is:

  • 10 times cheaper than owning (depreciation, maintenance, moorage).
  • Ties up no capital.
  • Predictable — no unexpected big costs.
  • No commitment — quit or rejoin anytime.
  • Flexible — choose from various boats in 4 ports at different price levels.
  • Time-efficient — no scrubbing decks, hiring repairmen, etc.

This tends to be true of other luxury items.  Don’t buy a vacation house — rent one through a site like VRBO, and stay in spectacularly nice places, anytime, anywhere in the world, with total flexibility and no commitment.  Don’t buy a Ferrari — rent one for a day, then rent a Maserati for a day next month, and so on.

So sensible.  Yet that immature little voice is still there.  I wanna boat.


The rogue urge for ownership

A challenge to minimalism is that rogue compulsion still seeking to buy things, when the rest of me is not looking.

Vacation home, boat, airplane:  the rogue compulsion would love to own all of these, even though it can be shown in seconds, on a spreadsheet, that renting is vastly cheaper.

Forget for a moment whether you or I can afford those — that’s not the point. In fact, the rogue is hardest to control with the most affordable things. Low price (or high income) makes it even easier to momentarily forget the various non-cash costs of ownership, the mental bandwidth, the maintenance time, the clutter, the loss of aesthetics.

There is no doubt Darwinian logic here.  We are clearly evolved from apes that survived jungle privation by saving torn t-shirts and yellowed newspapers.

Yet though I know this, the rogue persists, an evolutionary leftover, a mental appendix, uselessly urging me to accumulate junk, a squirrel gathering inedible nuts for a winter that will never come.

I let my guard down for just a moment, and suddenly find I’ve acquired yet another  free t-shirt, when I hardly ever wear t-shirts.  Or bought another physical book, after pledging years ago to use Kindle whenever possible.

I resort to desperate mental aikido to trick the rogue compulsion into acquiescence.  “If you buy something, you must sell something similar.”  Or, “just to be sure, let’s wait till tomorrow before buying.”  Or, “maybe there is something better you haven’t found yet.”

These keep my rogue at bay, but it persists, waiting for the next opportunity.

Capping my possessions

Every time I acquire an item, I try to get rid of a similar item.  Buy a shirt, give away or sell a shirt.

The idea is that ownership has costs.  Owning fewer things means less depreciation, storage, organization time, and more intense focus on the quality of the few things I still own.

The minimal family

In the previous post, I pointed out the connection between decentralization and minimalism.  This applies well to families.

By using simple signaling, simple rules and shared goals, you can massively simplify the organizational details of family life, freeing time for actual family life.  Here’s how we do it.

  • Divide financial responsibilities.  Agree in advance which spouse is responsible for which bills (including savings).  Set up separate bank accounts, and then just work independently.  Unless someone runs out of money (meaning the original agreement was flawed), then there is no need to argue or even talk about money.  It just takes care of itself.
  • “I spent more than 50 dollars today.”  Unexpected expenses do come up.  Agree on a daily trigger amount, above which each spouse must tell the other of an unusually high-cost day. This limits communication and/or conflict to things that matter.
  • Use shared e-calendars. Never verbally plan an event. Use Gmail, iCloud or similar shared calendars. Make one spouse the manager of the family schedule, and leave the other on a personal schedule. The family manager checks the spouse’s calendar, and if free, sends an invite to a new event. The non-managing spouse accepts, and then his/her phone or computer rings when it’s time to go. Pointless planning discussions vanish.
  • Teach your kids to do as many things as possible.  I’m not talking about chores, but just about knowing how to do things, so the kids can step in in a pinch.  I’m shocked at how many kids can’t wash clothes, cook a meal, walk or ride a bike to a friend’s house, change a flat bike tire, babysit a sibling, and so on.  The more they know, the more they can handle, and the less things will bottleneck on a frantic parent.
  • Everyone should have a phone.  Think of your family as a colony of ants.  Just as ants maintain an organized trail by signaling to each other by smell, humans now signal each other by phone.  Without a phone, there is no efficient signaling.  I used to think this was a huge waste of money, because smartphones are insanely expensive.  But then I realized that prepaid dumb phones are totally cheap.  You can buy your elementary-school kid a year of Tracfone minutes in advance, and it works out to under $10 a month — and you get the phone for free.  This may make it easier for some parents to get comfortable letting kids manage themselves.

That’s how we do it.  Well, OK, 80% of this is how we actually do it, while 20% (namely cellphones for all and learning to change a bike tire) is how I wish we did it.  But you get the idea.

This idea of decentralization, encapsulation and signaling is referred to in engineering as modularity.  It is the reason that object-oriented software is so much simpler than non-OO software.

Decentralization, signaling and minimalism

You can think of our giant human brains as an evolutionary cost of overcentralization.

To see why, compare us to ants and bees.  Those tiny brains accomplish highly organized group behaviors through extreme decentralization.  Individuals signal one other in simple ways — a scent, a dance — and peers respond robotically to those signals, resulting in amazing aggregate accomplishments.

Individual humans act with less coordination, and a cost of this is that we must be more flexibly intelligent, which requires carrying around a giant brain, which requires a giant body, which requires tons of food.  I’m not complaining, mind you.  Very glad I’m not an ant.  But this is an interesting way to look at the two forms of organization.

Decentralized decision-making, with simple communication signals, allows amazingly organized and economical group behavior, requiring little wasted effort, and little intelligence.

This idea — decentralization + signaling = minimalism — can be applied broadly.  It works in one’s personal life, in organizations, and even nation-states.  The more decentralized, the better, until you reach a point of diminishing returns (where the cost of signaling among all the independent actors is greater than the benefit).

One could argue that communism failed in the 20th century mainly because centralized supply forecasting was hopelessly complicated.  Without private ownership, independent buyers and sellers could not rely on price to signal supply and demand.  Instead, a huge government staff was needed to estimate — very badly — the supply requirements of a modern economy.  Computers would have helped, but still could not have kept up with the exponential increase in economic complexity over the past century.

Many of the most successful U.S. companies of the past 50 years — those that have built the most share value — have tended to run extremely lean, decentralized operations, except for a highly centralized capital allocation (investment) function.  The most extreme example:  Berkshire Hathaway owns over a hundred companies, employes over a hundred thousand Americans, pays about 2% of all U.S. corporate tax — yet it has only about 20 people in the head office, who are focused almost exclusively on investing.

This general idea suggests ways to manage the often pointlessly frantic suburban American family life.  See next post.