Crashing

A couple of years ago, to test out some software I was helping develop, I installed the MacOS and iOS developer betas on my iPhone and trusty MacBook. And, in short, it was an unmitigated disaster. Features suddenly disappeared (apparently still in development), both devices unexpectedly rebooted repeatedly, and my productivity ground to a near halt. Eventually, I ended up rolling back both to stable, released software, and all was well, save the week or two of lost time.

In the time since, I completely forgot about that episode. Until this past weekend, when I once again, with software to test-drive, installed the developer betas of iOS 11 and MacOS High Sierra. And, once again, both of my daily-use devices are a total mess.

Given their fairly late-beta stage, this time I may just try to limp along through subsequent releases. But, if nothing else, it’s a good reminder: apparently, I just never learn.

July 19, 2017

Less Messy

Real problems are messy. Tech culture prefers to solve harder, more abstract problems that haven’t been sullied by contact with reality. So they worry about how to give Mars an earth-like climate, rather than how to give Earth an earth-like climate. They debate how to make a morally benevolent God-like AI, rather than figuring out how to put ethical guard rails around the more pedestrian AI they are introducing into every area of people’s lives.

– Maciej Cegłowski, Notes from an Emergency

July 11, 2017

Download This

On Tuesday, I was waiting for the subway with Jess and my parents, when an MTA employee announced that uptown train service had been temporarily halted due to a subway incident further uptown.

I pulled out my iPhone, and opened the Citizen app, which provides a real-time stream of incidents reported to 911, as well as a geotagged map of those locations, and live-streamed video from any Citizen user who happens to shoot a given incident.

Citizen told me that a subway had derailed near 125th St. minutes earlier, injuring more than 30 people, and strewing split open train cars across a number of tracks. I sent vibes of health and quick recovery to the injured passengers uptown, and suggested we head to the street to grab a cab; subway service wasn’t going to resume any time soon.

My father asked about the app, which reminded me that I frequently use a bunch of small, helpful, lesser-known iPhone apps. While a number of them are pretty specific, they’re each great if they match your specific use case. So, on the chance that you’d benefit from any you haven’t already discovered, here a handful of note:

===

Apnea Trainer (iPhone): If you swim, SCUBA dive, or free dive, Apnea Trainer provides a great breathing exercise that, when done for 5-10 minutes three times a week, hugely increases breath-hold time. Also great for decreasing symptoms of mild to moderate asthma.

Citizen (iPhone & Android): As described, real-time info on emergencies nearby. Currently NYC only.

Coach’s Eye (iPhone & Android): Essential if you coach athletes of any stripe. Allows you to record video of a movement, then play it back in slow-mo / mark it up with highlights and lines. Huge for providing specific visual feedback.

Dark Sky (iPhone & Android): Scary-accurate hyper-local real-time weather forecasts. If you’re considering, say, whether it’s a good time to walk your two small dogs, Dark Sky can tell you it’s about to start raining heavily in 13 minutes, then stopping about 25 minutes later. Useful (though slightly creepy) to get ‘it’s about to start drizzling’ notifications, then hear rain outside the window thirty seconds later.

Editorial (iPhone): A full-featured text editor with Markdown support and powerful automation. If you live in BBEdit, TextMate, Sublime, etc. on the Mac, this is by far the closest you’ll find on mobile. Paired with Dropbox, it’s great for doing work on the fly.

Fantastical (iPhone): The best calendar app on the iPhone, and the Mac, as you can quickly and easily enter events using natural language.

Feedly (iPhone & Android): Given that I write a blog, it probably shouldn’t be surprising that I read them, too. Feedly is by far my favorite RSS reader, both through their website and their excellent app.

Foursquare (iPhone & Android): Less widely used than I’d expect, but my go-to if I want to both find a coffee shop or shoe store nearby, and find out if it’s any good. If you’re using something like Yelp, or just Apple or Google maps, default to this instead.

HRV4Training (iPhone & Android): Using the camera on your phone, this tracks trends in heart-rate variability, one of the best ways to prevent overtraining. As I’ve written about before, if you work out hard, you should definitely be using this.

The Infatuation (iPhone & Android): While I use Foursquare more generally, if I want to find a good restaurant, The Infatuation is my go-to. Based on at least dozens of field tests, I can say their recommendations and ratings have been 100% correct thus far.

Overcast (iPhone): If you listen to podcasts (and you should), this is an order of magnitude better than Apple’s built-in client in terms of discovery, playlist management, and playback.

Pocket (iPhone & Android): Not hugely obscure, but still not as broadly used as it deserves to be. If you find a longer article / video, use the browser plug-in to save it to Pocket, then read it offline from your phone when you’re in a subway, waiting at the doctor’s office, etc. Exponentially increases the amount of content I can consume.

Soulver (iPhone): Super-intuitive hybrid of a calculator and a spreadsheet. When I need to crunch numbers or brainstorm financial projections on the fly – for business or daily life – Soulver has completely replaced the built-in calculator app.

Todoist (iPhone & Android): Still my daily to-do list (though I use text files for longer-range planning, outlining, tracking, etc., I have a slew of recurring tasks, which are much easier managed when automated). Lightning-fast sync, powerful boolean search / filters, etc.

Ulysses (iPhone): Where I write all blog posts and long-form content on the Mac; I mostly use the mobile app for tweaks and edits. A powerful yet minimalist pair.

Zero (iPhone): The science behind intermittent fasting is pretty impressive, and it’s even more effective if your eating window ends as close to sundown as possible. Zero tracks the length of your daily fasts (I shoot for 16 hours), as well as the number of minutes and hours you eat post-sundown. A great way to cut fat, get healthier, etc.

And, finally, since this is already more than long enough, a bunch of other non-Apple apps I use – most well-known, a few less so. As I’m too lazy to link them all: 1Password, Amazon, Audible, Candy Crush, Caviar, Dropbox, Genius, Gmail, Google Maps, GuitarTuna, Instagram, Kindle, MiniBar, Netflix, Open Table, PayPal, Resy, Runkeeper, Seamless, Seconds Pro, Signal, SimpleNote, Tripit, Twitter, Uber, UberEATS, Washington Post, Zipcar. If you don’t know any of them, worth the Google.

June 29, 2017

Completely Foreign

One of the core ideas underlying Composite is that health is about habits – the small things you do day in and day out that add up over time to meaningful change.

To that end, we’ve been developing an algorithmic approach to acquiring new healthy habits. First, we assess clients’ current habits, in areas like movement, nutrition, and lifestyle. Then we automatically build a prioritized stack of new habits that would be maximally beneficial, introducing them a couple at a time. By monitoring compliance through an app (and egging clients on with accountability to a human coach), we can use learning techniques like spaced repetition to determine when we should be adding new habits, and when we should be doubling down on practicing ones already introduced.

As a result, I’ve spent a ridiculous amount of time looking at other algorithmically-based learning systems. Which is what led me to discover language learning site Lingvist.

Lingvist is the brainchild of a physicist at CERN, who had been living in the French part of Switzerland for years, but had never learned the language. He built the prototype system for himself, studied with it for a few months, and developed enough fluency to test out of the equivalent of a high-school French class.

The approach weighs the importance of words by real-world statistical occurrence, so you spend most of your time focused on the parts of the language you’ll actually use. (I discovered the value of this the hard way, when living in Japan as a high school exchange student; though I could say ‘kindly give me the fish of your brother Yamada,’ I couldn’t say ‘hey guys, I think we just ran out of toilet paper.’)

And it uses an adaptive algorithm: based on your real-time performance, Lingvist alters the pace of learning new words, and the frequency of re-testing old words, even within a single practice session.

I’ve spent some time playing with both the French and Spanish versions, and can confidently say Lingvist is très bon / muy bueno. Try it out yourself.

June 28, 2017

Singularity: Not Quite Yet

My favorite recent update from a friend in the AI world: machine vision algorithms are currently having a terrible time differentiating Goldendoodle puppies and fried chicken.

June 13, 2017

Neutrality vs. the Robots

On Monday, I posted about the importance of net neutrality, and of making your voice heard as the Trump FCC considers rolling back the existing strong enforcement policy.

Fortunately, that’s hardly a minority view, as more than a half million people have weighed in on the FCC’s public comment system to that end. (To reiterate, you should, too: go to gofccyourself.com, click “Express”, then leave a comment supporting “strong oversight of net neutrality based on Title II enforcement.”)

However, about ten percent of the comments have weighed in against net neutrality. And while that might elsewhere be a sign of healthy debate, it’s a bit suspicious that 58,000 of those comments use the exact same clip from a 2010 anti-neutrality press release, with posts cycling in perfect alphabetical order by posters’ names.

According to some crack reporting by ZDnet today, the supposed posters of those comments confirmed that they hadn’t left the comments themselves. Some didn’t even know what net neutrality was.

In other words, this looks like a textbook ‘astroturfing’ bot attack. Given the outsize role of bots during the 2016 election, I hope more media outlets will follow ZDnet’s lead, and give this issue the coverage it deserves. It’s bad enough that powerful internet service provider lobbies egged the FCC into considering scaling back enforcement in the first place; it’s even worse if those same players are resorting to underhanded tactics to try and make it seem like it’s what we, the people, want.

May 10, 2017

Neutral

The internet is, by design, a very robust system. Instead of a hierarchy that can be controlled from a central point, it’s a distributed network. So, in the words of legendary computer scientist (and Electronic Frontier Foundation co-founder) John Gilmore, “the Net interprets censorship as damage and routes around it.”

But there’s one real point of weakness in the internet’s design: the single pipe that connects from the distributed cloud directly to you, the end user. You depend on your mobile provider to connect your phone to the internet, or your cable provider to connect your desktop computer. Which, in turn, gives those providers unique power.

Imagine that the internet is like the island of Manhattan, and you live directly across the Hudson River in Fort Lee, New Jersey. For anything to make it to you from Manhattan, it would have to cross the George Washington Bridge.

Because the bridge is publicly owned, we take for granted that anyone who wants to can drive across. But what if that bridge was privately owned? All of a sudden, the owner of the bridge could start making rules about who could use it. For example, the bridge company could cut an exclusive deal with Domino’s Pizza, and prevent any other pizzeria from delivering to you over the bridge. Goodbye real New York slices, hello doughy circle of crap.

And that’s basically what net neutrality is about. Much like preventing the bridge company from making an exclusive deal with Domino’s, under net neutrality rules, Verizon isn’t allowed to make a deal with Amazon Prime that would then block you from accessing Netflix.

Back in 2000, the FCC put net neutrality rules in place using their Title I enforcement standard. Verizon, in turn, sued the FCC. And the courts ruled that, for the FCC to be able to actually enforce net neutrality, they would need to instead use the stricter Title II standard.

Three years ago, in a giant push that united the internet, consumers commented en masse and convinced the FCC to adopt that stricter Title II net neutrality standard. Victory!

But last month, Trump appointed a former Verizon attorney, Ajit Pai, as the new head of the FCC. And, as Pai said, “net neutrality’s days are numbered.”

Pai has now proposed moving net neutrality back to the looser Title I standard. Which, as the courts have already ruled, the FCC can’t actually enforce.

So, if you care about a free and open internet, about a level playing field in which new companies can compete against rich, entrenched players, now’s the time to act. Americans’ voices convinced the FCC to adopt a strong standard three years ago, and we can keep those protections in place by making our voices heard again today.

To help out, head to gofccyourself.com, and click the ‘express’ link. (Because of the huge wave of support thus far, the FCC’s website appears to be regularly crashing, so you may have to try again later if the site is currently down.)

Then leave a comment saying you want strong oversight of net neutrality based on Title II enforcement.

It takes two minutes, but it could have an immeasurable impact on the future of the Internet. That’s pretty solid ROI.

Again, gofccyourself.com, click express, then “strong oversight of net neutrality based on Title II enforcement.” Get to it.

May 8, 2017

Bootleg

Back in the olden days, when Napster was still a thing, record industry execs spent a whole lot of time and money trying to prosecute people for digitally downloading music. They contended that people were stealing music because they didn’t want to pay for it. But, in retrospect, it’s clear that people were stealing music because that was the only way to get it online. As digital album sales data demonstrate, once they were able to buy music digitally, people flocked to that option in droves.1

During the pre-iTunes Store period, I remember talking with Sean Parker, who compared the online theft of music at the time to bathtub gin. During Prohibition, people couldn’t buy liquor, so they started making it at home. Once Prohibition ended, they could have continued to home-brew inexpensively. Instead, nearly everyone was more than willing to pay for the quality, convenience, and consistency of store-bought brand liquor.

I thought of that again recently, when I came across a table calculating overall internet usage data for last year. Back in 2011, BitTorrent – the primary method for illegally downloading movies – accounted for 23 percent of daily internet traffic in North America, and the movie industry was tearing its hair out with distress about piracy. By last year, BitTorrent traffic was under 5 percent, while (legal, paid) Netflix and Amazon Video have now grown to account for more than 40% of daily traffic.2

In other words: the bathtub gin effect strikes again.

  1. While digital sales never rose to match album levels, that’s primarily a result of unbundling albums into individual songs – people often only want one or two songs from a given album – and moving heavily to a streaming model – which tends to increase consumption without increasing revenue as incremental consumption is free. While both are great for consumers, and less great for record company profits, they’re business model choices made by the industry itself.) ↩︎
  2. I expect things will push even further in that direction once studios give up the practice of ‘windowing’ – delaying the digital release of films until after their full theatrical run. I’ve long contended that a lot of people would be willing to pay fairly high prices (as two movie tickets now closes in on $40, even before marked-up popcorn) to watch new movies at home on the same day that they’re theatrically released. ↩︎

May 3, 2017

Wined Up

A lot of tech-world prognosticators have tagged virtual reality as ‘the next big thing.’ But just as many have pointed out that VR (the ability to interact with a virtual world) will pale in comparison to the sister technology of AR – augmented reality, or the ability to overlay virtual information over the real world.

With a pair of AR-enhanced glasses (or, eventually, contacts) on, you might be able to repair an engine with specs and labels for it digitally overlaid on the metal, or walk through a party with people’s LinkedIn profiles and recent social media updates floating above their heads, like a scene straight out of Super Sad True Love Story.

While that level of interaction is (perhaps fortunately) still a ways off, we’re now seeing some impressive early examples of AR on mobile phones. Consider the Google Translate app, which can translate real-world signs and documents on the fly through your phone’s camera:

Or the venerable SkyView app, which overlays constellations and star and planet names on the night sky:

Both are fun and (at least intermittently) useful, though neither hits as close to home as the Vivino app. The app has long allowed you to scan a wine bottle, to see ratings for it, solving the complete information vacuum represented by most wine stores. But the latest update expands that to an even more socially fraught situation: navigating a restaurant wine list.

Sure, you can fall back to wines you know, or make semi-educated guesses based on varietal and region. Or, with Vivino, you can just point your phone at the list, and find out about the specifics of every wine on it, based on the collective wisdom of the apps 20+ million users:

We may still be a ways off from living in an episode of Black Mirror, but I suspect we’ll be seeing a steady increase of these single-purpose, phone-based AR tools along the way. Cheers to that.

April 28, 2017

Poly-anna

I don’t have a great history of endorsing email apps, as the last two I jumped behind (first Sparrow, then Mailbox) were both acquired and then discontinued pretty much immediately after I plugged them.

Nonetheless, chancing fate, I’d like to once again make an email client recommendation: Polymail.

First, it’s clean and fast.

Second, it integrates a bunch of useful features otherwise only available as separate services: snoozing messages to reappear in the future, per-recipient read notifications on sent messages, the ability to send emails at a scheduled later time, contact profiles with integrated social media / past interactions, etc.

Third, it’s the only client I’ve found that also integrates two of Gmail’s best browser interface features: undo send, and inbox categories.

And fourth and perhaps best of all, it has a surprisingly effective one-click unsubscribe button at the top of any automated email. While most of those emails end up in the aforementioned inbox categories, rather than my primary inbox, I also find my email wrangling is far less stressful if I cut back on the volume of received messages overall. Between news alerts, messages from merchants I’ve bought from in the past, social networking notifications, etc., the amount of ‘bacn’ (i.e., one step up from spam) that I’d been getting daily was fairly mind-boggling. With just a few weeks of liberal unsubscribe button use, I’ve whittled those down by nearly 90%, to the point that I actually read (and want to read) nearly everything that shows up.

So, Polymail. It’s on Mac and iPhone / iPad now, and coming to Android and (for those living on the dark side – I’m looking at you, Ole) Windows shortly. Check it out!

April 25, 2017