Don't corrupt your objective function by Mike Hudack

I recently switched to Apple Music from Spotify.

The nominal reason for the switch was 4:44. Jay-Z's newest album wasn't available on Spotify. And it wasn't just 4:44 that wasn't available. He pulled his entire catalog. Jay-Z engaged in a classic windowing strategy in which the album was first made available to Sprint and Tidal customers, and then on three streaming platforms that aren't Spotify: Apple Music, Google Music and Amazon Music. 

It's meaningful that Spotify wasn't in on the second window. I think it's because they're not willing to pay. 

Spotify's unwillingness to pay goes beyond losing out on content like 4:44. I fear that their pursuit of margin in advance of a forced IPO is hurting customers through algorithmic distortion.

A while ago Spotify created a series of mostly excellent data-driven products like Discover Weekly that promised to revolutionize the way we discover music. They worked. By using what must be an incredibly rich library of music listening data Spotify was able to create incredibly powerful discovery products that "just worked." I'm sure I wasn't the only one who would regularly just play my Discover Weekly instead of doing work to find albums and playlists to listen to. 

Over time Spotify's algorithmic offerings expanded. I started listening to Daily Mixes. Spotify advertises Daily Mixes as "[T]he music you love, minus the effort."

They continue:

Daily Mix is the perfect line-up of tracks ready to play at the touch of a button. Based on the different styles of music you regularly listen to, each mix is loaded with artists you love, plus a sprinkling of new discoveries that fit the vibe too. What’s more, it grows with you. As your music taste evolves, so does your Daily Mix.

Daily Mix is a great product. It has just one problem: it doesn't include all the music you love. 

Every song Spotify streams costs them money. Some songs cost more money than others. The most expensive songs come from the likes of Jay-Z. The cheapest songs come from fake artists created by Spotify (Spotify denies they create fake artists, but they do). 

If you're Spotify and you want to become profitable and you have to IPO because you raised a billion dollars under punitive terms that punish you for every day you don't IPO and your margins are terrible because everyone just wants to listen to Jay-Z what do you do? 

Well, to start with, you don't carry 4:44. But then you take the fabulous data products that you created -- Discover Weekly and Daily Mixes -- and you distort them. You change the objective function from being one that purely optimizes for customer value to one that attempts to balance customer value and profit. 

I can imagine the conversation, and it probably all seems reasonable and good. People love these playlists. They also love generic piano music. What if we played a little more generic piano music and a little less Arthur Rubinstein? No one would really notice

Then the test gets designed and run. It's a 28 day test in which a control group continues to receive the traditional playlists with the traditional customer-optimized objective function while an exposed group receives a playlist which includes "royalty cost" as a feature in the algorithm.

I imagine that the test shows no or very little difference in engagement between test and control and no measurable effect on retention, free or paid. The conclusion that's reached is that it's okay to include Spotify economics as a feature in the algorithm. 

Months pass. The impending doom of a looming IPO combined with crushing unit economics continue to weigh heavily on the shoulders of Spotify executives. The objective function gets tuned to place ever more value on Spotify's economics. The knob keeps turning. The tests keep showing no significant impact on retention. 

But then 4:44 comes out. A few people hop over to Apple Music or Google Play or Amazon to listen to it, because they like Jay-Z and heard 4:44 is good. Then, if you're me, something funny happens when you jump over to listen to 4:44. You start listening to expensive music again. 

Apple Music doesn't care if I listen to Arthur Rubinstein or Relaxing Piano Music. So I listen to Arthur Rubinstein. His Chopin Nocturnes are better. And DAMN. is good. I didn't realize it, but it'd been a really long time since I listened to The White Stripes. And there you are, Pusha T. 

Apple Music has their own version of Discover Weekly now. It's called "My New Music Mix." Mine has a really nice mix of up-and-coming acts and brands you know. Their "My Favorites Mix" has Tom Waits and the Ramones and Bob Dylan and CCR and Radiohead and Kendrick and Jay-Z. 

There is no fear of major labels at Apple Music. And this means that the experience is better. I'm listening to better music now. Indie music is getting surfaced because the algorithms that power Apple Music believe I'll like it, not because it's cheaper for Apple to stream. 

I'm not going back to Spotify. When I pick up an Android phone again I'll do it with Google Play Music. Spotify has treated me like a lobster in a pot of water. The heat kept getting turned up, slowly, and I barely noticed. But now that I've tasted an alternative streaming service I realize how mistreated I've been. 

This should be a lesson to anyone building algorithms of this sort. Don't optimize for yourself. Optimize for your customer. Figure out your own economics and your business separately. The moment that you cross that line your algorithm becomes corrupt.

You'll try to measure the degree to which you're hurting your customers, but your measurement will be wrong. When people do bad things for customers they never run the retention tests long enough to see the true impact of their changes. They say it's because they want to move fast, but it's really because they don't want to know the truth. If they wanted to know the truth they'd have a long-term holdout. They almost never do. 

Also: you should consider exclusive content. It's a smart marketing tool. Exclusive content causes customers to keep sampling your product, and allows you to iterate towards something good while getting multiple cracks at customers. Apple Music wasn't that great when it launched, but its content strategy has kept me coming back periodically for treats like 4:44. I kept doing that until I stuck around. 

But there's a broader point here. You should never let your interests diverge from the interests of your customers in the first place. Your focus in building software should be in improving customer experience, not increasing profit. Profit is a lagging indicator of customer happiness, not the other way around. 

Google Home pulls further ahead by Mike Hudack

I’ve been saying it for a while: Google Home is going to win the home assistant battle. Google Home and Google Assistant will win because Google is better at conversational AI than Amazon and Apple, and will improve at a faster rate than either of them because of its huge data advantage.

Google Home can now tell who’s talking to it, and answer them differently. This is a huge deal. It’s a first step towards artificial assistants becoming truly conversational. It helps make them seem human-like. It’s also hugely practical: I want to hear my schedule when I ask, and my wife wants to hear hers when she asks.

There are lots of other things that Google Home already does better, but this is a first example of what will become an insurmountable strategic advantage. Further differentiation will follow, at accelerating pace, as Google’s expertise and data advantage compounds. This is because in AI applications data advantages lead to exponential improvement over time.

 

Disturbing literature, disturbing elections, and the tyranny of leverage by Mike Hudack

Western culture is becoming pessimistic. Our literature is tilting towards near-future disaster. Classics like 1984 and It Can’t Happen Here are newly popularThe Mandibles and American War feature on bookstore tables around the United States. 

These books are popular because they explore a linear extrapolation of the current moment into the future: they plum the depths of our current politics, our economics, our sociology. 

The FT on The Mandibles

Known for tackling big contemporary issues head-on, Shriver deals skilfully here with the implications of economic meltdown. The novel, set in a near-ish future, tells of the plight of the once wealthy Mandible family and the decline of four generations into penury, thieving and prostitution.

It Can’t Happen Here and American War even feature the same barbed wire on their covers. I’ve had to put away my copy of The Mandibles because it’s just too painful at the moment.

Trump sits in the White House. Brexit and Theresa May rule Britain. More than one in five French voters voted for Marine Le Pen, leader of the National Front.

When my daughter reads history she will read that the populist moment of the teenage years of the 21st century was an inevitable and natural reaction to free trade, globalism, technological advance, the disappearance of low- and medium-skilled jobs and the 2008 crisis. Tribalism is the natural reaction to economic and social dislocation. But we still have a problem: the opposition is intellectually bankrupt.

Macron might triumph in France, but he won’t do so because people are particularly excited about his message. Macron is a former Goldman Sachs banker. He served as economy minister to the least popular president in the history of Western democracy. Macron is the candidate of the status quo running against a dangerous, racist, anti-Semitic, anti-Europe, anti-Euro populist. The only thing Marine Le Pen is not against is Vladimir Putin. This she has in common with Trump.

If Macron wins it won’t be because people like or appreciate him or his policies. It will be to save the Republic:

In order to save the Republic we must vote for the status quo cosmopolitan globalist who will continue to destroy it! https://t.co/01kDoMvrhk

-- Mark Sleboda (@MarkSleboda1) April 23, 2017

This is the problem. The world has changed in profound ways and no politicians anywhere are acknowledging the changes and offering a path forward. Low and semi-skilled manufacturing jobs are not coming back. Machine learning is fundamentally changing the nature not only of factory work, but of office work. Most college degrees aren’t worth anything. The sheer leverage enjoyed by the most talented amongst us drives staggering amounts of income inequality, and this secular trend is unlikely to reverse any time soon.

Dwell on this last point for just one second.

Think about the amount of technology that exists today to leverage the intelligence of a single person. One person can write brilliant code that out-performs a million people trading in the open markets. One person can design a factory that churns out new cars with one hundredth of the human involvement necessary only generation ago. One person can have an idea that changes an algorithm which changes the news people see which changes an election. 

Compensation naturally follows. We are now capable of having our impact so leveraged that we have dramatically different replacement values in professional settings, which in turn leads to dramatically different compensation. Historically the difference in performance between the best accountant in the world and an average accountant was relatively small because the leverage with which they could operate was limited. These days the best accountant writes software that scales her brilliance and obviates the need for the average accountant. This is the impact of Information Technology on our world. It is more fundamental even than the impact of the shipping container

My daughter will study the populist movement as a natural outbirth of globalism and 2008. She will also study the emergence of new political philosophy that helps us deal with the social and economic consequences of massive personal leverage. But no politician — or author, or philosopher, or economist — is yet offering us this political philosophy. So we are left with a choice between Macron and Le Pen while we read stories like The Mandibles which explore the world we’ll live in if we don’t figure this shit out. 

Spoiler alert: it’s fucking depressing.

Police violence strikes the middle class by Mike Hudack

The video of a man being forcibly removed from a United flight is one of the most disturbing things I’ve watched in quite a while. There’s another video of him back on the plane, bleeding, repeating “just kill me.” Eventually he was removed on a stretcher.

The man who was brutalized at the order of United Airlines and Chicago’s transport police was a doctor on the way home with patients to treat in the morning. His crime? Refusing to give up his seat for a United employee because he had patients to see the next day.

Middle class Asian doctors on airplanes are supposed to be safe from the violence of the state. There is something seriously wrong with the behavior of American police if they act in this way against this man in this setting. Seriously wrong. If they’ll do this to this man on this plane what will they do to a black man on a deserted street in the south side of Chicago?

The information bubble by Mike Hudack

There’s been a lot of hand-wringing about filter bubbles lately, but not information bubbles.

Information bubbles exist when people inside an institution have access to information that’s very different from the information available from the outside. Information bubbles cause people within them to make decisions differently than those outside, and they frequently cause the institutions to be popularly misunderstood.

A common feature of information bubbles is that people within the affected institutions often find reporting about them to be inscrutably wrong.

The article Daniel Ellsberg, Edward Snowden, and the Modern Whistle-Blower in the New Yorker describes information bubbles exceedingly well.

In the article the author quotes from Daniel Ellsberg’s auto-biography about preparing Henry Kissinger for his time in government:

You will feel like a fool for having studied, written, talked about these subjects, criticized and analyzed decisions made by presidents for years without having known of the existence of all this information, which presidents and others had and you didn't, and which must have influenced their decisions in ways you couldn't even guess…

It will have become very hard for you to learn from anybody who doesn't have these clearances. Because you'll be thinking as you listen to them: "What would this man be telling me if he knew what I know? Would he be giving me the same advice, or would it totally change his predictions and recommendations?" And that mental exercise is so torturous that after a while you give it up and just stop listening. . . . The danger is, you'll become something like a moron.

This resonates. When I worked at Facebook almost everything I read about the company was wrong. Mots of the “facts” reported about the company were wrong unless they were cribbed directly from earning announcements. Strategy analysis was almost always off the mark. The tone of coverage had more to do with the reporter and our relationship with them than the company’s actual actions. When the company’s actions were responsible for the tone of coverage it was usually due to how we communicated our actions rather than the actions themselves.

This must also be true of governments. Who knows what’s going on behind the Russo-American election drama. We have no way of knowing what other actions have been taken, or what information either side is acting on. We have only reportage and tweets from the Russian Embassy in Washington.

Is the US government stealing Putin’s money from his bank accounts? Disabling oil pipelines? Messing with Putin’s own upcoming election? We won’t know for a generation.

Similarly, we can’t know what it’s like to be inside Apple as it’s lapped by Google, or what it was like to be inside Microsoft as it was lapped by Apple. We can’t know why Musk really decided to have Tesla buy Solar City.

This does not mean we should give up on understanding the world around us. It just means that we should be humble.

It is worth repeating this adage frequently: “Everything you read in the newspaper is absolutely true, unless it’s about a subject you know something about.”

The Pixel has put my iPhone 7 in a drawer by Mike Hudack

Google Home is so good that it inspired me to try the Pixel.

I’m two weeks into using the Pixel and I don’t know if I’ll ever go back to iPhone despite having been a (mostly) loyal iPhone user for almost a decade.

The hardware is just as good

Pixel is the first phone to challenge iPhone on its own turf. Its build quality and software quality is as good as iPhone. Apple’s previously insurmountable advantage in product design and manufacturing has been bridged.

Pixel’s hardware feels solid and expensive. The screen is at least as good as iPhone’s. The camera is also as good or better. The fingerprint sensor is faster and more reliable. Battery life is longer and better managed.

The software is better

Pixel’s software is excellent. It’s better than iOS. Android has matured, and Google’s finishing layer that’s unique to Pixel is also excellent. Pixel’s OS is tasteful and well thought through. The jank previously native to Android has been excised.

More importantly, a lot of the software is materially better than iOS:

  • Notifications are excellent. You manage them in the notification tray, on the notification. This is the paradigm that makes most sense. You can also take myriad actions directly from notifications. You can reply to messages, upvote a location in Foursquare, or look at a snapshot from your Nest cam in your notification tray. Apple tried to copy this functionality in recent versions of iOS but their copies are poor substitute’s for Google’s real thing.
  • The OS share dialog on Android is also better. The share dialog offers the ability to share through apps, but also directly to people. You can choose to open WhatsApp or share directly to a frequently contacted friend on WhatsApp. Totally brilliant.
  • Google Now integration is very good. Like with notifications, Apple’s attempt to copy this functionality in its left drawer and in the notifications tray falls far short of the real thing.
  • Google MapsGmail and Google Calendar are all better on Android.
  • GBoard is phenomenal. I’m swipe typing for the first time, and it’s awesome. Integrated GIF search and the like is also epic. Again, Apple’s attempts to copy this functionality fall far short.
  • Assistant is epically good. Moving from Siri to Alexa felt like moving from an Apple II to a Lisa. Moving from Alexa to Assistant is like moving from Lisa to a Mac.
  • Even small things are good and well thought through. Two factor authentication with Google services works with a fingerprint instead of a shared code from Google Authenticator. This is a very welcome and really nice flow that shows just how far Google has come in user experience design.

It’s also worth mentioning that I got a free Daydream View with the Pixel. It’s a better and less expensive version of the Gear VR. It’s still a novelty, but it’s also obviously part of the future.

But it’s really about machine intelligence

These things by themselves might not be enough to get you to throw your iPhone away for a Pixel. It’s the AI that’ll get you to do that.

This past weekend Caroline and I were in Paris. We traveled home by way of the Eurostar at Gare du Nord, a Parisian train station. When we arrived at the station Android automatically pushed a card to me offering a translation of “Where do I buy train tickets?” from English to French. It also offered a convenient speaker button that I could use to have Google speak the phrase for me in French if I was unsure of my pronunciation.

Pixel knew that I was an English speaker, that I was visiting France, that they speak French in France, that I was in a train station, and that maybe I would like to be able to ask someone where to buy train tickets but might not have all the words to do so. That’s service.

This is only a small example of the machine intelligence baked into Pixel. Google Photos automatically organizes your pictures with facial recognition and other features.

The phone recommends coffee shops to you, tells you what museums to go to, and tells you what time you need to leave for you next appointment. It tells you when your bills are due.

You can ask Assistant “Who is Mariano Rivera?” and then “What’s he worth?” and it’ll answer both questions perfectly.

Apple is trying desperately to copy this functionality and is failing. To use Pixel is to understand how machine intelligence should be integrated into a phone and into your life.

Android phones used to be poor copies of iPhone. Now iPhone feels like a poor copy of the Pixel.

There are only two reasons not to switch

There are only two reasons not to switch to the Pixel: services lock-in (iMessage, FaceTime and iPhoto) and concern about switching to an unfamiliar platform.

Services lock-in is hard to fight. WhatsApp, Messenger and SMS are poor substitutes for iMessage and FaceTime. Android doesn’t offer a competitor, which is why Google released Allo and Duo. Pickup of Allo and Duo have been disappointing, which means that Google is at a structural disadvantage in this department. They’ve got to catch up here in order to truly compete effectively.

But on familiarity, Google has a winning strategy. You’re already using Android apps on your iPhone. Google Maps, Gmail, Google Calendar, Chrome, Sheets, Docs and Gboard are all iPhone apps built by Google in the style of Android, using Material Design. They are identical to their Android versions, only not as good.

This means that once you’ve used the Google suite of software on an iPhone you’ll barely notice the change to Android. You’re already used to Material Design. And other apps aren’t a problem either. Almost all of the apps you use on iPhone are available on Android, and these days they’re just as good.

This is now Google’s game to lose

At some point you’re going to see something on a Pixel or another flagship Android phone that wows you. It may be Assistant, or it may be in Google Photos or somewhere else. It’ll probably be an application of Google’s lead in machine intelligence, although it might be better notification handling or something like that. You’ll say “I want that,” and Apple’s excellence in hardware design won’t be enough to outweigh Google’s excellence in software. You’ll switch.

I still believe that Apple is ending a super cycle and starting a new one. They’re at a low point from which they’ll likely recover. But after experiencing Google Home and Pixel I’m convinced that recovery won’t be easy. Google has colonized my pocket with Pixel and our home with Google Home. Our TVs might soon get Chromecast

Living with two AIs by Mike Hudack

We have been living with Alexa since it came out. We have three Echos: one in the kitchen, one in our bedroom and one in our living room.

Over time our house has gradually gotten smarter, and Alexa has become more useful. We’ve gone from just using the flash briefing and occasional music streaming to using her to control our kitchen lights and the temperature in various rooms around the house using Tado smart thermostats and radiator valves.

For a while I was the exclusive user of Alexa in our house, but over time Caroline started using her more and more. Her use of Alexa was accelerated massively when Amazon formally launched Echo in the UK and an English localization became available. Alexa suddenly understood Caroline’s mid-Atlantic accent better than before. The BBC led our flash briefing. Usage soared.

Then, on a recent trip to New York, I picked up two Google Homes. I put one in the kitchen and one in our dressing room.

We don’t use Alexa anymore. If you’d like an Echo, we’ve got one to sell you. Three, in fact.

Most reviews claim that Google Home and Echo are roughly equivalent, and that at the moment Echo is marginally better. We have not found this to be the case in our house.

Google Home’s lack of “Skills” — the little voice applets that populate Amazon’s Alexa app store — is more than made up for by support for SpotifyPhilips HueSamsung Smarthings and IFTTT.

The IFTT integration is particularly sick: it allows you to hijack any phrase and use it to trigger an IFTTT recipe. You can even pass variables, so you can say things like “Make the bedroom 20 degrees,” and pass the values “bedroom” and “20” on to whatever heating system you’re using which is also IFTTT-compatible. It feels like magic, and it’s instant. It beats the hell out of developing a custom Alexa skill with Lambda “in under an hour.”

The Spotify integration is also excellent. “Hey Google, play music” starts playing your Discover Weekly. You can ask for any of your Spotify playlists. You can say “Play Willie Nelson” or “Play the Goldberg Variations” and it just works. Alexa routinely plays the wrong music and forces you to be very specific: “Play the album The Goldberg Variations by Glenn Gould.” Google needs no such specificity. You can talk about music with it like you would a friend.

Google Home is aware, unlike Echo, that your home probably has multiple rooms. You can ask it to “Play The Goldberg Variations in the bedroom.” Better yet, if you’ve got a Chromecast hooked up to your stereo, you can say “Play the Goldberg Variations on the stereo.” It just works.

And this is another key difference between Google Home and Echo. Alexa makes you feel like you’re typing at a command line. Your syntax must be perfect. Google Home is forgiving, and seems to legitimately get better at understanding you over time. “Hey, Google, what’s my day look like?” works. You’ll be told what the weather’s like, what your first meeting is, how long your commute will take and then Google will start your news report. Bam.

Alexa feels like a novelty that offers a glimpse of the future.

Google Home feels like the future. We actually use it. Often. It’ll even translate for you. “Hey, Google, how do you say ‘I don’t speak French’ in French?”

Google wasn’t kidding when they said that they were behind in releasing a home product but that their AI was much more sophisticated than anyone else’s and that they’d quickly catch up. This race is now Google’s to lose. And it’s making me want to get a Pixel.

Apple and Amazon have their work cut out for them. If you can’t even get AirPods to work as your ubiquitous interface to your sub-par AI then you’re really going to have a hard time competing against something as fantastically good as Google Home.

[The Alexa/Echo, Assistant/Home thing is weird, and so I’ve tried to write around it.]

Coming Apart by Mike Hudack

We were warned. We should have known. No one has any excuse.

Brexit showed that polls are useless when one option is less socially acceptable than the other. People were ashamed to tell pollsters they were going to vote for the crude, lewd, unqualified Trump. They lied in telephone polls and they lied in exit polls. But they didn’t lie in the voting booth. Brexit should have taught us this lesson, and shame on the pollsters for ignoring it.

But more importantly, Charles Murray called yesterday’s election a couple of years ago in Coming Apart: The State of White America, 1960-2010. In Coming Apart Murray talks about the decay of white working class life in America. He also talks about how the “winners” in American society are increasingly sorted from the “losers,” such that they barely touch. The losers are aware of the winners. The winners are not aware of the losers.

Murray explores the increasing homogenization of individual neighborhoods: the spread of super-rich and super-poor ZIP Codes. He talks about Washington, DC and Virginia and New York City in one breath and Michigan in the next. He compares, he contrasts.

His conclusions are inescapable:

The top and bottom of white America increasingly live in different cultures, Murray argues, with the powerful upper class living in enclaves surrounded by their own kind, ignorant about life in mainstream America, and the lower class suffering from erosions of family and community life that strike at the heart of the pursuit of happiness. That divergence puts the success of the American project at risk.

Murray was not ignored by the political class or the mainstream media.

David Brooks wrote in the New York Times that “I’ll be shocked if there’s another book this year as important as Charles Murray’s Coming Apart.”

Bloomberg Businessweek wrote “Charles Murray has written an incisive, alarming, and highly frustrating book about the state of American society.”

Publisher’s Weekly: “A timely investigation into a worsening class divide no one can afford to ignore.”

But it was ignored. Murray warned us. He was ignored.

Murray wasn’t the only one. While Murray wrote about the growing segmentation and alienation of American society and the decay of rural white America, Robert Putnam wrote about the decline and fall of American civic society in Bowling Alone: The Collapse and Revival of American Community, all the way back in 2001.

Here’s a jacket quote for Bowling Alone from Amazon:

The conclusions reached in Bowling Alone rest on a mountain of data gathered by Putnam and a team of researchers since his original essay appeared. Its breadth of information is astounding–yes, he really has statistics showing people are less likely to take Sunday picnics nowadays. Dozens of charts and graphs track everything from trends in PTA participation to the number of times Americans say they give “the finger” to other drivers each year. If nothing else, Bowling Aloneis a fascinating collection of factoids. Yet it does seem to provide an explanation for why “we tell pollsters that we wish we lived in a more civil, more trustworthy, more collectively caring community”. What’s more, writes Putnam, “Americans are right that the bonds of our communities have withered, and we are right to fear that this transformation has very real costs”. Putnam takes a stab at suggesting how things might change, but the book’s real strength is in its diagnosis rather than its proposed solutions. Bowling Alone won’t make Putnam any less controversial, but it may come to be known as a path-breaking work of scholarship, one whose influence has a long reach into the 21st century.

Putnam wasn’t ignored either. The Washington Post called him “The de Tocqueville of our generation.” The New York Review of Books wrote “Rich, dense, thoughtful, fascinating…packed with provocative information about the social and political habits of twentieth-century Americans.”

Cass Sunstein sounded the alarm about social media and the Internet’s impact on our civic life in Republic.com and Republic 2.0. From the Amazon jacket:

Republic.com 2.0 highlights new research on how people are using the Internet, especially the blogosphere. Sunstein warns against “information cocoons” and “echo chambers,” wherein people avoid the news and opinions that they don’t want to hear. He also demonstrates the need to regulate the innumerable choices made possible by technology. His proposed remedies and reforms emphasize what consumers and producers can do to help avoid the perils, and realize the promise, of the Internet.

Again, Sunstein wasn’t ignored. The New York Times Book Review wrote that the book “Raises important and troubling questions about the effect of the Internet on a democratic society.”

The Financial Times wrote “Cass Sunstein sounds a timely warning in this concise, sophisticated account of the rise of the internet culture. He argues that it is our very ability to wrap ourselves in our own tastes, views, and prejudices with the aid of technology that constitutes a real threat to the traditional democratic values.”

Those three books sum it up:

  • The 21st Century economic order has been kind to an elite few but unkind to most;
  • Institutions of civic society that once brought us together — from fraternal organizations to bowling clubs to churches — have disappeared;
  • The Internet, and social media particularly, has pulled us apart rather than pushed us together.

But something even more important is at play here: Millenials are making less than their parents. American productivity growth is stalled. The great promise of liberal western democracy and economics is, for the first time in modern history, failing to deliver the goods. This is what you get.

And make no mistake: This situation is global. It’s true in America, in Britain, in France. A lot of people have been left behind by the 21st century. God is dead. Humanism as our new religion doesn’t give people the safety and security that Christianity did. The factory is no longer employing the town. Global elites are increasingly segregated and isolated in Paris, in London, in New York, in San Francisco. Yesterday it was Brexit. Today it is Trump. Is tomorrow Le Pen?

Who's running the Protonmail honeypot? by Mike Hudack

Protonmail is a service that offers “secure email based in Switzerland.” It offers features like end-to-end encryption, anonymous accounts and “Swiss privacy.” Read: it’s subpoena-resistant. It’s also “100% free,” and seems well designed. Mr. Robot uses ProtonMail.

Anyone want to take a bet that Protonmail is run by an intelligence agency? Secondary wager can be placed on which of the major powers the intelligence agency is aligned with: American (includes allied countries like France and Israel which would have this capacity), Russian or Chinese.

My bet’s on American-aligned.

Marketplaces come and go, but sellers maintain their reputation by Mike Hudack

Joel Monegro’s Deep Web Marketplaces is an excellent read.

I was struck by these three points:

  • A seller's brand and reputation are extremely important in a system where the intermediary (the marketplace) does not guarantee trust and safety.
  • This is largely decentralized in deep web marketplaces, as vendors make sure their brand is spread across multiple websites and forums.
  • Marketplaces come and go (or get seized by the FBI) but sellers need maintain their reputation.

I recently got a profile on Onename, which is powered by Blockstack’s blockchain identity product. I suppose this is meant to be the pan-marketplace identity and reputation system, at least in USV’s blockchain portfolio. I don’t have much use for it right now (Facebook is fine and will even store your PGP public key for you) but I can certainly see how, if I were selling drugs on a “dark” marketplace, I would want something like Onename to keep my identity safe as the Feds play whack-a-mole.

The deeper question is whether law abiding citizens who aren’t cypherpunks might want such a thing as decentralized identity maintenance. If and when they do, it won’t be because it’s “decentralized identity maintenance.” They’ll care about the blockchain about as much as they care about http, which is to say not at all.

Cookie warnings and Brexit by Mike Hudack

Yesterday, while trying to buy tickets from my phone for Schoolboy Q at Brixton Academy, I kept getting full-screen and difficult to dismiss modals warning me that TicketWeb uses cookies. This terrible modal is necessary because of the EU Cookie Regulation

From the UK Information Commissioners Office, the office responsible for enforcement of the regulation:

You must tell people if you set cookies, and clearly explain what the cookies do and why. You must also get the user’s consent. Consent can be implied, but must be knowingly given.

The regulation is idiotic. Nearly every web service (and app) in the world uses some mechanism for tracking users and almost all those mechanisms use cookies. Tracking is used to provide utility to end users and also to make money. 

The cookie regulation causes a ton of damage, the worst being that I almost didn’t buy my Schoolboy Q tickets because I got about half a dozen cookie modals and they made me so frustrated that I nearly abandoned the purchase flow. That would have created about £75 in economic loss attributable directly to this regulation.

There’s an old Erwin Knoll quote that goes like this:

Everything you read in newspapers is absolutely true, except for that rare story of which you happen to have first-hand knowledge.

There should probably be a corollary: “Every regulation issued by the European Union is absolutely proper and correct, except for the rare regulations in areas in which you are expert.”

If the EU Cookie Directive is at all indicative of other areas of EU regulation then maybe Brexit has some sense to it. There must be an equivalent of the EU Cookie Directive in the Common Agricultural Policy and in so many other areas of EU jurisdiction. Maybe all of them.

And before you say that all regulation has its equivalents of the Cookie Directive and that an exited Britain would just introduce its own equivalents, let me suggest one more rule: the larger and less representative a regulatory body is, the more likely it is to produce stupid regulation.

Estonia probably has less bad regulation than larger countries like the United States, and the United States probably has less bad regulation than supranational blocs like the European Union. So theoretically a smaller and more representative state should produce less bad regulation and more good regulation, if only because those being regulated have more of a direct voice in government.

Which, after all, was the reason why America exited the United Kingdom a few hundred years ago.

PS: Schoolboy Q is awesome, and while I’ve never been, I hear Brixton Academy is too. You should go to the show so we can hang out and grab a drink.

Relying on bank lending for stimulus is inefficient by Mike Hudack

It’s struck me for a long time that much of monetary policy that’s meant to stimulate lending actually benefits people who are already wealthy and who are also, therefore, less susceptible to stimulus intended to increase their spending.

I also think that these policies are responsible for rising inequality in the world in the past decade or so as money has become cheaper and cheaper for people who already have it, with little real change for people who don’t already have it (since their wages have been stagnant and their borrowing ability is therefore limited).

A much more efficient program would be a form of helicopter money in which the government or central bank simply gave money to people who most “need” it, and who will quickly spend it. Such an approach would be much more efficient: it would get rid of all the middlemen and make every dollar much more likely to enter circulation since the poorer people receiving the checks would get immediate benefit from spending it (and, unlike rich people, relatively little benefit from saving it).

Such a direct stimulus program, however, is difficult to implement using current money technology: the government would literally have to write everyone checks, or physically hand them cash. This is part of the reason I believe we should build a new cryptocurrency-based backend for our money system — to make policy response to the next recession better and more efficient.

Anyway, the St. Louis Fed, as part of their ongoing research into the policy response to the Great Recession, has published a new analysis showing that Fed policies designed to stimulate bank lending weren’t all that efficient at achieving the policy goal:

To help address the Great Recession, banks were provided with lower-cost capital and liquidity. One of the goals was for banks to pass these benefits through to consumers by extending extra credit, encouraging them to borrow more to stimulate the economy.

Recent research shows that these credit pass-throughs made their way to some consumers. However, they didn’t go to the consumers who were most likely to spend and boost the economy.

The Fed and other central banks should be moving much faster to transition the monetary system so that they can deploy helicopter money effectively in the future.

Filter Bubbles and the Universal Hipster by Mike Hudack

The Verge captures the homogenization of culture that’s been bothering me for a long time:

Igor Schwarzmann is the German co-founder of Third Wave, a strategy consultancy based in Berlin that works with small-scale industrial manufacturers. The company’s clients range across Europe, the United Kingdom, and the United States, so Schwarzmann often finds himself moving between poles of the global economy. While traveling, he turns to Foursquare for recommendations about where to eat and drink. “It knows what I like,” he says.

Every time Schwarzmann alights in a foreign city he checks the app, which lists food, nightlife, and entertainment recommendations with the help of a social network-augmented algorithm. Then he heads toward the nearest suggested cafe. But over the past few years, something strange has happened. “Every coffee place looks the same,” Schwarzmann says. The new cafe resembles all the other coffee shops Foursquare suggests, whether in Odessa, Beijing, Los Angeles, or Seoul: the same raw wood tables, exposed brick, and hanging Edison bulbs.

It’s not that these generic cafes are part of global chains like Starbucks or Costa Coffee, with designs that spring from the same corporate cookie cutter. Rather, they have all independently decided to adopt the same faux-artisanal aesthetic. Digital platforms like Foursquare are producing “a harmonization of tastes” across the world, Schwarzmann says. “It creates you going to the same place all over again.”

I wrote about this in a different context the other day. Filter bubbles and increasing intra-class connectedness (but not much extra-class connectedness) are making our world, at least the cosmopolitan corners of it, homogenous.

Independent coffee shops around the world are all the same. Every city has good New American food. Our intra-class connectedness means that Brooklyn and Dalston and Kreuzberg and the Mission are extremely similar. We’re losing local character and losing extra-class empathy. The Dalston hipster doesn’t have much empathy for the inhabitants of post-Industrial Britain, and the Brooklyn hipster couldn’t care less about the former West Virginian coal miner.

This has always been true to at least some extent, but it’s getting worse. Filter bubbles and cheap jet fuel make it so.

Schwarzmann’s coffee shops are all the same because of the network and its filters. The relationship between them is as circular and emergent as a ranked feed. We all have one aesthetic now. We are all one. Except for those of us who don’t get the technology and don’t live in a major cosmopolitan city. They inexplicably vote for Brexit and Trump.

This will take a while to be sorted out. There will be pain in the meantime. But at least during this process we know that we can get a good quality single origin pour over coffee almost anywhere in the world, and with some nice smashed avocado toast on the side. With chili flakes and lemon.

Refuting Picketty by Mike Hudack

A new IMF working paper refutes Picketty:

Thomas Piketty’s Capital in the Twenty-First Century puts forth a logically consistent explanation for changes in income and wealth inequality patterns. However, while rich in data, the book provides no formal empirical testing for its theoretical causal chain. In this paper, I build a set of Panel SVAR models to check if inequality and capital share in the national income move up as the r-g gap grows. Using a sample of 19 advanced economies spanning over 30 years, I find no empirical evidence that dynamics move in the way Piketty suggests. Results are robust to several alternative estimates of r-g.

From the conclusion:

On inequality, the evidence against Piketty’s predictions is even stronger: for at least 75% of the countries, the response of inequality to increases in r − g has the opposite sign to that postulated by Piketty.

This finding makes sense to me since Picketty doesn’t make sense to me. I find it much more likely that power is shifting  away from financial capital than that financial capital is accumulating unchallengable power. 

What I mean by this is that capital is no longer scarce — it’s productive applications of capital that are scarce. If this is true I think it means that Picketty is wrong. 

Rate cuts and fiscal policy by Mike Hudack

The Bank of England cut rates by 25 bps this week. Their action was a response to a torrent of terrible economic data in the aftermath of the Brexit vote.

I think that for most people the promise of a marginally cheaper mortgage probably isn’t enough to get them buying a new house when they’re unsure about their job.

Similarly there doesn’t seem much point in taking out debt to buy new factory equipment — even if the debt is really, really cheap — if profits are falling and your long-term sales outlook is emphatically cloudy.

The British economy doesn’t seem like it needs more asset bubbles spurred by cheap money. It needs customers, jobs, markets to expand into (and maintain access to), quality infrastructure and a friendly regulatory environment.

Brexit is a political problem and needs political solutions. It’s time for fiscal stimulus and clear statements of future government commercial and trade policy.

The government must speak with one voice, clearly and loudly. And best of all, money is historically cheap! There’s never been a better time than now for government to borrow to fund some new infrastructure.

One could maybe start with supporting a nationwide fiber upgrade since BT doesn’t seem to be doing it. Committing to an accelerated Crossrail 2 would be nice too. We live just down the block from the station, and such a commitment would do wonders for our post-Brexit property values. Thank you in advance, Mrs. May.

The biggest problem with Alexa by Mike Hudack

We have three Echoes in our house. The one in the kitchen gets the most use. We use it to listen to the radio in the morning, to play music, set timers and do kitchen conversions.

We don’t use it for anything complicated. We’ve tried, but gotten frustrated quickly and given up. The reason is that Echo is bad at keeping state. It can’t have a conversation. I think part of this is a simple UI problem: it’s hard to have a conversation with a black cylinder. Humans offer lots of non-verbal feedback during conversations, and Alexa offers virtually none. Even if her software were capable of maintaining state during a complicated conversation as a human does a person talking with her would likely lose track because there’s no feedback mechanism other than words. She doesn’t even have different tones of voice.

Changes in intonation, posture, raised eyebrows, glances and shifting weight all help us keep track of where we are in a conversation. Absent these cues we use conversational pauses and tone. Absent these rougher cues — in messenger platforms or e-mail — we use written history to keep track of our conversation.

Alexa offers none of these affordances, and so she’s hard to have a conversation with. It strikes me that until she offers some kind of feedback other than words spoken in a monotone we won’t find it easy to do much with her other than ask for Radio 4 or an alarm when the boiled eggs are ready.

The rise and fall of American productivity by Mike Hudack

The review (and summary) of The Rise and Fall of American Productivity in the New York Review of Books is excellent, although pessimistic.

It argues that economic progress is generally terrible and that 1870 – 1970 was wonderfully exceptional… and that the party is now over. The author’s somewhat counter-intuitive explanation is that the pace of technological advancement has slowed since 1970. 

I prefer to think that modern economics is as bad at measuring today’s progress as the author suggests it was in his 1870-1970 “special century.” 

Just as electricity wasn’t properly accounted for in records of progress over that special century, access to all the worlds information isn’t today. And so on and so forth. 

Still, the review is well worth reading.