All posts by Nathan Taylor (praxtime)

About Nathan Taylor (praxtime)

I blog at on tech and the near future. The next 5-20 years. Follow @ntaylor963 for me the person. Follow @praxtime for my blog posts.

Saturday links 30-Mar-2019: status as a service, niche superstars, screentime, Model T as hatkiller

Here’s links and commentary on recent reading.

1. Status as a Service. It’s a nearly 20,000 word (70 page) post by Eugene Wei. Long but great! At least if you’re into understanding tech and social media. Wei explains how social networks can be thought of Status as a Service, acronym StaaS. Some quotes below in bullet point form, mostly taken from Wei’s (shorter) twitter summary:

  • Social networks can be analyzed along three dimensions of competitive advantage: social capital, utility, and entertainment. A whole class of companies are better understood as Status as a Service (StaaS) businesses. This is especially true of social networks.
  • Status as a Service businesses are best analyzed along dimensions of social capital.
  • People are status-seeking monkeys.
  • All StaaS businesses have their own Proof of Work to enforce scarcity of status. Clever lip synch and dance number on TikTok , short quip on Twitter, beautiful photo on Instagram, insightful answer on Quora, etc.
  • Proof of Work is critical in StaaS business because it defines the scoreboard which users compete on. It separates users by their skill at that Proof of Work and creates a relative status ladder.
  • Early Twitter was a lot of dull life updates (my first tweet was “Doing my taxes”). Then Favstar and Favrd came along and provided a global scoreboard for all to see. This launched comedy Twitter. The entire tone of content on the service changed.
  • The greatest social capital explosion in history was the launch of Facebook’s News Feed. It simultaneously increased efficiency of distribution of posts while pitting all of them against each other in a global, unified status competition.
  • Durable status games need mechanisms to reward new and skilled users so that more experienced users don’t dominate. Need mechanisms to clear out “old money” that has stagnated, to make way “new money.”

What stuck with me most was how social networks are driven by status seeking through proof of work. This proof of work flywheel keeps all the status-seeking monkeys going. A useful (if perhaps depressing) lens by which to view social media. And the internet more generally. Full post here. Shorter twitter summary here.

And yes, my chosen proof of work is this blog. 🙂

2. Kobalt music label as niche superstar. Fine post by Mark Mulligan. In particular the point below:

Superstar niches: In the old model, labels would (and often still do) carpet-bomb TV, radio, print and digital with massive campaigns designed to create global, superstar brands. Now, labels can target more precisely and be selective about what channels they use. Kobalt’s business is based around making its roster superstars within their respective niches, finding a tightly-defined audience and the artists they engage with. The traditional superstar model sees an artist like a Beyoncé, Ed Sheeran or a Taylor Swift being a mass media brand with recognition across geographies and demographics. The new superstar can fly under the radar while simultaneously being hugely successful.

This idea of superstar niches harkens back to the now venerable 1000 true fans. But the larger point is “superstar niche” describes not just how music works, but how much of the internet economy works. It’s nearly all superstar niche now. The industrial/mass media age is over.

3. Why Tech Didn’t Stop the New Zealand Attack From Going Viral. This is in reference to the shooter who killed 49 people in two mosques in New Zealand earlier this month. The reason is simple enough. Scale. From Facebook “In the first 24 hours we removed 1.5 million videos of the attack globally, of which over 1.2 million were blocked at upload…” Facebook uses machine learning to tag videos that are similar, but if you take a new video with your phone of the video on the screen, you can fool it. Nothing too surprising. It’s just a useful reminder of the sheer scale of modern social networks. And of course, with end-to-end encryption on messaging services becoming all the rage, monitoring encrypted messaging networks won’t be possible. Which means avoiding blame for sharing. link

4. Lithium-ion battery prices are going down. Good news in a single chart.

source: BloombergNEF

5. Screen time and mental health. There’s a genre of media story best understood by just reading this title: Have Smartphones Destroyed a Generation? This response from Vox is good: Have smartphones really destroyed a generation? We don’t know. I think a reasonable position right now is not to worry too much. There is a rise is suicide among young people, but it’s hard to know what’s causing it. One caveat is Jonathan Haidt argues “We should all be careful about moral panics over tech, but I have been conducting two semi-open-source lit reviews which, so far, support the claims made in video: rising depression, for girls mostly, and some link to social media” And “It’s not so much screen time for all teens, that is bad; it’s specifically social media for girls.” To the extent we should worry, perhaps that’s where to look.

6. Podcasts. If you’re looking for podcasts, below are good episodes from three podcasts I listen to regularly.

  • a16z – Brian Koppelman interviewed by Marc Andreessen and Sonal Chokshi. Koppelman is the showrunner for the TV show Billions. He draws useful parallels between creating new shows and creating new companies.
  • Manifold – Kaiser Kuo interviewed by Steve Hsu and Corey Washington. Kuo hosts the Sinca podcast, on China. It’s good too. But here they discuss Kuo’s background, which includes a stint playing in a rock band in China, as well as working for Baidu.
  • Conversations with Tyler – Emily Wilson interviewed by Tyler Cowen. Wilson wrote an excellent new translation of Homer’s Odyssey. What’s great about this interview is the evident joy Wilson brings to her work. It made me smile.

7. How the Model T drove hats out of fashion. Link (which has old, hat filled crowd photos). Before sunscreen and foldable umbrellas, everyone wore hats to protect themselves from the sun and rain. Quote:

Hats in fashion hit their peak around the 1920s, and set on a downhill path from there. At the same time, the FMC was able to flood the market with their Model Ts and almost completely dominate the American car market. Cars were widespread and used for everyday travel, taking people off of their feet and into cars with vinyl roofs (the steel roof was later applied in the mid-1930s). People suddenly realized that they really didn’t need hats inside their cars. They could go from house to car and from car to work. The car’s roof protected drivers and passengers from all kinds of weather. Why did they need to wear hats anymore?

8. Pee, not chlorine, causes red eyes from swimming pools. You read it here first. link

That’s all for now. Thanks for stopping by.

Sunday 17-Mar-2019: College admissions, Waymo, Wigner’s friend, Zebra flies

Here’s links with commentary from recent reading.

1. Notes on the College admissions scandal. The New York Times has a good explainer. The primary ways people cheated on college admissions where: 1) paying someone else to take SAT tests, and 2) bribing coaches to fake sports resumes. The context here is elite colleges face unrelenting pressure to raise their rankings. In practice this makes them lure in max applicants but only accept a flat cap, because lower acceptance rates equate to greater eliteness. But somewhat surprisingly, this pressure also leads to easy grading, with students rarely flunking out. Why? Because if particular demographics of students flunk out, people will suspect unqualified students must have snuck in. This dynamic makes cheating to get into an elite school work. Because once you sneak in, the university doesn’t want you to flunk out, especially if you’re underqualified. Everyone needs to get an A.

In terms of commentary, Arnold Kling argues the “fundamental scandal is that elite colleges are a positional good for parents. The whole process is built around that.” That is to say, kids weren’t cheating. Parents were cheating so they could brag to their friends. Often not even telling their kids. Thus the people most outraged are rich parents with elite degrees, who have been grinding their kids down playing by the rules. Kevin Carey points out that statistically elite college degrees tend not to matter very much for the wealthy or those with high test scores, as they’ll get fine salaries either way. It makes the biggest difference for the underprivileged, as an entree into upper class culture, networking and jobs. This is true. But downplays the social status allure of the elite college credential, which drives of the problem. Parents cheat for status, not money.

Ross Douthat say many “misunderstand the real function of meritocracy, which is not the facilitation of upward mobility but the legitimation of a ruling class.” That is to say, Harvard’s purpose is not merely to select applicants, but to justify the legitimacy of those who govern. What we’re seeing now in US politics is a breakdown of trust in leadership of our institutions. This breakdown encourages a nihilistic populism.

One proposed response to the scandal is to eliminate test scores on applications. Daniel Friedman walks through why this will makes things worse. The easy way to see this is by asking whether it would be easier or harder for wealthy parents to rig the system if they didn’t have to provide test scores. One less thing to lie about.

Having read a lot of commentary, I suspect nothing will change. The dynamics behind it have been growing for decades, with no end in sight. So if you want to relax bit about getting your kids into an elite college, I’d suggest Bryan Caplan’s book. 🙂

2. Waymo’s self driving blunder. An excellent piece by Timothy B Lee on how Google Waymo is repeating one of silicon valley’s most classic mistakes, Xerox PARC and PCs. In the 1970s, Xerox PARC famously created the Alto computer, with a mouse, copy/paste, windowing, and all the other things that would start the microcomputer revolution. But they failed to commercialize it. Why? Because the Xerox PARC “workstation started at $16,595—more than $45,000 in 2019 dollars.” Instead Apple released the Apple I for $666, the Apple II for $1298. As a side note, I got an Apple II+ as my first computer in high school. It was great! But felt super expensive at the time. Apple did not release a personal computer with all the Alto features until the Lisa in 1983, a flop at $9995. But then the first Mac in 1984 for $2495 was a hit. Lee argues that Google Waymo is making the same mistake by trying to jump directly to fully self driving cars (jumping to the Mac instead of releasing the Apple I). Waymo should go to market with a lessor solution and iterate. If you are into technology and disruption theory, a great read. link

3. Level 2 self driving is pretty silly. On a related note, the levels getting to self driving cars are defined as: level 0 – manual, level 1 – cruise control, level 2 – partial automation (hands on, may have to take back control), level 3 – eyes off (car can stop if needs help), level 4 – full (doesn’t need help), level 5 – no steering wheel. This is ok except for level 2, which is silly. Why? Because with level 2 you’re supposed to let the car drive but still pay attention and keep your hands on the wheel at all times. But human natures being what it is, when it was tried in 2012 by Google “In-car cameras showed users napping, putting on makeup and fiddling with their phones.” Pretty much. Level 2 shouldn’t exist.

4. Survey of political divide in US. Unsurprising voter survey results, but still makes me sad:

How about: “Do you ever think: ‘we’d be better off as a country if large numbers of the opposing party in the public today just died’?”

Some 20 percent of Democrats (that translates to 12.6 million voters) and 16 percent of Republicans (or 7.9 million voters) do think on occasion that the country would be better off if large numbers of the opposition died.

We’re not finished: “What if the opposing party wins the 2020 presidential election. How much do you feel violence would be justified then?” 18.3 percent of Democrats and 13.8 percent of Republicans said violence would be justified on a scale ranging from “a little” to “a lot.”

Since we’re doing voter surveys. Maybe part of this is the demographics of age. See below. If age is the driver, things will get worse. From America’s Defining Divide Isn’t Left vs. Right. It’s Old vs. Young.

sournce: Huffington Post

5. New paper on Wigner’s friend quantum mechanics. Ok. If you’re not into foundations of quantum mechanics, skip this and jump to #6 below about Zebra stripes and flies! But if curious, plow ahead.

Wigner’s friend is a thought experiment in quantum mechanics. Observer a measures the outcome of a quantum experiment, which collapses the wave function, so observer a records a particular result. Now. If there’s an observer b who watches observer a, what observer b sees is the quantum superposition from the experiment is contagious, in the sense that observer b sees observer a is still in superposition. At least until after observer b does their measurement. This is related to the famous Schrödinger’s cat thought experiment.

At a technical level the contradiction between observer a and b manifests as a violation of the Bell inequality. The experiment in the paper did an extended version of Wigner’s friend, with an outside observer c.

source: Experimental rejection of observer-independence in the quantum world

From the abstract “In a state-of-the-art 6-photon experiment, we here realise this extended Wigner’s friend scenario, experimentally violating the associated Bell-type inequality by 5 standard deviations. This result lends considerable strength to interpretations of quantum theory already set in an observer-dependent framework and demands for revision of those which are not.” The last sentence says the observer is part of the quantum system. So the naive Copenhagen interpretation of wave collapse is no good. This paper won’t change any minds. But to me is another small push towards many worlds being reasonable. Your mileage may vary. link to paper

6. Zebra stripes discourage biting flies. One longstanding hypothesis (among many) for why Zebras have stripes is to discourage flies. A new paper supports this idea. Apparently flies can see the Zebras, but the stripe pattern makes them struggle to land. Excellent write up by Ed Yong. Plus lovely pic below, dressing up a horse with stripes to test the theory.

source: Tim Caro at US Davis

And that’s all for now. Thanks for reading.

Sunday links 10-Mar-2019: Facebook privacy, DNA shell casings, workism, our baby shark future

Here’s links/commentary on interesting things I’ve read recently on the internet.

1. Zuckerberg’s social networking privacy statement. Last week Mark Zuckerberg wrote A Privacy-Focused Vision for Social Networking. Much commentary ensued. Below is my summary, where anything in quotes “” is Zuckerberg:

  • Growth in public social media (think Facebook proper) is dead. Private messaging is the new growth. “private messaging, ephemeral stories, and small groups are by far the fastest growing areas of online communication”
  • Facebook proper won’t change. It will still make money from ads, and people using it will have as much (or as little) privacy as before. Facebook will continue to print money. “Public social networks will continue to be very important in people’s lives”
  • Facebook’s two messaging apps, WhatsApp and Messenger, are the new growth. They will have end-to-end encryption, which means no ads, at least in the messages themselves. “I believe the future of communication will increasingly shift to private, encrypted services.” “In a few years, I expect future versions of Messenger and WhatsApp to become the main ways people communicate on the Facebook network.”
  • End-to-end encryption means hate crime, pedophilia, drugs, criminality, etc are no longer Facebook’s problem. Facebook so happy. Whew! But why are they off the hook? Because the messages themselves will only be unencrypted on the end users phones. Nowhere in between. Facebook can’t read them, even with a police warrant. That said, Facebook will still cooperate with police, meaning for example they could still pass along the metadata (eg timestamps of when a message was sent). “We are working to improve our ability to identify and stop bad actors across our apps by detecting patterns of activity or through other means, even when we can’t see the content of the messages, and we will continue to invest in this work.”
  • Encrypted messaging is ephemeral, and will get automatically deleted after some time period. “set a new standard for private communication platforms — where content automatically expires or is archived over time…. For example, messages could be deleted after a month or a year by default.”

So what was not said? Answer: money! How will messaging make money without ads? Facebook itself prints money from ads. Instead, the default money making model for messaging is Line in Japan, and WeChat in China. These started as messaging apps but are now complete platforms with payments and mini-apps running inside them. So unless we hear otherwise, that’s the best guess on where Facebook wants to go. While that will be harder to pull off in the US, in many parts of the world Facebook is the internet, so their chances outside the US are higher than within.

Furthermore, as Antonio García Martínez showed in his excellent book Chaos Monkeys, Zuckerberg has always disdained grubbing money from ads. Yuck. Dirty. So dirty. Now Zuckerberg can finally funnel his Facebook ad money into the platform he’s always longed to become. The Microsoft of the social networking age. Skimming money off every platform payment transaction, as every good software monopoly’s heart truly desires.

I’d say the best write up on the topic is from Ben Thompson. Thompson points out that with Facebook proper still printing money, they can afford to subsidize their WhatsApp and Messenger play. This flips the privacy narrative. Making privacy a winner for Facebook. Short term it’s for PR. Long term it’s for (they hope) total social networking platform domination. Quote:

Why can Facebook deliver most of the value? Because they are still Facebook! They still have the core Facebook app, Instagram, ‘Like’-buttons scattered across the web — none of that is going away with this announcement. They can very much afford a privacy-centric messaging offering in a way that any would-be challenger could not. Privacy, it turns out, is a competitive advantage for Facebook, not the cudgel the company’s critics hoped it might be….

Relatedly, and most importantly, there needs to be much more appreciation for the anti-competitive trade-offs inherent in an absolutist approach to privacy. Facebook is doing what its fiercest critics supposedly want, and enhancing its competitive position as a result.

2. DNA tests on envelopes and shell casings. If you told me it was possible to track down criminals by picking up DNA from licked envelopes, I’d have paused, but nodded. And perhaps yes, even 100 year old envelopes, sure. Story here. But even though I know it’s a fast moving field, how about this: a lab is able to get “usable DNA off of about 30 percent of the casings it analyzes.” Those are shell casings. From bullets. So the criminal touches the bullets to load the gun, leaving trace DNA. Then shoots the gun, which expels spent shell casings at the crime scene. And 30% of those casings retain enough DNA to identify someone. Wow! Story is here.

3. Workism as substitute religion. Good piece from Derek Thompson on how work has now become a source of life meaning, as other social institutions like religion have fallen by the wayside. In particular this paragraph jumped out:

In 1980, the highest-earning men actually worked fewer hours per week than middle-class and low-income men, according to a survey by the Minneapolis Fed. But that’s changed. By 2005, the richest 10 percent of married men had the longest average workweek. In that same time, college-educated men reduced their leisure time more than any other group. Today, it is fair to say that elite American men have transformed themselves into the world’s premier workaholics, toiling longer hours than both poorer men in the U.S. and rich men in similarly rich countries.

As I mentioned on twitter, I was in high school in 1980. And yet the past is a foreign country. Even my own. I find it incredibly jarring and hard to believe the world used to be like that. A world where the rich goofed off, instead of flaunting their status and moral virtue through their elite workaholic hours. link to the piece

4. Small science teams are better, or…maybe not. Bloomberg: Small Groups and Individuals Are More Innovative Than Large Teams. And from Ed Yong at the Atlantic: Small Teams of Scientists Have Fresher Ideas. Both are good write ups of the paper. But then Sean Carroll pointed out “The causal factor doesn’t seem to be size of the collaboration, it’s receiving funding from top agencies!” And “Disruptive science is likely to seem speculative or wrongheaded at first, and agencies don’t like to go for that.” And hence “funding some long-shot ideas is more important than worrying about team size. We need a diversity of funders and approaches, and a willingness to let people chase occasional longshots.” Sean Carroll is right. A good catch, and good correction to the narrative. Diversity of science funding matters more, or at least as much, as team size.

5. Lithium Ion battery prices. The chart below makes me happy.

source: BloombergNEF

6. Our baby shark music future. People use voice interface to select music more than ever before. Hey Siri, play xxxx. Alexa, play yyyyy. But what are the second order effects on music songwriters and producers? Let’s see:

But Grainge, who has attracted many of the world’s most popular artists into his fold, has a little tip for songwriters and musicians who want people to stream their music through smart speakers. He uses the example of the timeless hit, perennially popular among the global toddler population, “How Much Is That Doggy In The Window?” to explain that the title of a song must be front and center in a song’s lyrics.

“If you’ve got something that is a brand, is a soundtrack, is a song where the title is in the chorus and the melodies, we’re seeing really explosive data and activity,” he said. “That helps us in the creative process because it enables us, with the data and with consumption, to use the technology to say to the talent, you need to have something as basic as the song title […] in the chorus.”

In other words, baby shark. We’re talking baby shark.

I can’t recall where I saw it, but someone predicted a dance club in 2040 will play Baby Shark late at night, and all the cool kids will go wild as they recall their misspent kindergarten youth. Asking Alexa to play baby shark over and over and over. Truth.

And that’s all this time round. Thanks for reading.

Saturday links 16-Feb-2019: Campus disinvitations, reducing suspensions backfires, college loans, Fortnite is all about the social

1. Campus disinvitations are way down. People have been freaking out about disinviting college speakers (in particular conservatives) for the past two years. The chart below is great, because it means now we can all go fight about something else instead. link

source: Niskanen Center

2. Stopping school suspensions backfires. New results from a large scale test of “restorative practices”, which means greatly reducing school suspensions. With a goal of helping minority students. Unfortunately it backfired. The policy disproportionately hurt African American students. The common sense reason is simple. Disruptive kids make it hard to teach. And if you don’t suspend disruptive kids, their behavior hurts at-risk kids the most. Kevin Drum is a supporter. So he says we should “take the results seriously enough to try to figure out how programs like this can be improved. We should give up on them only if we do that and they continue to fail.” If you want a harsher take, go here.

3. College student loan burden. Let me quote from the piece:

For many of us, student debt means delaying — if not entirely forgoing — homeownership, marriage, and parenthood. This new form of social stratification — between those who have student debt, and those who do not — will have ramifications for generations to come.

Millennials didn’t adopt new ideas about college. We inherited and internalized the old ones. But the cost of those ideas has risen exponentially in the past few decades. In 1983, the average full-time student borrowed $746 ($1,881 in 2018 dollars) per year. The most recent statistics from the College Board indicate that in 2018, the average annual undergraduate loan is now $4,510, while the average graduate loan hit $17,990. In 2016–2017, the average borrower left college with $37,172 in loans.

The above is an excellent point. The old norm was go to the best (most expensive) college you got into, and don’t worry about choosing a major with an eye on salary. Any degree automatically pushed you into the upper middle class. Student loans are no big deal. This was true. But isn’t any more. Yet the enticing norm lives on, zombie-like. Along those lines, College May Not Be Worth It Anymore:

Since 2000, the growth in the wage gap between high school and college graduates has slowed to a halt; 25 percent of college graduates now earn no more than does the average high school graduate. Part of the reason is oversupply. Technology increased the demand for educated workers, but that demand has been consistently outpaced by the number of people — urged on by everyone from teachers to presidents — prepared to meet it.

Why do employers demand a degree for jobs that don’t require them? Because they can.

4. Trilemma of speech moderation. Tyler Cowen argues regulation of speech on the internet faces an unsatisfactory trilemma: “When it comes to private platforms and speech regulation, you can choose two of three: scalability, effectiveness and consistency. You cannot have all three. Furthermore, this trilemma suggests that we — whether as users, citizens or indeed managers of the platforms themselves — won’t ever be happy with how speech is regulated on the internet.” Kevin Drum has a useful illustration, from his experience moderating blog comments:

  • If the blog was small, I could easily moderate comments and do it consistently.
  • If I was willing to spend lots of time on moderation, I could manage a large blog with consistent comment policies.
  • If I decided not to worry about consistency, I could manage a large blog without putting a lot of time into comment moderation.

5. Planet Labs has 200 Earth observation satellites. The smallest satellites are the size of a shoebox. So it’s all about volume. They have the largest satellite constellation in history, and can image the entire Earth once a day. Nothing special here, just thought it was cool. link

6. Three notes on China. First: China faces a housing glut, leaving 65 million empty apartments. Second: Why did Americans ignore the Unkrainian famine caused by Stalin in 1932-33 and do nothing? For the same reason we’re doing nothing about Uighur concentration camps now. We gots way too many other things to worry about. Third: “Xi Study Strong Nation” is a mobile phone app that gives you points for reading/watching Communist Party content. In particular I found this striking:

When I went home over winter break, I discovered that my mom, who generally didn’t use her mobile so much, was on it every single night, and even would not sleep until late into the night (while she generally was asleep by 10PM). I found this really strange, and only after I asked about it did I learn that the education committee at the subdistrict level had ordered teachers at all schools to download an app called “Xi Study Strong Nation,” and to earn points by fulfilling various tasks every day.

7. Fortnite is all about the social. Fortnite is a massive gaming hit. But you knew that. What I didn’t know was how important social was to the game’s allure. Quote: “Fortnite intended to merge specific shooter dynamics with the sandbox nature of Minecraft so that players could define their own style of play.” And: “Fortnite has become a daily social square – a digital mall or virtual afterschool meetup that spans neighborhoods, cities, countries and continents. This role is powered by Fortnite’s free availability, robust voice chat, cross-platform functionality, and collaborative gameplay. Accordingly, examples abound of kids, adults and families simply hanging out or catching up on Fortnite while they play.” Interesting longform if you’re curious. link

8. NASA’s Mars Opportunity rover is dead. Opportunity had a planned 90 day mission, but lasted 14 years on Mars. Main discoveries were gypsum and hematite, signs of Mars’ far wetter past. My favorite image is below.

Opportunity's northward view of tracks
source: NASA

And that’s all for now. Thanks for reading.

Saturday links 9-Feb-2019: climate and epidemics, ur DNA r belong to us, Alzheimer’s gum disease, Aphantasia

Here’s a list of links/commentary on what I found interesting this week.

1. European arrival in Americas may have caused enough death to cool climate. When Europeans discovered America in 1492, they brought along smallpox and measles. The resulting mass epidemics killed roughly 90% of the population, dropping from 60 to 5 million. Previously cultivated land, roughly the size of France, became reforested. The new plant growth lowered CO2 levels, cooling the climate. In particular causing the temperature downturn in the 1590s. This idea has been around since 2003. A new paper adds to the argument. It’s a disturbing and headline worthy idea.

What’s certain is the mass American dying and resulting reforestation happened. And on the margin this lowered CO2 levels. Fine. What’s less certain is how big this effect was compared to everything else. The paper argues noticeably big. It’s a cool paper. The BBC news story is here. And if interested, read a critique here.

2. All ur DNA r belong to us. In April last year the police solved a murder cold case by searching GEDmatch, a public DNA database. They only found 3rd and 4th cousins, but that was enough to narrow the search to 1000 people. And from there they eventually solved the murder. GEDmatch has about a million people in it. It’s open source, in the sense that people who upload their DNA have agreed to allow anyone else to use it.

What’s new is Family Tree DNA, a consumer DNA testing company, with 2 million DNA profiles on file, has been allowing the FBI use their DNA database to find criminals. And they did this without informing their users until Buzzfeed broke the story. People are rightfully upset. This is a breach of privacy and terms of service. That said, as we saw in the GEDmatch case, the public seems generally supportive of using DNA to identify criminals. So I don’t think things won’t go back to how they were.

People believe their own DNA is private. But even though only 3% of your DNA is shared by 2nd cousins, that’s enough to identify your pedigree. And even 3rd or 4th cousins are enough to find you if the FBI is motivated enough. Given shared ancestry, DNA information is far more of public good than most naively assume. All ur DNA r belong to us.

Of course this means foreign governments can use DNA to track down relatives of dissidents, and tabloids can use DNA to snoop on celebrities. So which justifications for a search are legally allowed matters. But the larger point is we’re not heading toward a world where anyone can be tracked by their DNA. We live in such a world already.

3. Jay Rosen is worried journalism for the 2020 presidential election will be like 2016. Here’s Rosen: “it’s become clearer and clearer to me that, without intervention, coverage of the 2020 campaign is likely to be a disaster for everyone except Trump and his core voters, who want to watch it all burn anyway.” It’s hard to know what’ll happen in 2020 given how much is in flux now. But agree the dynamics around journalism and Trump are unchanged from 2016. Rosen says journalists should spend less time on horse race journalism, pumping out articles on who’s ahead and who’s behind, and spend more time focusing on issues voters care about. We’ll see. Twitter thread is here, and interview is here.

4. Gum Disease and Alzheimer’s. The more sensational version of the news was We may finally know what causes Alzheimer’s – and how to stop it. (Answer: bacteria behind gum disease). This was then countered by No, we don’t know that gum disease causes Alzheimer’s. With the subtitle “A new study suggests a link between oral bacteria and Alzheimer’s, but it’s far from proven.” Which is the correct take. As for me, am I flossing more often now? Yes. Yes I am.

5. Cowen on Pinker on the Enlightenment. Steven Pinker’s book Enlightenment Now is, in my view, excellent. But the book has been attacked for being naive. Pinker responded. Then Tyler Cowen got to the heart of the matter when he noted Pinker reserves the name Enlightenment only for those aspects he likes. Not enough grey. In particular I like this bit: “I am very much an admirer of Pinker and his work, and I consider myself an optimist, especially across longer time frames.  But what is sometimes called progress does also have a dark side, and we will do better fighting that dark side if we are clearer — in our own minds and with each other — on how things have run to date.”

6. Aphantasia, I have it. Aphantasia is an inability to visualize images in your mind’s eye. So imagine an apple. Close your eyes and see it. If you have aphantasia, you can’t do that. There are matters of degree. Some people can vividly imagine pictures. Others less clearly. But if you can’t to it at all you have aphantasia. It was a shock a few years back to read about it because, having aphantasia my whole life, I had always assumed everyone was like me, and “mind’s eye” was just a metaphor. If you’re curious, this account of how shocking it was to find out about aphantasia was similar to my own. Oh. I see. I see! So that’s why book authors waste so much time on elaborate visual descriptions of the countryside. Who knew? Remembering faces is a bit harder for me than most. I can’t visualize, but if I make an effort, or just see someone often that’s fine. GPS is a godsend. My kids sometimes complain I use it for places we’ve been many times, but I love it. And my case is relatively mild. If I focus hard I can sometimes sort of imagine seeing a picture. But it’s rarely worth the bother.

I suppose this is neither here nor there. But perhaps I wanted to share because it made me realize we can go through life assuming too much. Leading separate lives, we but dimly perceive our own mind’s eye. Best to reach out and share and learn from others, letting their outside views make us see more fully a complicated, if occasionally beautiful, world.

That’s all. Thanks for reading!

Apple revoking Facebook’s certificate was a strategic err which may come back to haunt the tech industry

image credit: Peter Sellers in Dr. Strangelove, Or: How I Learned To Stop Worrying And Love The Bomb

Let me explain why I think Apple’s decision to revoke Facebook’s developer certificate was a strategic err. Not just for Apple, but for tech in general.

First the background. For an app to run on an iPhone, it has to be signed by an Apple certificate. Apple provides app developers with both an enterprise certificate, with extra privileges for internal iOS app development, and a general certificate, for the app store. Facebook broke Apple’s rule in using the enterprise certificate to sign a Facebook VPN app (giving it more privileges) for an app going to the App Store for regular customers. This app could then collect information on which other apps were installed on user’s phones, and what data traffic was sent. This was a violation of the rules, a privacy breach, and was done knowingly. Make no mistake. Facebook was bad bad bad.

In response Apple revoked Facebook’s enterprise certificate, which in turn shut down all iOS development and testing inside Facebook. It’s the nuclear option. Then it turned out Google had done something not quite as bad but similar, so (to be fair), Apple shut down Google as well. And now it turns out there are other companies which violated the terms too. After a few days, Apple re-instated the certs for everyone, so now things are getting back to normal.

Here’s my point. As Ben Thompson noted: “I wouldn’t be surprised if, in the long run, the company comes to regret this move.” This is exactly correct. I believe Apple made a strategic err they will regret. Even though it’s popular right now. For example John Gruber “Give me a break. Everybody knows that Apple’s enterprise certificate program does not in any way permit distribution of this kind of software, which wouldn’t be allowed in the App Store. But that’s exactly how Facebook was using it. There are pros and cons to Apple’s iron-clad control over native apps on iOS. This incident with Facebook is one of the pros.”

Both can be true: 1) Facebook gets away with too much and absolutely deserves severe punishment, and 2) Apple going nuclear was a mistake. No matter how deliciously satisfying. No matter if the rules said they could revoke Facebook’s certificate. Rules should be enforced with an eye toward precedent, with an assumption bad actors will eventually latch on to those same precedents.

Here’s an analogy. Suppose back in the day when Netscape had just invented the web browser and was riding high, and Microsoft was the evil empire (think circa 1997), Microsoft shut down all Netscape windows development for a rules violation. It’s similar to what Apple (good company) did to punish Facebook (bad company). But in this hypothetical the villains are swapped. It’s Microsoft (bad company) punishing Netscape (good company). While this couldn’t happen back then, with cloud centralization it can now. As Apple just demonstrated.

The cold war between the Soviet Union and the US was very fraught, but the one thing both sides got right was showing restraint with nuclear weapons, either directly or in proxy wars. The cold war bomb threat was invoked all the time. Fine. But leaders showed restraint. Knowing if one side (the good side) used them, even if justified, then the other side (the bad side) would run with that precedent and also use them. Apple could have played the brinkmanship game, publicly threatening to revoke the cert in a Bay of Pigs style confrontation with Facebook, and gotten much the same result, but without actually firing nukes.

It’s unclear exactly how and when this may come back to bite. But as a hypothetical imagine a startup using modern centralized platforms such as Amazon AWS, Microsoft github, Google maps, etc etc. And Amazon, Microsoft, Google put in place complex developer rules on what’s not allowed. So it’s hard to comply. Then if that startup later competes against one of those companies, the platform owner has the temptation to stop that company dead by disallowing development on their platform on the pretext of a rules violation. Threatening may be enough. And I’m not saying Amazon, Microsoft or Google in particular will do this. Rather that in the cloud era all centralized cloud platforms face this temptation.

Apple set a precedent. Nukes are now allowed. Even if the letter of the law says you can do it, restraint is often a wiser choice. There will be increased temptation to create platform rules which are easy to violate. Laying a trap. Just in case. Then, if the mood strikes, bombs away.

image credit: Major Kong (Slim Pickens) riding a nuke in Dr. Strangelove

Wednesday links 30-Jan-2019: Cognition models, Marx and virtue signalling, slime, iterated embryo selection, screentime

1. Scott Alexander on cognition models. Reading the book On Intelligence by Jeff Hawkins when it came out in 2004 was a revelation to me. That’s because Hawkins provided an overarching theory for how cognition works, with his memory/prediction model. From page 99: “The recalled memory is compared with the sensory input stream. It both ‘fills in’ the current input and predicts what will be seen next. By comparing the actual sensory input with recalled memory, the animal not only understands where it is but can see into the future.” Reading this in 2019, it seems unsurprising. And the framework is not original to Hawkins. But it was the first time I’d encountered a well written, convincing and unified framework for human cognition. Mind is prediction. At a base level, sense data constantly flows in, and predicted future sense data constantly flows out. At a higher level, mind is prediction on how potential possible actions can affect the world. An overarching framework is critical for science, because it stops you from drowning in details. It’s like completing the outside edge of a puzzle. Once that’s done, the frame guides you on where to place all the other pieces.

So I really liked this post by Scott Alexander, where he created his “Grand Unified Chart” below. It 100% matched my priors 🙂

Scott Alexander explains:

All of these are examples of interpreting the world through a combination of pre-existing ideas what the world should be like (first column), plus actually experiencing the world (last column). 

All of these domains share an idea that the interaction between facts and theories is bidirectional. Your facts may eventually determine what theory you have. But your theory also determines what facts you see and notice. Nor do contradictory facts immediately change a theory. The process of theory change is complicated, fiercely resisted by hard-to-describe factors, and based on some sort of idea of global tension that can’t be directly reduced to any specific contradiction.

Why do all of these areas share this same structure? I think because it’s built into basic algorithms that the brain uses for almost everything (see the Psychology and Neuroscience links above). And that in turn is because it’s just factually the most effective way to do epistemology, a little like asking “why does so much cryptography use prime numbers”.

Last point. The full title of Hawkins’ book is On Intelligence: How a New Understanding of the Brain Will Lead to the Creation of Truly Intelligent Machines. That is to say, we could add a new row to Alexander’s table for artificial intelligence. It slots in perfectly. (And yes, I’m aware some would consider today’s AI a subfield of Bayesian stats). We’re still a long way from general AI. But the point here is the framework to judge AI progress already exists. If mind is prediction from sensory streams, generalized mind (AGI) requires an ability to go beyond narrowly specific domains.

2. Tanner Greer on virtue signalling. Greer first reminds us how Marx believed economic structures and institutions drove class consciousness, which in turn determined people’s ideas. Ideas are nothing more than an expression of class interest. Which means, as Greer says:

A straight line can be drawn from statements like these and the great terrors imposed upon captive populations by Marxist dictators like Stalin, Mao, and Pol Pot. Men like Stalin did not oppose freedom of expression simply because it threatened their personal power. They opposed freedom of expression because they earnestly believed it was a pointless exercise. There is no point in debating with representatives of the bourgeoisie when the ideas of the bourgeoisie were not pliable to debate. If the opposing stance was not the product of reason, it could not be undone by reason.

Then to Greer’s analogy to signalling theory:

The central proposition of signalling theory is that the vast majority of arguments made in the public sphere are not made in good faith. An outraged tweet is not written to express genuine emotion, but to signal solidarity with the ‘right’ side. A verbose blog post is not written to persuade its readers of its argument, but of the cleverness of it author. A well circulated censure of some racist act is not written to convict the racist, but to display the Wokeness of the censor. The connecting string in all of these cases is that your arguments are less about your ideas than shaping other people’s perceptions of you. Whether you believe you are writing primarily to shape other people’s perceptions of you is immaterial. As with the Marxist theorists, signalling theorists are happy to conclude that signalling does not need to be a fully conscious process. In place of a class consciousness imposed by the material circumstances of an individual’s social status, signalling theorists trace the origins of self-interested arguments to mental social-status ‘modules’ imposed by the material circumstances of an individual’s evolutionary heritage.


Civilized discourse depends very much on both parties in duel of words recognizing the self-serving intentions that lie behind the other party’s speech… and then deciding not to mention them. This is Reason’s Pact: the bare minimum required for fruitful rational discourse to take place. We underestimate how many institutions in our society depend on us maintaining this illusion. Jettison the ideal of ‘reason’ as a governing principle, and all you have left are words as war. 

One quibble. Greer notes “Robin Hanson, doyen of modern signalling studies, fights a perpetual battle to convince his readers that none of the ulterior motives he ascribes to human sociality writ large are behind his twitter polls.” Fair point on the irony of Hanson being attacked for impure motives and signalling. But I would add in Hanson’s defense, I’ve not seen Hanson use that style of argument against his accusers. See this example. In fact the central argument of Hanson’s book on signalling is social policy could be improved if we take signalling into account. That is to say, I read Hanson as a nerdish theorist who writes about signalling with the hope of finding ways to work around it, leading to better social policies. Though perhaps that doesn’t matter. It’s being weaponized regardless. Quibble aside, an outstanding post. Worth reading in full.

3. No One Is Prepared for Hagfish Slime. It’s yucky. Has a great picture of a slime-covered Prius. And is written by the excellent Ed Yong. Recommended.

4. Iterative embryo selection. CRISPR gene editing is not reliable enough to do the many 100s or 1000s of edits it would take to dramatically impact complex traits. But one way around this problem is through iterative embryo selection: “conducing multiple generations of embryo selection in a petri dish by exploiting gametogenesis or stem cells.” So you use IVF, get an embryo, do gene edits, then replicate those embryos to get a new set of cells, select those without side effects, and repeat. The selection step allows you to weed out bad edits. This is not completely workable now. And will be applied to livestock first. But the technique will accelerate when CRISPR can do large numbers of complex edits. This was a new idea to me. link

5. Screentime worries overblown. New paper with N=355,358 argues that screen time worries are overblown, if hard to pin down. The effect of wearing glasses is bigger. Paper here. News article here.

6. On desiring your enemies to be evil. I was going to link to some Covington stories, but decided not to. Here’s C.S. Lewis from his 1952 book instead:

Suppose one reads a story of filthy atrocities in the paper. Then suppose that something turns up suggesting that the story might not be quite true, or not quite so bad as it was made out. Is one’s first feeling, ‘Thank God, even they aren’t quite so bad as that,’ or is it a feeling of disappointment, and even a determination to cling to the first story for the sheer pleasure of thinking your enemies are as bad as possible? If it is the second then it is, I am afraid, the first step in a process which, if followed to the end, will make us into devils. You see, one is beginning to wish that black was a little blacker. If we give that wish its head, later on we shall wish to see grey as black, and then to see white itself as black. Finally we shall insist on seeing everything — God and our friends and ourselves included — as bad, and not be able to stop doing it: we shall be fixed for ever in a universe of pure hatred.

And that’s all for this post. Thanks for your time!

Saturday links 19-Jan-2019: Gravitization of quantum mechanics, Revolt of the public, income sharing for tuition, Big 5 personality quiz

1. Roger Penrose on the gravitization of quantum mechanics. Physicist Roger Penrose was on Sean Carroll’s podcast (recommended). I want to highlight a particular point Penrose made at the 1:16 mark. As you may know, the theories of gravity and quantum mechanics are not reconciled. The problem is general relativity (gravity) treats space as continuous, not quantized. While quantum theory makes all interaction quantized.

One approach is to reformulate gravity to match quantum mechanics. This approach is called quantizing gravity. String theory is an example. Testing String theory (unless an indirect method is discovered) requires getting down to the Planck scale (very tiny, meaning super energetic particles). As Penrose put it on the podcast, it would require a particle accelerator the size of the solar system. Indeed.

Penrose argues we should go the other way around. Rather than reformulate gravity to match quantum mechanics, reformulate quantum mechanics to match gravity. This approach is call the gravitization of quantum mechanics (see paper). Penrose said testing this is just about within experimental reach now, using bose einstein condensates.

So what’s the point here? Funding, funding, funding! The world’s largest particle accelerator is the Large Hadron Collider (LHC) at CERN. CERN has just announced a proposal for an even larger collider, at the cost between €9 and €21 billion. Sabine Hossenfelder argues “it would make more sense to put particle physics on a pause and reconsider it in, say, 20 years to see whether the situation has changed, either because new technologies have become available or because more concrete predictions for new physics have been made.” Exactly so. I have no idea if Penrose’s ideas on the gravitization of quantum mechanics will pan out, but at least we can test them for way less than €9 billion.

2. A good Bandersnatch move review. Bandersnatch is a Netflix movie (technically an episode of Black Mirror) that lets you interactively make choices, resulting in different endings, depending on what you pick. I enjoyed this review more than the movie itself, since it clarified why video games are better suited for interactivity than movies. Here’s one bit:

The reason that letting the audience choose its own story keeps failing when the entertainment industry tries it is that it’s a bad idea. It’s the author’s job to write the story. They can then choose a way to convey that story that gives the reader freedom in how they experience it. But if the story itself is merely a loose collection of different options, each in a different genre and with a completely different tone, then what they’ve created isn’t a coherent work, but a self-indulgent mess—like “Bandersnatch”, in fact.

3. Martin Gurri’s book The Revolt of the Public. I’m about halfway through and it’s excellent (link). Gurri explains why internet publishing has led to nihilistic worldwide populist revolts. I’ll write more once I’m done. In the meantime, here’s a paragraph that captures the core idea:

The information balance of power has changed, of course. A generation ago, the public could exist only as a passive audience Information was dispensed on the industrial model: top down and one to many. That was the great age of the daily newspaper and famous anchormen on the model of Walter Cronkite. The advent of digital platforms, in a sense, created the public. People from nowhere, free of institutional entanglements, pushed the elites out of the strategic heights of the information sphere. Almost immediately, great institutions in every domain of human activity began to bleed authority a process that, as we have seen, in now approaching the terminal stage for many of them. That is my thesis for the revolt of the public.

4. CRISPR gene editing can’t make super-smart babies. This is a pretty basic point, but worth making since gene editing has been in the news. CRISPR/Cas9 gene editing is err prone, resulting in “off target” effects. So editing a single gene works only because you can make multiple attempts, starting over until you get it right. But complex traits like height or IQ are highly polygenic, depending on 1000s of genes. With each gene of very small effect. And editing 1000s of genes isn’t practical. At least for now, it won’t work. link

5. Telecoms companies sell user data to third parties. While many are upset about Facebook privacy, telecom companies are worse. If you care about this kind of thing, this article title says it all: I Gave a Bounty Hunter $300. Then He Located Our Phone.

6. NCAA should allow college athletes to cash in on endorsements. Financial exploitation of college athletes in big name sports is a tricky topic. But this seems like a good way to finesse the problem. Let Nike pay them. link

7. Income sharing to pay college tuition. The way income sharing works is you pay no college tuition up front. Then after you graduate, if you make above a threshold, you pay a fraction of your income for several years. Quote:

Lambda School gives students three payment options. They can either pay $20,000 upfront, pay $10,000 upfront and forgo 17 percent of their salary for a year (with the maximum payment capped at $15,000) or pay zero dollars upfront and forgo 17 percent of their salary for two years (with the maximum payment capped at $30,000). Allred estimates that more than 90 percent of students have opted to pay via an income share agreement.

It’s easy to dismiss this as no big deal. Why would paying before versus after graduation matter? And the New York Times piece is full of hype. But if you believe incentives are critical to making an organization perform, you can see why colleges getting paid depending on someone’s post college job success could be transformational. Incentives are critical. For example see Austen Allred talking about how the school is helping disadvantaged kids do better at job search. On the flip side, I think there’s also an incentive not to get students enrolled in programs where there’s no chance they’ll graduate. That seems harsh, but the current incentives are to enroll and accumulate debt, which is worse. Better to align all students to the programs that will help their careers the most, rather than over promise. Aligned incentives are great!

8. Personality quiz based on big 5 traits. Personality quizzes are normally junk science. And yes, that includes the popular Myers-Briggs test. Psychology breaks personality into the “big 5” traits: extraversion, agreeableness, conscientiousness, negative emotionality and openness to experience. 538 published an online quiz. It’s quick and fun to take. I got no surprises back. I’m more disagreeable and introverted than most. But very open to new experience. A good fit for writing about cool things on the internet, if I may say so. 🙂 Here’s the 538 test if you want to give it a try: link

And that’s all for this week. Thank you for reading!

Sunday links 13-Jan-2019: Apple services, China climate change, superstar cities, falling murder rates

Here’s my weekly list of links with commentary, the first for 2019.

1. Apple’s new services strategy. My new post from earlier today. I argue Apple’s shift to put their movies/TV shows/music services on other hardware platforms was a decision to commoditize the complement. link

2. Climate Change is about China. Noah Smith: “This leads to a painful but inescapable truth — no matter how much they spend, no matter how dramatically they change their societies, the U.S. and Europe won’t be able to put much of a dent in global warming on their own. Yes, the U.S. should ban coal power, tax carbon heavily and spend lots of money on building green energy infrastructure. But without a huge change in China, none of that will matter — the battle against climate change will be lost.”

In particular I liked these two graphs:

Smith’s post is here. Plus also see his related post Saving the Planet Doesn’t Mean Killing Economic Growth.

3. Nuclear power is bad, but better than the alternatives. Many people hate nuclear power. I think the best argument for nuclear power is not that it’s safe. It’s not. But rather all kinds of power generation have risks and problems. Arguing for nuclear power is arguing we don’t live in a perfect world. Solar is great, not magic. Here’s an interview with pro-nuclear activist Michael Shellenberger. One bit:

Look at France and Germany, the two biggest economies in Europe. France has 92% of its electricity from zero emission sources – Germany has 46%. And yet France pays half as much for electricity as Germany because it depends heavily on nuclear and Germany is phasing it out. Germany saw its electricity prices rise 50% over the last 15 years. France’s prices has stayed stable. So introducing huge amounts of solar and wind onto the grid makes electricity expensive; we see it all over the world. The reason has to do with trying to manage all of that unreliability.

4. Economics of superstar cities. Tech naturally clusters, creating superstar cities. Which has been going on since the 1980s. This has coincided with tightening house building restrictions, which has led to decreased mobility into growing cities, at least compared to the past. This is a bit of a hobby horse for me. So am glad to see more people writing about it. Here’s some recent articles which I thought worth reading, if you want to click through:

5. US murder rate decline. The chart below speaks for itself.

New York Time article with chart is here. Plus a nice tweetstorm by Noah Smith on same topic.

6. Some good longreads. I’d like to link to some longer articles which I enjoyed reading over the holidays. I’ll comment on why I thought each one might be worth your time.

  • Online quiz on how you pronounce certain words. Attempts to identify where you grew up. Fun, and it worked for me. link
  • A son discovers only after his death the true extent of his dad’s friendship with NBA basketball star Charles Barkley. link
  • On the poisonous allure of social media shaming. An older topic, written in first person. Surprisingly good. link
  • A personal story on life when weighing 460 pounds. Poignant. The kind of topic that would be maudlin if done badly, but when done well, is universal. This is done well. link

And that’s all for this week. Hope you have a great 2019.

Apple’s new services strategy. Apple is commoditizing (some) of its complements: music, movies, TV.

Apple announced their music and movie services will be available on non-Apple hardware. Whoa! This is huge. Ben Thompson correctly highlighted the importance, saying this shift in strategy (link $) “is not just fascinating, it’s frankly a bit stunning.” And “I think these announcements are a much bigger deal than people realize: Apple is absolutely sacrificing its traditional hardware business model, at least when it comes to TV black boxes and smart speakers, in favor of a content-centric services strategy.”

Why such a big deal? Apple’s incredible success was built on a consistent model: use proprietary software and services to make money selling hardware. That was true for the Apple II, Mac, iPod, iPhone, iPad, Watch. Despite near bankruptcy in 1997, the Apple business model remained sacrosanct. Side note: I’d argue iTunes for Windows with iPod was an exception that proved the rule.

Apple’s vertically integrated software/hardware stack allowed them to create premium experience. And premium profits. But also meant Apple services were hindered strategically. A strategy tax if you will. Because the optimal services strategy (think Google) is horizontal, putting your services on all platforms and devices. The marginal cost of extending an existing service to another platform is low, so in some sense services “want” to be everywhere. More users -> more data -> better services.

Suddenly Apple is putting Apple music on Amazon Alexa speakers. Apple movie and TV shows on Samsung, LG, Vizio and Sony TVs. And allowing AirPlay to those TVs, the ability to send content from your phone to your TV. Apple is destroying the proprietary value of their hardware. Most significantly Apple TV and HomePod, but also to a lessor extent the iPhone itself. What gives?

What’s happening now in 2019 is something I wrote about in 2013, in a post which I think holds up well.

But longer term, say 5, 10 or even 15 years, the Apple services strategy tax creates a third possible Apple bear story. Phones are becoming more and more about services. Let’s say the phone-services split was 90-10 for the first iPhone. It was primarily about what you held in your hand, not what’s in the cloud. Let’s say it’s now at 50-50. You can see where this is going. It’s possible that in a world where the phone-services split is 10-90 the other way, a company which focuses exclusively on services could wind up with a better overall premium phone experience than a company which focuses on complete systems. The strategy tax could actually dislodge Apple from the premium end of the market. Clearly this is a long term speculative scenario, but it’s a plausible future. Especially if you segment the market, where segments focused more on services will reach their tipping point earlier.

Apple makes money selling premium hardware, but what the people are paying for is premium experience. That’s the job to be done. And right now Google maps, Google email, Google voice assistant, etc are all better than Apple. Nonetheless, since I wrote that post in 2013, Apple has continued making enough moderate (if lagging) progress on its services, plus has innovated on its hardware (Face ID, camera, displays, integration of Mac and iPhone, custom ARM processors, integration of Apple Watch and Airpods) to keep their overall experience better. But this approach is showing cracks. Hence the new services strategy.

My guess is Apple’s new services strategy is, in select areas, to commoditize the complements. From Joel Spolsky’s classic post:

A complement is a product that you usually buy together with another product. Gas and cars are complements. Computer hardware is a classic complement of computer operating systems. And babysitters are a complement of dinner at fine restaurants. In a small town, when the local five star restaurant has a two-for-one Valentine’s day special, the local babysitters double their rates. (Actually, the nine-year-olds get roped into early service.) All else being equal, demand for a product increases when the prices of its complements decrease. Let me repeat that because you might have dozed off, and it’s important. Demand for a product increases when the prices of its complements decrease. 

By putting music, movies, TV on all platforms, Apple has decided to stop making those services a premium differentiator for their hardware. They are commodities. Anyone can get them. Even customers who never buy Apple hardware. Hmmm….. what about the $3 billion Apple spent buying Beats in 2014? I’d say flushed down the toilet.

To be fair, there’s another possibility. Apple may be attempting to simultaneously create premium horizontal services with music/movies/TV, and retain their traditional premium vertical stack hardware business. Ben Thompson suggests this possibility when he says “In short, it seems that Apple can have its (services) cake and eat it too: offer exclusive content across all platforms, while preserving its integrated experience (and iMessage) for its moneymakers.”

Perhaps. But the strategy conflict remains. Making a cross platform service means avoiding proprietary customization. Exactly what made those services work better on Apple hardware. That said, I know enough about huge corporations to know how this will play inside Apple. The people in the music/movie/TV services group want it to be best in market. And that’s fine. Good for them! Perhaps they’ll succeed. But even if they become the best, they won’t differentiate Apple hardware. To the hardware teams at Apple, those services are now pure commodities, and the cheaper those commodities become, the more valuable their hardware complement.

The big open question here is which other services at Apple are candidates to go horizontal. If my commoditize the complement theory is correct, it should be all services which don’t move the bar on selling Apple hardware. Let’s try three buckets:

  1. Cross platform now: music, TV, movies, itunes
  2. Cross platform likely: books, all media content, parts of iCloud
  3. Cross platform unlikely: Siri, iMessage, maps, App store, Apple pay

At first I wanted to put Siri into bucket #2. The strategy trade off is acute for voice interface. Google and Amazon can partner with everyone, getting more users and data, making their voice interfaces better and better. Apple lags. But on the other hand, Apple is famously future product oriented. And I started thinking about which future product might benefit most from tight hardware/software integration. The (or at least one) answer is augmented reality (AR) glasses. So even though the pressure for Siri to go horizontal right now is large, I suspect it’ll remain proprietary. I moved Apply pay into bucket #3 for the same reason. Whenever Apple AR glasses arrive, buying stuff with Siri voice and Apple pay will be frictionless. And in tech, frictionless is money. Pure money.

Update January 15 (two days after original post): Apple just announced they are partnering with search provider DuckDuckGo on Apple maps. Maps is a critical strategic capability, so perhaps my guess on Apple’s new services strategy was wrong. That said, I stand by my fundamental point that to the extent a service is available everywhere, it becomes a complement to selling hardware, creating an economic incentive to keep the margin on those widely available services low. And to be clear, partnering with a particular search provider is very different from putting Apple maps on Android. Which remains extremely unlikely. In any case, a lot happening right now with Apple services strategically. Something to keep a close eye on for 2019.