40 acres and a mule

November 05, 2018

One day, a farmer was tilling his land. The work was hard, but satisfying: he was in great shape, and rarely got sick. Life wasn't too stressful: he lived simply in a small house with two children and a wife who loved him.

Life was good; very Jeffersonian.

Until one day, the farmer met the economist.

The economist noticed how hard the farmer was working, and said to him, "I'll make you a deal: I'll give you equipment that will double your yield: 80 acres worth of crop from your 40, in exchange for 15% of your harvest."

The farmer thought about it. 15% of 80 acres meant he'd give 12 away, and keep 68. That means he'd get 68 acres worth of crop, when he used to get just 40.

"And I wouldn't have to work any harder?", the farmer asked.

"No, not one bit. We'll supply the equipment, you use it, and your yield will double."

"What happens if my yield doesn't double? Do I still owe you 15%?"

"We'll take our 15% no matter what. Most people who use our equipment double their output, but a few find it doesn't work. On the other hand, we've seen some people triple their output. We can't say exactly how well your farm will do."

The farmer wasn't too sure about this. Why ruin a good thing? He was happy, he had enough, and just wanted to run his farm.

On the other hand, he liked the idea of getting more output for the same amount of work. 15% seemed like a small price to pay for nearly double the output. And he knew that if he didn't find a way to work more efficiently, the farm wouldn't be enough for the next generation; his kids were college-bound, and he didn't have the income for that in his current situation.

Also, if he could work just a few years with much greater output, he could retire a little early: more time to visit the kids at college, travel with his wife, and read all those books he never quite got to. That is, if it worked.

"I'm not sure", the farmer replied. "I'll have to talk to my wife about it."

"No problem", replied the economist. "Just don't wait too long. I'll be here, but every day you work, you're leaving a lot on the table. Why not take the deal and double your output? The risk is on us; if things don't work out, we're out the equipment, you just lose a little bit of your harvest."

The farmer liked the deal, and yet, he always wanted to be his own man. That meant controlling his destiny, not relying on magic beans he didn't understand from strangers.


Office Space

November 04, 2018

Office Space is coming up on its 20 year anniversary: almost old enough to drink.

I'm not surprised Mike Judge's went on to create "Silicon Valley"; his depiction of technology work is so spot-on, so utterly right, it could've been written today, in late 2018.

Some of the highlights:

  • The traffic: the movie opens with a scene of soul-crushing commuter traffic, just like the kind every weekday on 280 or 101. There's a reason I've avoided driving to work
  • Peter lives in a bland, dystopian box of an apartment; my friend John used to live in one of these places, while working at Intel, in Sunnyvale.
  • Peter has eight bosses: the joys of matrix management!
  • The office is loud, cramped, and requires fighting off a never-ending stream of interruptions ("Corporate accounts payable speaking, just a moment!")
  • "No one in this country can ever pronounce my name right. It's not that hard: Na-ghee-na-na-jar. Nagheenanajar"
  • The emotional tone: great people, somewhat meaningless work, and the sense of sameness: every building looks the same, every day starts and ends the same way, an identical sea of cubes.

Not all tech work is like this, but if you spend any time doing it, you're certain to see a few things you'll recognize.

20 years has passed, and I can't believe how little change there's been.



November 03, 2018

If the supply of risk capital declines, what then?

I think we'll see a lot more Craigslists: sustainable "businesses", "operations" really, that earn enough money to fund themselves on an ongoing basis, even without much growth.

It strikes me that we lack a category for Craigslist. They clearly aren't trying to maximize profit. And yet, nonprofit doesn't feel quite right, either: the connotation is too "humanitarian" (arts, healthcare), with funding from donors. Somehow a website for apartment and job listings doesn't seem to sit right, in that category.

But maybe my thinking is too narrow. Nonprofits do everything from disaster relief (Red Cross) to looking after real estate (my condo association), to museums and even international aid (Rotary).

I get a lot of sideways stares when I call a tech company a "nonprofit". But there's no reason we couldn't have more Craigslists and given how efficient modern software stacks are, and how things are stabilizing, I think we'll see a lot more stuff like this: not hockey-stick venture growth, but too many active users to shut down.

As I thought about it more, I came up with the following taxonomy of funding models:

  • Traditional venture capital: always understood to be temporary. Growth extremely important.
  • Revenue, for profit: How the Fortune 500 runs. Long-term stable. Growth important.
  • Revenue, nonprofit: same as above but less emphasis on maximization and growth.
  • Patronage: solicit donations from wealthy benefactors (many arts organizations use this)

The distinction is fuzzier than it first seems. A lot of venture companies won't hit their numbers (most don't) and might end up with more of a nonprofit/patronage model if they aren't shut down outright. In any case, sites like Indiegogo and Kickstarter show that revenue and venture capital aren't the only game in town.


Risk capital

November 02, 2018

I can still remember how ragey the business community got over facebook's early valuations. facebook is not worth $33 billion, they wrote. There was an oddly righteous tone to it, like, I can't believe these guys have the arrogance to raise at that valuation. Who do they think they are? No revenues and they're asking for HOW MUCH!?

I can't say I blame them; it still seems absurd, six years later, that facebook paid billions for two social networks in 2012-2014, neither with significant revenue: first $1 billion for Instagram (2012), then an even more eye-popping $19 billion for WhatsApp, just two years later (2014). I remember standing in my kitchen in San Francisco thinking about it; facebook paid half the value of Ford (the car company) for a company with around 50 employees and minimal revenue. It was unprecedented; never had a company with so few tangible assets or cashflow been purchased for that much.

I took me a long time to see this, but I realized it's all just supply and demand. Business school teaches that securities (e.g. stock, how ownership is transferred) should be valued on more objective criteria, things like discounted cashflow: project the cash flows, and calculate based on those. Perhaps that's a good for calculating what we ought to pay—a finance academic would probably argue that's correct—but what people do pay is based on a market process, just like houses or public companies.

And that market process runs on supply and demand. Specifically, when raising funding for early-stage companies, it seems to work like this:

  • At any point, a certain amount of equity (ownership) is on offer, "for sale" to investors, and
  • A given level of money—"risk capital"—is on offer, which purchases the ownership.

As I said, no different from houses or public companies.

The key insight of this model: supply of risk capital drives valuations (prices).

So how do these factors (supply and demand) vary over time? Interesting question. One thing I've noticed: company formation happens much more slowly, and at a much more constant rate, than investment, which is why capital availability predicts valuations. Company formation is affected by things like:

  • Confidence—how many people are leaving their jobs to found companies?
  • Perception of opportunity: are investment bankers and lawyers spilling over into long-term careers in tech? (I've seen a lot of this over the past decade)
  • Do people think there's specific opportunity in a space? (Social networking, robots, AR/VR)

Whereas risk capital is more of a "Wall Street" thing, driven by:

  • Is early-stage tech perceived to be a good investment? (2008: Nobody investing in anything. 2012: INVEST IN ALL THE THINGS. 2018: Most people say "yes", smart money beginning to wobble)
  • What else is available? Are other asset classes yielding enough? (2008-2016: No. 2016-now: interest rates going up)

In the past year, I've noticed a bit of tapering. Still lots of money flowing, but more to later-stage companies. Apartment rents are softening; I see more office vacancies. The overall tone is still optimistic, but more cautious than a few years ago. Fewer new things seem to be getting traction.


Opportunity recognition

November 01, 2018

I was talking to my friend Tom over the weekend about opportunity recognition, the skill of "seeing what's next". Though I live in Silicon Valley, I've never had the confidence to "bet the farm" on a particular idea, doing something crazy like dropping out of Harvard and moving halfway across the country, or running up $25K in credit card debt, to pursue a business idea.

Nonetheless, opportunity recognition is an important skill in business; if you're going to commit resources—time, money, attention—you'd better be able to tell whether something will work.

That's when I remembered this picture from my friend Kurt:

Clara at the sound board

Kurt's daughter Clara is 2 or 3 years old, but she's already "running sound" at church, just like her dad. He sent this as kind of a joke; obviously a 2-year old can't operate a sound board. But she's learning young, and if this is something she enjoys, she'll have logged a lot of time by young adulthood. I have no idea whether she'll be interested in sound design later in life, but were she to be, she'll be like the boy who did free throws with his dad starting at age 5; practice matters a lot.

Thinking of high achievers I've known, they benefit in numerous ways from their families:

  • They learn skills: painting, writing, negotiation
  • They're taught habits: grit, hard work, persistence
  • They receive access through their family's social and professional connections
  • They receive cash which lets them focus on what matters

I still think people are mostly responsible for their own success, but being born in the right family certainly helps. And I hope I can do these things for my children someday.

Looking at it this way, I'm not surprised opportunity recognition wasn't someting I had in my 20s; we had no businesspeople in my family, so I'm not sure where I'd learn it? I also think truly great businesspeople don't so much "take risks" as "identify situations with tons of upside and very little downside".

Nonetheless, I'm getting better at this, though still on the lookout for ways to improve more quickly; it helps to write it down.


Pick your neighbors carefully

October 31, 2018

Dockless scooters are currently the biggest nuissance in my life.

They wouldn't be, if their riders were considerate of others, as I'm sure they are in many places. I have no doubt there are many places where scooters have brought nothing but delight. Whereas here in Oakland, it's a daily game of pedestrian bowling, as kids careen down 3-foot wide sidewalks at 15-20 mph, doing their best to dodge pets, delivery drivers, young children, and people entering/exiting offices.

Riding scooters on sidewalks in Oakland is not only dangerous, it's illegal. But since most people who do it are young Black Americans, law enforcement action is deemed "hateful" (children) and "racist" (due to the color of their skin). Even if the cop simply stops the kid and asks him to slow down, or ride in a bike path, on the street.

I don't really see a way out from this problem. Technology is another word for "tool"; as more tools (technologies) come to market at an accelerating rate, their users will be empowered to do more things, with greater power and speed, than they were before. The same force that makes scooters a nuissance—increasing access to more and better tools—underlies fake news on Twitter and facebook. Giving everyone access to powerful tools might be a net positive thing, but there will be some negatives, too.

Pathological optimists (like the people here), think the law can solve these problems. I doubt it; there are too many new things appearing too quickly, and even if we could keep up, regulation of every tiny aspect of life is unenforceable. And also, not very fun. I don't want to live in a place with police everywhere.

In Code 2.0, Larry Lessig lays out four ways to regulate behavior: law (make it illegal), norms (make it taboo), economic (make it too expensive), or what he calls "architecture" (make it impossible via fences/gates/software). Of these, I think architecture is our only hope to preventing misuse. That requires active effort from the businesses who make this stuff, which seems unlikely. Can you imagine a cigarette company making cigarettes that are less pleasurable to smoke? I can't.

Maybe the only solution is to live with people whose norms are similar to yours—in other words, pick your neighbors carefully. In any case, law won't solve our problems.

In fact, even that won't be enough. We'll have to worry not just about new things, but also about existing things becoming more addictive. That's what bit me. I've avoided most addictions, but the Internet got me because it became addictive while I was using it.

—The Acceleration of Addictiveness


Dad sayings

October 30, 2018

My dad "isn't a rocket scientist", as we'd say in Chicago, but he knew a few things. His wisdom came to us through choice sayings, always a bit dour (in the characteristic Lutheran way):

  • Parking lots are complete chaos. Chaotic environments are dangerous. It would be hilarity to see an autonomous vehicle try to navigate the sea of lane-line crossings, angle-in parkers, backward drivers, strollers, pedestrians lurching into the street, and tons of small kids, in front of any big box store.
  • Moisture is the enemy of the home. Water causes many (most?) household problems: mold, rot, foundation damage, etc. Never ignore water. Larger lesson: be responsible and don't ignore taking care of things. You'll save a lot of money and time fixing small problems before they become large, and urgent.
  • Call them first. Time is precious; optimize. Be smart. Don't do today what you could do in half the time tomorrow (via batching or going somewhere closer to home). Avoid making special trips. Don't go to stores that aren't open or don't have what you need. Unforced wastes of time are the plage; avoid them at all costs (to this day, they're a major peeve of mine).
  • Discretion is the better part of valor. Sometimes it's OK to walk away from a fight. Don't start big projects when you're tired or try to lift too much by yourself; you'll just end up hurting yourself, or doing sloppy work you'll have to redo.
  • It's always something. Murphy's Law says "anything that can go wrong, will"; there are a lot of things, so expect something is always going wrong. Human vices—forgetfulness, carelessness, sloppyness—always get in the way of things. Be tough; deal with it, keep going, and above all, remain cheerful.
  • Dollar bills flying out the door/window. Said when a door/window was left open during the winter, letting the heat get out; a reminder that big wastes of money are often small, ongoing things (e.g. credit card interest, water leaks), and to avoid that cardinal Midwestern sin: wasting money.

My dad embodies what Stripe (the payments company) calls "micro pessimism but macro optimism": everything is always broken, messed up, and wrong, but in the end, things will be better than you can possibly imagine.

I love that.


Switching lenses

October 29, 2018

Another thought on pathological optimism: "switching modes". It's an ability I'm trying to develop.

My friend's mom, a Black American, was voting in the US midterms in Atlanta. As he explained it,

they had a 3+ hour wait because they didn’t have power cords for the voting machines. She asked a guy and he said “oh well the state never gave us cords for these so when the battery dies that’s it”

I make a point of assuming the best intentions, so my first inclination is to take the explanation at face value. But something else could be going on; maybe someone is trying to suppress votes?

Most people I know see it one way or the other; it doesn't occur to them, as a matter of habit, that not only might other people see things differently, but that both could be right. Such half-truths melt my brain—there's something so difficult about training one's brain to habitually see both sides. Not "he thinks X, she thinks Y, but I agree with Y", but "he thinks X, she thinks Y, I think it's something like 65% X and 35% Y".

I find it incredibility difficult, yet useful, to see things this way, a sort of quantum superposition for everyday opinions and gossip.

Rubin writes about "switching lenses" a lot in his book, In an Uncertain World; I might be thinking about it because I'm reading that book again (one of my favorites).



October 28, 2018

I was discussing the midterm results with some friends yesterday, one of whom is a Black American (a term I love, as it carries so much history, and pride). He had this to say:

Ending on a positive note: this is my grandmothers voter registration card from 1966. Her and my Granddad went to get it the day after the Voting Rights Act was passed. She told me a year before she died that my grandfather took a bunch of his coworkers in his truck to vote for the first time because the white men who owned the businesses in town would drive past the polling place and if they recognized your car, you got fired as soon as you got back to work.

The right to vote

Progress is possible; don't forget it.


Starsky Robotics

October 27, 2018

I started a new job at Starsky Robotics yesterday. Their people ops manager asked me to write an answer to the question, "Briefly describe your past work or school experience that led you to this work".

My answer:

I was hand-etching printed circuit boards and working on cars since before I had a driver's license. I was pretty sure I'd end up doing hardware after college (I have several close friends who went into chip design) but I met Bill Gates in 2005 and he convinced me the real action was in software. Software also seemed to be where most of the smart people I knew ended up. I spent the next ten years building just about every kind of software system—a database engine, a hotel booking system that moves millions of dollars, an email productivity tool, analytics that tracked app performance across billions of mobile app installs, a natural language-based booking system, rental property management software, a music-based ad network, a project dashboard system that generated Powerpoint slides automatically, even an e-commerce paint site that matched colors based on metamerism—but never really found a "specialty".

When I wrapped up my last thing, I decided I wanted to go really deep in something, so I looked for what I could see myself doing for the next 10 years or longer. It had to be something that was going to be relevant on that time scale, something with a lot of technical depth, where I could apply my breadth of knowledge across hardware, probability/stats, nuts-and-bolts software engineering, and something with a lot of business impact. I settled on fixing healthcare or autonomous robots and ultimately chose robots because I think the challenges with healthcare are more regulatory/political, but we're on the cusp of a massive golden age of innovation in autonomous robots.

So that's what I'm doing now.

The place is better than it looks. They're a little under-the-radar—no mass-market brand like Coinbase, Stripe, or some of the other places I considered—but they're a small team that's been able to do a lot.

These guys and gals have real breadth; autonomous vehicle control means executing across (1) hardware design (lots of cameras, sensors, and ruggedized electrical design), (2) perception (making sense of all the data from the sensors using machine learning and lots of math), (3) motion planning (both local and global), and (4) cloud services/ops.

One thing I noticed already: Starsky has a very different engineering culture than I'm used to. Most places I've worked have had teams of people who, while not exactly interchangeable, could at least "pinch hit" for each other, e.g. an infrastructure developer might be able to throw together a crappy web UI or mobile app. At Starsky, if the cloud ops people wanted to work on the low-level C code that turns the truck's steering wheel, they wouldn't have a clue how to do it. That's new.