The WordPress and open source communities have lots of thoughts on Matt Mullenweg’s actions after his talk at WordCamp. Much of this commentary has overlooked what I think is a core concern at the heart of Matt’s argument: the growing role and influence of private equity in tech, particularly in open source.
PE firms raise money from institutional investors and high net worth individuals to buy companies, restructure them, and sell them later for a profit. They’re all about maximizing returns ASAP, usually in a 3-7 year window. That’s a shorter timeframe than the typical founder or public company shareholder, and far shorter than Matt’s time horizon:
I would like future generations to grow up with a web that is more open, more free, gives more liberty, and so open source is really my life’s work, even above WordPress and anything else. I hope to work on it the rest of my life.
When a PE firm acquires a tech company, a few things happen. The new owners focus on aggressive cost-cutting and “streamlining” operations to boost short-term profitability. This is extremely destructive, leading to mass layoffs, reduced benefits, and wholesale outsourcing. At the same time, PE firms often load up acquired companies with debt, extracting cash while leaving the businesses more financially unstable. With a myopic focus on an exit, long-term investments in innovation, customer service, and employee development fall by the wayside.
The result? Private equity backed companies are 10 times more likely to go bankrupt than public companies.
The basic business model of private equity firms often leads to disasters like that at ManorCare for three fundamental reasons. First, private equity firms typically buy businesses only for the short term. Second, they often load up the companies they buy with debt and extract onerous fees. And third, they insulate themselves from the consequences, both legal and financial, of their actions. This leads to a practice of extraction, rather than investment, of destruction, rather than creation. While not every company owned by private equity firms goes bankrupt, the chance of disaster meaningfully increases under their ownership. — Plunder by Brendan Ballou
Open source communities are particularly vulnerable to this model. They’re delicate ecosystems of volunteer contributors, transparent governance, and values-aligned sponsorships. Applying the PE playbook can easily destabilize that entire balance.
Take, for example, Vista Equity Partners’ $1 billion acquisition of Acquia in 2019. Acquia is the main commercial backer of Drupal, another popular open source CMS. In the years since the acquisition, the Drupal community has visibly struggled. Acquia appears to be contributing less code and providing less financial support to the ecosystem.
With fewer resources, core development and maintenance have slowed. This coincides with a sharp decline in Drupal’s market share compared to competitors like WordPress:
The Drupal story demonstrates how applying PE tactics to open source projects can undermine their health and sustainability. By nature, these communities aren’t a fit for rapid extraction of outsized returns. Their vitality depends on reinvestment, transparency, and long-term stability — the opposite of the PE model.
Toys “R” Us. Simon & Schuster. Even America’s ports. Not everyone can push back against private equity, but I believe we should support companies like WordPress that have a long history of admirable values and prescient leadership. WordPress powers more than 40% of the web — its community’s health is too important to ignore.
I recently ordered a Comma.ai kit to add semi-autonomous driving capabilities to my 2018 Mazda CX-5. It took some trial and error (and help from the Comma.ai community on Discord) to get it working properly, but I highly recommend it as I’ll explain later.
I’m sharing the steps I took in case it’s helpful for others looking to install the Comma 3X on the same or a similar vehicle. It’s a pretty intuitive process, all things considered, but there weren’t a lot of guides with photos on how to do the actual installation process.
A false start
When I initially received the Comma 3X kit and installed it (follow the installation video and Jafar’s post, it’s pretty simple) I thought I’d be up and running quickly. However, the device seemed stuck in dashcam mode and didn’t move the steering wheel at all. After digging into the Discord server and reading more, it appeared to be an issue with fingerprinting my specific Mazda model.
I reset the device and installed a special fork I found on the Discord server which had more Mazda car fingerprints to help identify how to interface with the car. This worked on a test drive, but the car would beep aggressively every 5 seconds if you weren’t holding the wheel very deliberately, making highway driving difficult. Even without Comma installed the car would complain when you had cruise control enabled and weren’t holding the steering wheel in the right place, but adding the Comma made it complain much more.
The Discord community recommends trying a wheel cover with weights underneath, which worked somewhat but was quite bulky. Ultimately, I decided to pursue the ultimate solution I had been trying to avoid: swapping out the EPS motor.
Swapping the EPS motor
The EPS motor on pre-2022 Mazdas is designed for very lightweight lane departure vibrations. I always thought it felt a bit underpowered, and you could feel the car lightly bouncing between the lane markers. On any serious turn, the car would just continue to depart the lane.
Newer Mazda EPS motors can apply more torque to the wheel, allowing for real LKAS (lane keep assistance) that can turn the car even through 10+ degree curves. I was initially hesitant to replace the EPS motor because I wasn’t sure exactly where it was located. But after reading the pinned messages on Discord and asking a few questions, the community reassured me it was pretty straightforward.
The process ended up being easier than I anticipated, although there were a few tricky parts. Here’s a detailed breakdown of how it works:
Order the new EPS motor from eBay — the LKQ store has several options, but you have to make sure the vehicle is model year 2022 or higher. As an example, here’s a search on LKQ’s eBay store, or just look for “2023 Mazda CX-5 Power Steering Pump OEM”. You’ll also need a 10mm and 12mm wrench for later.
You’ll receive a hunk of metal similar to this. The only part needed is the circular gearbox-looking piece on the left of this picture.
Remove the two bolts pictured here, which will detach the shaft from the EPS motor. Use a 12mm wrench for these and be careful not to lose the small plastic gear in the center of the motor once you remove it.
While not required, I recommend removing the cover to the left of the steering wheel. It pops out easily and improves access to the EPS motor.
Now that you’re in, you’ll see the same motor mounted, as pictured here:
Here’s the trickiest part: you need to remove the same two bolts on the old gearbox, but one bolt is close to a piece of metal, making it hard to fit the wrench in. It’s doable but will take a few attempts. Both bolts are on quite tight, so don’t be afraid to use some force to loosen them.
Once the bolts are out, unhook the 3 cables at the top and detach the old motor.
One tip: the small gear kept falling out or misaligning as I tried to install the new motor. Rather than fumbling with alignment, I positioned the gear on the other (already mounted) side so it was already lined up correctly. You’ll see what I mean.
Screw the bolts back in, reconnect your battery, and you’re all set! You’ll want to install this modified fork on the Comma 3X, which tells older Mazdas to use the new EPS motor.
Review
After installing the new EPS motor, I’ve been really impressed with its performance. I wouldn’t even bother installing Comma.ai on an older Mazda without it — the experience is just too annoying otherwise.
The upgraded EPS motor enables stop-and-go adaptive cruise control, which by itself is a fantastic improvement for city traffic. It also provides LKAS that keeps the car smoothly centered in the lane without any warnings to keep your hands on the wheel. These enhancements make long drives much more pleasant.
With the new motor, I’ve completed a few hundred-mile road trips with only three or four interventions to assist with turning, and just a handful of additional interventions in busy city traffic. It really is quite capable.
Since I had Comma installed before the EPS motor upgrade, I spent some time observing the Openpilot lane markings as we drove. There was never a point where the green outline indicating the car’s projected path was misdirected or where the detected lane markers deviated from reality. I also drove at night and in the rain, with no issues whatsoever. This gave me a lot of confidence in the system’s capabilities.
Special thanks to the Comma team for building an incredible hardware product, and to all the maintainers of Openpilot and volunteers in the #Mazda Discord channel.
Citymapper is what happens when you actually understand user experience.
This post was originally published on Medium in 2014.
Every so often, an app comes along that just completely understands the way you think. I don’t normally write long posts about an app I’ve used. But Citymapper is so incredibly well-made that I decided to put together a list of common use cases of a maps app, and how both Google Maps and Citymapper handle them.
Scenario 1: I need to get somewhere as fast as possible.
Citymapper, 10 seconds: I can see all 4 options (public transport, walking, car/taxi, and biking) side-by-side, allowing me to compare options quickly.
Google Maps, 20 seconds: I need to click through each individual option and remember which is the fastest. Also, the flyover animation is cool but unnecessary when I’m trying to see what the fastest time is.
Scenario 2: I want to see the time needed to get from point A to point B.
Citymapper, 10 seconds: I can enter my origin and destination at the same time.
Google Maps, 30 seconds: I need to first enter my destination, then specify how I want to get there, and then change my origin from “My Location” to a different address.
Citymapper, 2 seconds: Since I know the subway system pretty well, I don’t always need full directions — I just need to know where the closest subways are so I can judge which one I should take. With one click on the train icon, I can see all the trains near me, sorted by distance. Same thing for buses.
Google Maps, 30+ seconds: There’s no way to see all nearby trains, so you have to manually look around your waypoint and click on each station. There is a public transit option in the sidenav, but it only highlights the lines on the map and you still need to click through to see which stations are closest. And at transit hubs (in this case, Union Square) it considers different subway entrances and lines as separate, so it won’t let you see aggregate information for a particular station.
Scenario 4: How long will it take me to get somewhere by Citibike?
Google Maps, can’t do it: Google Maps doesn’t have bikeshare integration, so I’m forced to switch apps and use the equally scatterbrained Citibike app to find availability.
Citymapper, 15 seconds: Not only does Citymapper find where the nearest Citibike location is and let me choose whether I want the fastest route or the quietest (read: safest) route, it also notices that the closest Citibike location to my destination has no available spaces and reroutes me to the next closest one that does.That’s magical.
Scenario 5: I just searched for directions at the office, and now I want to see those directions again.
Citymapper, 5 seconds: Previously searched destinations can be found right underneath the search bar for easy access.
Google Maps, 10 seconds — if you know where to find it: For some reason, previously searched destinations are found under your profile (the person icon next to the search bar) instead of when I search. I never noticed that until I started writing this post.
Scenario 6: I’m late to a meeting so I’m sprinting as quickly as possible down the street, looking at my phone for directions.
In any other app, shaking my phone might mean “undo last action” (e.g. Mailbox) or “submit feedback” (e.g. some beta builds of mobile apps). In a maps app, though, there’s pretty much one thing a shaking phone means — I’m running, biking, or just moving. And when I do any of that with Google Maps, this happens:
It blocks my view, it shuts off turn-by-turn directions, and is altogether a pretty terrible interaction.
Other design details
City icons
Going to a different city? Citymapper’s customized icons for each of its supported metro areas gives the app personality.
Offline subway map
If you’re in the subway, you can access a full offline zoomable subway map by clicking on an icon on the homescreen. They also have a “Save trips” option on the directions screen so you can save directions for offline use (e.g. on a subway).
Humor
First, their release notes are legitimately funny (see Twitter examples).
And when presenting your transit options, Citymapper will occasionally throw in a funny option:
or sometimes:
But I guess I’m easily amused.
Final thoughts
Citymapper is more thoughtfully designed for urban travel. But for people who drive cars, something like Waze — with its fullscreen interface and crowdsourced realtime information — is a better choice.
And there are still features I would love to have in Citymapper. Turn-by-turn directions. Spoken directions for when I’m on a Citibike. Maybe personalized time estimates that are tailored to my walking, running, and biking speeds. But whatever Citymapper adds to the app in the coming months, I know it will be thoughtfully engineered and designed to make finding directions almost fun.
If you’re so inclined, you can download Citymapper here. (And disclaimer: I have no relation to Citymapper. Just a happy user. They didn’t even know I was writing this post.)
This post was originally published on Medium in 2015.
In 2014 I was lucky enough to speak at Morning Prayers, a secular Harvard tradition that has existed since its founding in 1636. Below is a copy of my remarks.
Good morning. My name is Zachary Hamed, and I am a senior in Leverett House studying computer science. I’d like to begin with a reading from Ecclesiastes:
The race is not to the swift or the battle to the strong, nor does food come to the wise or wealth to the brilliant or favor to the learned; but time and chance happen to them all.
Chance—and its sister, serendipity—are the under-appreciated forces that drive the twists and turns in life. Serendipity is defined as the “the occurrence and development of events by chance in a happy or beneficial way.” It was first penned by Horace Walpole in a letter to Horace Mann in 1754, who said he was inspired by an old Persian fairy tale called “The Three Princes of Serendip.”
As the story goes, the King of Serendip had three sons, whom he had educated with the best tutors in the land. After several years, the king felt his sons had mastered all the knowledge of the arts and sciences, but feared they had been too sheltered and privileged in their education. He sent his sons to the desert, where they wandered for days.
Late in their trip, they met a farmer who asked if they had seen his missing camel. The princes asked the farmer if his camel was blind in one eye, had a gap in its teeth, and was injured in one leg. When the farmer heard this, he accused the princes of stealing his camel and took them to the king.
When the king asked how his sons knew such intimate details of a stranger’s camel, they explained that their days of wandering had led them to notice small details in the environment. An animal had eaten the tougher and browner grass on the left side of the road, leading them to believe the animal was blind in one eye. There were lumps of chewed grass along the road that were the size of a camel’s tooth. And there were only 3 camel footprints on the road with the fourth being dragged, leading them to believe the camel was disabled in one leg. As soon as they finished, a traveller entered and reported he had found a lost camel in the desert. The princes were rewarded handsomely and their intuition trusted for years to come.
The story hits close to home because Cambridge resembles Serendip all too often. It is easy for us to be so ensconced in our research, our work, our meetings and our routine that we don’t make time for productive and creative wandering.
In the 1950s, a pair of medical researchers—one at NYU and one at Cornell—accidentally injected rabbits with a certain enzyme and noticed that the rabbits’ ears folded over. Neither had the time or funding to pursue the anomaly, so they buried it away in their lab notebooks. Several years later, only one of the professors decided to review his research notes when he realized the flopping ears were actually more important than his initial research was. His work led to a Nobel Prize, and a subsequent sociology paper on the topic coined the terms serendipity gained (for example, reviewing your notes and deciding to pursue a question further) and serendipity lost—making the same discovery but never following up.
In short, serendipity enters our lives several times a day. We can gain it or lose it. But the best part is—we can engineer it. MIT Professor Ethan Zuckerman covers this phenomenon of creating luck. “Engineering serendipity,” he says, “is this idea that we can help people come across unexpected but helpful connections at a better than random rate. And in some ways it’s based on trying to reassess this notion of serendipitous as lucky — to think of serendipitous as smart.”
Harvard engineers serendipity at every point it can. You’re placed with a random group of your classmates when you enter as a freshman. Even when you’re allowed to choose your roommates in sophomore year, you’re placed randomly into houses and a lottery randomly places you into rooms. The lectures in your general education class could help you on a completely unrelated assignment in a different class. And have you noticed how many walking paths there are outside this very church? Walking through Harvard Yard is a case study in serendipity—it’s almost impossible to avoid seeing someone you know.
We meet best friends at events in Ticknor Lounge, significant others in early morning sections in Sever, mentors over lunch in Leverett dining hall, and overhear job opportunities in Lamont Cafe. Every minute on Harvard’s campus is a chance encounter waiting to happen.
Harvard’s secret is it takes a diverse group of people and squeezes them into a small, confined space. But the real world is a big place. And the scariest part is that serendipity will no longer be engineered for us. In fact, New York City, where I’ve been living for the past 4 months, seems to be engineered against serendipity. Have you ever tried meeting someone new on a New York City subway train? In a city of 8 million people, how do you find your new best friend? Your new job opportunity?
As always, Ralph Waldo Emerson provides the answer. “Shallow men,” he says, “believe in luck or in circumstance. Strong men believe in cause and effect.” It is up to us to engineer serendipity once we leave Harvard’s gates. So talk to the person behind you in line at the coffee shop. Knock on your neighbor’s door in the otherwise anonymous apartment building you live in. Ask your friends to introduce you to someone new and go to lunch with them.
Be open to the unexpected today. Relish the uncertain. And embrace serendipity.
Last spring, I took a class on social psychology. It was one of the most useful classes I’ve ever taken, so I figured I’d share some of the more interesting findings. Each paragraph has a citation to its right so you can get more information or read the original study.
1. Reciprocity has a strong effect on us.
20% of people send Christmas cards back to people they’ve never met, just because they received one from them. For the same reason, tips to waiters go up 3.3% when an after-dinner mint is provided with the receipt. And when the server looked the diner in the eye and gave them a second mint? Tips went up 20%.
2. You attribute a higher value to things you already own—this is known as the endowment effect.
Willingness to sell was twice as high as willingness to pay in one study. In other words, participants were willing to buy a mug for $5, but once they owned it, they wouldn’t sell for less than $10.
3. Heat makes us angry, and sadness physically makes us colder.
When you feel rejected, you report the room as being colder and you prefer warmer foods over colder foods. Crime rates are higher in hotter regions, and crime is more likely on warmer days. Baseball pitchers are more likely to hit batters when it’s hot. This occurs because heat causes arousal, but people misattribute that arousal to situations around them and not to the heat.
Duchenne smiles (example B) that are exhibited in high school yearbook photos are correlated with better life outcomes 30 years later. Here, Paul Ekman—an expert in facial psychology—exhibits both non-Duchenne (exhibit A) and Duchenne smiles.
4. Smiling is contagious—and can predict your happiness, professional success, and lifespan.
Humans laugh more at movies when other people laugh. Additionally, many people smile at getting a strike in bowling only after they turn around to their friends—you smile for the social approval, not for doing something successfully. In another study, students who exhibited “Duchenne smiles”—a more authentic type of smile that engages the eye and mouth muscles—in their high school yearbook were more likely to get married and were more likely to self-describe as “happy” 30 years later. Students with less intense smiles were more likely to be divorced. And in any given year, people who exhibited Duchenne smiles in their high school yearbook were half as likely to die.
5. How we’re approached and our desire to be consistent affect our decisions.
If I asked you to volunteer for an “Experiment at 7AM,” would you do it? What about a “7AM experiment”? 56% of people asked to volunteer for the first did so, but only 24% volunteered for a “7AM experiment”—fewer people want to wake up early, so the ordering of the words matters. In another experiment, some participants were called and asked if they would hypothetically volunteer for the American Cancer Society. When they were contacted a few days later and asked to volunteer, 31% agreed—versus 4% of people who were cold-called and asked to volunteer for the first time.
6. We act differently when reminded of who we are.
When participants were told that men and women scored differently on a particular test, female participants’ performance dropped dramatically. Male participants’ performance on a task dropped after interacting with an attractive female participant. When children are in a group on Halloween, they take more candy on average—but when children were singled out and asked their names, they took far less candy.
7. Being watched sometimes helps—except when eating.
Having an audience of people watching you complete a task improves performance on simple tasks but hinders performance on more complex tasks or when learning a new skill (they showed this with both humans and cockroaches—don’t ask). The mere presence of someone in the room causes this effect; even a repairman working on something in the corner slowed people down. Yet when it comes to eating, a full chicken will overeat in the presence of another chicken, and animals eat more in pairs than when alone.
8. Comparing people to their friends is the most effective way to make them do something.
When an electric company tried to encourage people to save energy at home, telling them “your neighbors are reducing their energy use” led to a 2% reduction in household usage. Telling people “save energy to save money” or “save energy to save the environment” did not decrease, and in some cases increased, energy usage.
9. Context—where we do something—has a substantive effect on what we do.
56% of actual voters voted for a pro-school budget when voting in a school vs. 53% otherwise. While that effect may not seem huge, it’s statistically significant and was reproduced in a lab environment (64% of people voted for a fake pro-school budget when shown pictures of a school vs. 56% who voted for it otherwise).
10. The more you’re exposed to something, the more you like it—this is called the mere exposure effect, and it works in milliseconds.
Participants shown a foreign word frequently were more likely to say the word had a positive connotation. The most immediate application of this effect is advertising; the more often you’re exposed to a commercial or ad, the more positively you will rate the company. Flashing images that elicit positive or negative emotions for only a few milliseconds subliminally conditions your attitude.
People liked familiar objects more than abstract patterns, but in both cases participants overwhelmingly preferred curved objects over objects with sharp edges. Objects in category C (featuring both curves and edges) fell in between.
11. Curves > Edges.
Humans overwhelmingly prefer curved visual objects over objects with jagged lines.
12. Don’t get hurt when there are a lot of people around you.
Bystanders are less likely to intervene in a crime or help in an emergency as the number of observers increases, as each individual feels that someone else will help and responsibility is diffused. When a victim is bloody, people help less often—likely because there is a chance they would be exposed to pathogens. But victims who scream receive more help than those who don’t—clear and unambiguous danger is helped far more often than not.
13. We really want to be happy, but being too happy can negatively affect our work.
In a study of 10,000 participants from 48 different countries, happiness is rated as more important than any other personal outcome—more than finding meaning in life, becoming rich, or getting into heaven. Happy people more often label themselves as “curious,” and depressed people are more likely to notice small changes in facial expressions. Yet extremely happy people (9/10 or 10/10 on a happy scale) got worse grades and had lower salaries than moderately happy people (6/10, 7/10, or 8/10 on a happy scale).
14. We do stupid things because we want to conform.
In one study, a participant was placed in a group and asked to answer a seemingly simple question. The rest of the people in the group were all told to respond with the same incorrect answer to the question, after which the participant was asked to answer in front of the group. 37 of 50 participants gave the same incorrect answer as the rest of the group (even though it was very clearly wrong), either because they wanted to be “liked” by the group or because they thought the rest of the group was more informed than they. This effect was dampened by having just one other person in the group agree with the participant.
15. We have trouble separating out traits in a person.
Globally positive or negative reactions on a person (“he’s a nice guy”) affect our judgment of a person’s specific traits (“he’s attractive”). This is called the halo effect, and is particularly noticeable in celebrities; their attractiveness or fame also leads us to believe they’re intelligent, happy, or honest.
16. We’re influenced by very particular types of rewards.
Expected rewards reduce motivation on a task. Surprise rewards increase motivation on the same task. Fixed rewards are less powerful than performance-based rewards, even with creative tasks.
17. Authority can fundamentally change our emotions and behavior.
In the Stanford Prison experiment, participants were split into prisoners and guards and placed into a mock prison. In just six days (of a planned two weeks), the experiment had to be shut down because guards were harassing and abusing prisoners, and prisoners began showing signs of emotional breakdown.
65% of participants knowingly delivered a lethal dose of electricity to a participant (who they later learned was fake) simply because the instructor in the room told them to.
18. Authority can also make us be obedient and do things to other people we could never imagine.
In the famous Milgram experiment, participants were told to administer a shock of increasing strength when a participant in another room gave incorrect answers to a series of questions. About halfway through, the shocks were labeled “danger: severe shock” and a recording was played begging the experimenter to stop the experiment. Yet in 63% of cases, the participant administered the maximum shock, even when the person they thought they were administering a lethal dose of electric shock to another human being.
A recreation of the original Stanford Marshmallow experiment is predictably adorable. Longitudinal studies have shown that students who can resist eating the marshmallow are better behaved and get better grades later in life.
19. Self-control at an early age might be indicative of success later in life.
In the famous Stanford marshmallow experiment, a group of children participants were asked to wait in a room with a table full of marshmallows and cookies. If they wanted, they could have one treat now and the experiment would be over. Otherwise, if they could wait for the experimenter to return in a few minutes, they could have two treats. The children who couldn’t delay their urges—either they asked for the treat right away, or tried to sneakily eat a treat when the experimenter left—had more behavior problems, lower SAT scores, more trouble paying attention in school, and found it difficult to maintain friendships. In fact, a child who could wait 15 minutes scored 210 points higher on the SAT than children who could wait only 30 seconds.
20. People aspire to round number goals.
I tried to make this list 20 bullet points long instead of 19, and you do the same thing when trying to run 2.0 miles instead of 1.9. In Major League Baseball, players were four times as likely to end the season with a 0.300 batting average than 0.299. And when looking at over 4 million SAT scores, students who scored a 1290 were more likely to retake the test than students who scored a 1300—even though admissions offices did not statistically favor one score over the other.
