Universal language translation: thx, Google

Google is working on improving translation to the point where we have a universal language translator. In the Smithsonian Magazine, Kissing Language Barriers Goodbye:

“One thing that surprises people when we talk about Translate is our team doesn’t have any linguists on it,” Estelle says. “We’ve launched 71 languages, and I would say our team doesn’t know how to speak the vast majority of them. A human translator is not going to be able to learn all these terms and things as fast as our [data] can learn from the web.”

And with the recent release of their Pixel Buds, Google is trying to produce the sci-fi level real-time language translator. From Engadget’s article Google’s Pixel Buds translation will change the world:

Google packed its headphones with the power to translate between 40 languages, literally in real-time. The company has finally done what science fiction and countless Kickstarters have been promising us, but failing to deliver on, for years. This technology could fundamentally change how we communicate across the global community.

Real-time translation can do a lot for breaking down barriers. Engadget continues with the praise:

You’ll be able to walk up to nearly anybody in another country and be able to hold a fluid, natural language conversation without the need for pantomime and large hand gestures, or worry of offending with a mispronunciation. International commerce and communication could become as mundane as making a local phone call.

Universal translation might not be universally good, however. The Smithsonian Magazine article continues:

She thinks the device could be somewhat useful with travel, business and international relations but not groundbreaking. At a certain level, we already have translators (people) in place, and most who work in foreign relations know the appropriate languages. A device, Murphy believes, could have negative consequences.

“I think it can make people lazy,” Murphy says. Translating languages can be mentally challenging by forcing the brain—especially one that knows more than two languages—to work in a different way, but the exercise is rewarding, nonetheless. The brain pulls from a place of linguistic empathy that even the finest voice translator could never reach.

While this universal communication could be a positive, Murphy acknowledges, “it might lead to people thinking they’re communicating when they’re not.” Culture is not always completely embodied in language (take sarcasm, for example), and communication is not always about the information being passed.

Translation is the “key” to unlocking a “truly global” web, but this sort of translation could also lead people to think they’re communicating when they’re not (to paraphrase the article). If you think you’re communicating with someone, but you’re talking past each other in a series of miscommunications, the UX of your world is off, so to speak. This happens in same-language communication as well. With the additional layer of auto-translation interference, it might be easier or harder to detect when such miscommunications happen.

Borders on the Web Series

The web is sometimes spoken of as a borderless place. Through the ~magic~ of technology, the internet, the “decentralized” web, borders would be eliminated and the world would become truly flat.

I’ll share links as part of this series that reinforce and challenge that notion.

  • Linguistic borders, reinforced by lack of multilingualism yet challenged by machine translation successes.
  • Geographic borders, reinforced by internet infrastructure yet challenged by novel methods of providing internet access.
  • Legal borders, reinforced by censorship and international agreements yet challenged by citizens and corporations.
  • National/political borders, reinforced by internet service providers, top level domain names, and governments yet challenged again by citizens and corporations.


Data as a Gift: Implications for Product Design

The idea of data as a gift, and the act of sharing data as an exchange of a gift, has data ethics and privacy implications for product and service design.

Recent work by Kadija Ferryman and Nick Seaver on data as a gift in the last year addressed this concept more broadly and brought it to my attention. Ferryman, in her piece Reframing Data as a Gift, took the angle of data sharing in the context of health data and open data policies. Seaver, in his piece Return of the Gift, approached it from the angle of the gift economy and big data. Both make great points that are relevant in the context of data collection and ethics, especially as it relates to data security and privacy more generally.

Ferryman introduces the concept brilliantly:

What happens when we think about data as a gift? Well, first, we move away from thinking about data in the usual way, as a thing, as a repository of information and begin to think of it as an action. Second, we see that there is an obligation to give back, or reciprocate when data is given. And third, we can imagine that giving a lot of data has the potential to create tension.

When you frame the information that we “voluntarily” share with services as a gift, the dynamics of the exchange shift. We can’t truly share data with digital services—that implies that we retain ultimate ownership over the data. You can take back something after you share it with them. But you can’t do that with your personal data. Because you can’t take back your data after you share it, you can more accurately conceptualize the exchange of data with digital services as a gift. Something you give, and which cannot be returned to you (at least not in its original form).

Data as a gift creates an expectation or obligation for a return, Seaver makes clear. Problem is, when we’re sharing data on the internet, we don’t always know exactly what we’re giving and what we’re getting.

The gift exchange might be based on the expectation that your data is used to provide the service to you. And the more data, the better the service (you might expect). For this reason, it seems easier to share specific types of data with specific services. For example, it’s easier for me to answer questions about my communication or sexual preferences with a company if I think I’m going to get a boyfriend out of the exchange, and sharing that data might make it more likely.

But what happens if a company stops seeing (or doesn’t ever see) an exchange of data as a gift exchange, and starts using the data you gift it for whatever it wants in order to make a profit? By violating the terms of the gift exchange, the company violates the implicit social contract you made with the company when you gifted your data. This is where privacy comes in. Gifting information for one purpose and having it used for other unexpected purposes feels like a violation of privacy. Because it is.

A violation of the gift exchange of data is a privacy violation, but it feels like the norm now. It’s common in terms of services to be informed that after you gift your data to a service, it is no longer yours and the company can do with it what it wants.

Products and services are designed so that you can’t pay for them even if you want to. You must share certain amounts of data, and if you don’t, the product doesn’t work. As Andrew Lewis put it, “If you are not paying for it, you’re not the customer; you’re the product being sold.” We didn’t end up there because we are that dedicated to free things on the Internet. We were lured into gifting our data in exchange for specific, limited services, and the companies realized later that the data was the profitable part of the exchange.

Nick Seaver refers to this as “The obligation to give one’s data in exchange for the use of “free” services,” and it is indeed an obligation. To avoid gifting your data to services that you might not want to enter into that type of exchange, you have very few ways to interact with the modern Internet. You’d likely also have to have a lot of money, in order to enter into a paid transaction rather than a gift exchange with a company in return for services.

For those of us working in product or service development, we can use this perspective and consider the social contract of the exchange of data gifts.

  • Consider whether the service you offer is on par with the amount of data you ask people to gift to you.
    • Do I really need to share my Facebook likes with Tinder to get a superior match?
  • Consider whether the service you offer can deliver on the obligations and expectations created by the gift exchange.
    • Is your service rewarding enough and trustworthy enough to where I’ll save my credit card information?
  • Consider whether you can design your service to allow people to choose the data that they want to gift to you.
    • What is the minimum-possible data gift that a person could exchange with your service, and still feel as though their gift was reciprocated?
  • Consider the type of gift exchange that you design if you force people to gift you a specific type or amount of data.
    • Is that an expectation or obligation that you want to create?

When you view each piece of information that a person shares with you as a gift, it’s harder to misuse that information.


Note: Thanks to Clive Thompson for bringing Kadija Ferryman’s piece to my attention, and Nick Seaver for sharing his piece Return of the Gift with me on Twitter. 

Feature Names Matter

When someone starts using your software, they need to build an understanding of how it works and how the pieces interact. The UI text you write and the feature names you choose can build or break a mental model.

From a marketing perspective, the importance of the name is clear. You want something catchy, marketable, searchable, memorable, all these things. But most importantly, a feature name must help a user build a mental model of what your feature does.

The mental model helps the user understand why they might use this feature, and what for. One of the riskiest part of shipping something new is adoption. If people don’t know what it does or how it works, they won’t use it. A crucial element to that understanding is what you call the new thing and how you describe it in the product. If I can’t guess based on the name what it does, I might not click on it at all and explore it.

Let’s look at some feature names…

  • Google+ vs GoogleDocs. One of these is pretty opaque, and the other is pretty clear. I might think that Google Docs is google FOR docs, but as soon as I click into it, I’ll see what it is and understand that it’s for writing docs. I might never click into Google+ because I have no idea what it is based on the name.
  • Dropbox vs Box. There’s a reason both of these companies are named practically the same thing. Because you put things in boxes that you want to share and store. It’s a super evocative mental model, so it gets a bit overused, perhaps.
  • Slack vs HipChat. HipChat is a bit more descriptive, but you know automatically that it’s a chat app. Slack turns a verb into a noun, and hopes that you start using it and understand that you slack off while using it… kind of.

It’s harder to come up with examples in software of things that truly failed, because they aren’t very well known. But the example that brought this to life for me is from a card game I learned how to play recently. Red7 uses the concept of a “canvas” and a “palette” to tie the metaphor of color across the game. But combining those concepts with the established mental model that you have in a card game with a discard pile and a hand of cards took quite a bit of work. In reality, the clever metaphor broke down and impeded what could have been quick understanding by burdening an existing card game mental model with a mental model of painting ephemera. It was marketable, but not intuitive because it didn’t help people build a mental model to understand how the game works.

The simplest way to pick a good feature name is to test them out. Do some word association exercises with your team, but also with people that don’t work on your team and don’t even work in software. Diverse teams matter a lot in this exercise. This can help identify names that build mental models, break them, or are irrevocably associated with irrelevant mental models.

Another way to pick good feature names is to rely on scenarios when building features. That way, you’re less likely to conceptualize a feature based on its architecture, or your internal team structures, and more likely to think of it from a problem-solving perspective. If you know exactly what the feature is doing, and for whom, it’s easier to pick a useful name.



My impressions from the 2017 Bay Area Book Festival

For the second year in a row I attended the Bay Area Book Festival. A collection of authors, publishers, readers, and other bookish sorts also show up for the weekend in Berkeley. This year, like last year, I discovered some great sessions that led me to think about things from a new perspective and that I might not otherwise have learned about.

One great thing about the book festival is that, true to its slogan, it connects readers and writers. Much like the podcast Song Exploder, the sessions caused me to think far more about the process behind creating the things I love than I might have otherwise.

Something that struck me and my friend, however, is how few of the attendees were young. There are a lot of attendees (of sessions, especially) that seem to be retired, quite a few middle-aged people, and in the YA-fiction sessions I attended, a good number of children and their parents. In fact, much of the festival seemed designed to be family friendly.

However, in a reflection of publishing at large (perhaps) the festival neglects the population that is most clearly present in Berkeley and the Bay Area to me—the ~ * millennials *~ that attend the university and work in tech and other industries all around the bay. I saw a few other people in their mid-twenties and mid-thirties, but not nearly as many as I’d hope to see. No one that I talked to in that age range (save my friend) knew about the festival that weekend, and I saw almost no marketing for it aside from the emails from the organizers (thanks to signing up for the email list last year) in the social media sites and feeds that I follow. Hopefully in future years this will change.

I also struggled to determine who the actual target audience of the festival was based on the sessions I attended and the booths I witnessed wandering around (I’m not much of an actual booth-visitor). The fest again, advertises itself as connecting readers with writers, but it seemed to be connecting writers with other writers just as often. Whether on panel discussions or through writer-oriented booths in the park, that was a very present and understandable theme. I’m not sure it matters that the target of the festival is so broad, and perhaps it might flourish even more if it were more explicitly targeted to this wider audience. But again, perhaps not.

I attended two sessions each day of the festival, and here are my impressions of each of them.


The first session I attended on Saturday was called Worlds We Create: Young Adult Fantasy Writers on Creating Alternate Realities and Memorable Characters. It’s only in the last year or so that I started realizing that I do read fantasy and sci-fi novels, and while I didn’t seek them out when I was younger (sorry, Tamora Pierce and Phillip Pullman), I always enjoyed historical fiction and fairy tale retellings. This session was explicitly about how fantasy writers go about creating the worlds in their novels. The moderator had some great questions, they asked about how authors develop characters to reflect the world they inhabit, whether the story or the characters or the world come first, and more.

They discussed using “sensitivity readers,” a concept I hadn’t heard of before this panel, who you can hire to provide feedback and criticism about the way that you’ve told the story of a character with experiences that you haven’t had directly. One of the authors, who set her latest book in turn of the century New York City, spoke of the balance of being sensitive to certain life experiences and having to confront discrimination with the focus of blowing it up in the book.

Needless to say, I’ll be thinking a lot more about the quality of worldbuilding after this panel. I was in the midst of reading a series that was set in a world that took awhile for me to understand, and seemed to use a glossary to attempt to offload some of the worldbuilding heavy lifting, yet the story and the characters were fairly well-developed enough for me to appreciate it and continue following through. The balance of world-building time and story-telling time was also covered in the panel, with one of the authors mentioning that she’ll use worldbuilding as a way to get started writing a new book if she’s struggling with how to start.

The second session I attended on Saturday was rather different from the YA fiction panel. This session featured Masha Gessen in conversation with Orville Schell on Truth, Lies and Totalitarianism in Russia and the U.S.. This session featured quite a few great moments. My friend pointed out that she hadn’t before heard someone speak with such attention to the precision of their language, yet that is exactly how Gessen spoke. She was very precise when agreeing or disagreeing with Schell, and didn’t mince words.

Some paraphrases of what she said in the session, along with my own characterizations about what she said, follow.

Totalitarianism succeeds through destruction… not durability of the presence of totalitarianism or genetic makeup or expectations but of the absence of shared understanding and society and other things that help you move forward in the face of it.

She took care to distinguish her perspective from that of Schell’s. He asked a question about whether totalitarianism or autocracy persisted in some countries due to the genetic makeup of people or due to a durability of the people, and she disagreed, though recognized that they agreed on the state but not the cause of the state. The absence of certain things, rather than the presence of something, is what is helping totalitarianism succeed and persist in these nations.

We’ve constructed a story about what happened in post-soviet states but not what happened in the west… democracy isn’t a state achieved by a country but a spectrum (becoming more or less democratic)

I really appreciated that she mentioned this. As someone who studied post-soviet states in college, I am quite familiar with the stories constructed there, but I hadn’t considered this perspective until she mentioned it. In my recent travels and events in the United States, it’s become clear that the Western states didn’t pay much attention to developing an identity that didn’t position other states as enemies and actively maintaining democracy.

Gessen also mentioned, perhaps in response to an audience question, why Trump admired Putin so much. In response to that, she referenced Tim Snyder in the NYRB, who wrote “Putin is the person that Trump plays on TV.” Expanding on that point, she continued that “Trump talks of power, off the charts popularity, and control… but doesn’t really understand where that comes from.” And where that comes from is autocracy. They also discussed her own article in the NYRB about autocracy.

Also discussed was the differing perspectives of politicians, businessmen, and others in an autocratic state compared with a democratic state.

There’s this belief from Russian politicians, journalists, businessmen that everything can be bought or is for sale or is transactional… a few years ago that didn’t fly here but that’s changing here… apparent struggle to understand non-transactional motives in authoritarian societies and possible explanation for a lack of understanding of civil society… If public service and nonprofit motives are questioned or not understood then democracy is challenged

Schell also brought up countries that don’t deal with history in an honest way (like many totalitarian states, Russia or China especially have quite censored and limited views of their own past), she spoke of WWII as being a touchstone in Soviet history in rewriting the past, but also continued to discuss whether a society can move forward with a distorted view of the past.

All societies have distorted views of the past…that’s trauma that societies have to deal with. Estonia for example has done a good job with this: “we trust one another, we were under occupation, and we’re going to rebuild based on that trust.”

Germany had the advantage of short period of time and everyone could remember pre-totalitarianism and talk about perpetrators, bystanders, and victims…. but it’s totally muddled in the Soviet Union; people were both perpetrators and victims and no one was a bystander.

How do you take responsibility for your own past if you messed it up? Whitewash the history…

One of the things that she mentioned as being most challenging for rebuilding in (or after) a totalitarian or autocratic state was imagination.

How do you rebuild when you can’t imagine a future or a past…and sometimes you can’t imagine the present because it’s so awful that you can’t fathom it…you reach for the past and confuse lack of imagination with understanding

When asked about her hopes for Russia in the future, she was blunt. “Russia is hopeless, scorched earth.” She did expand in response to a follow-up question to discuss the small-scale successes of some friends of hers in starting focused community organizations, but acknowledged that people seeking to make change on a small scale, while inspiring, is a way of admitting defeat. After you start focusing almost exclusively on small-scale change, you’re no longer trying or dreaming of making those changes on a large scale.


Sunday I slept in, so I skipped an earlier morning session I considered attending so again attended two sessions.

First I attended another YA fiction session, Reality Bites: Fiction About Teens’ Real Lives, which was really great. Another panel discussion, the authors discussed what they do when they write most of a book and realize it isn’t working, sometimes all you can do is “upcycle” some of the scenes into other books. The moderator also asked about recurring themes and iconography, and while one author mentioned trees (she writes largely thrillers that tend to involve the woods), another author mentioned that she often writes about the sky, likely because it represents us against the vastness.

Most relevant of all, I thought, was the discussion on how writing novels teaches you to be compassionate about multiple points of view. Kim Culbertson especially made some great points about this. Sometimes a character has to do something differently than the way you would. She mentioned that “Lots of people are uncomfortable when they hold themselves up to another person and the edges don’t match.” But she likes to challenge that in a way, to help people get to the understanding and consideration that other people are different from you and that may seem strange but it’s actually awesome, and we need to recognize that!



The second panel I attended was about immigration and identity, Living in Two Worlds: Crossing Borders and Identities to Create Home. The moderator asked each author in turn to speak about their book which led to a great focused discussion on the themes in each one.

First, Laleh Khadivi discussed her book and the challenge about writing about two worlds, especially that “one world is never going to believe the other world.”

Next, Lesley Nneka Arimah discussed her book of short stories. She mentioned that she included Facebook in one of her stories because she was tired of reading stories set in contemporary Africa that didn’t mention social media. She underscored that while traditional infrastructure in Nigeria and other countries may be lacking, everyone has a cell phone and electronic communication is crucial to how they live their lives. It seemed out of touch not to acknowledge that reality of life in her fiction.


Carolina de Robertis spoke next about her novel, but my favorite quote by her came from the Q&A after the moderated panel concluded, which was that “Culture is not static; we can shift culture and push it open with the narratives that we live and write.”

Pajtim Statovci spoke last about his novel, and addressing the fact that living through the trauma of immigration, especially as a refugee, doesn’t automatically make you stronger.

Sometimes our negative experiences don’t make us stronger. Sometimes they make us weaker and sad and pathetic.

That resonated with me, because it’s easy to look at those who have survived something traumatic and assign strength to them merely because they survived. But that isn’t always the case. This thread resonated with the other books on the panel as well, because the main character in Khadivi’s book is an immigrant who radicalizes after immigration. Lesley Nneka Arimah mentioned about wanting to write about flawed characters as well, in the vein that we often think of ourselves as better than we are, so she wanted to write about characters that do the wrong thing in service of finding their own truth.

Overall impressions

The festival was worth my time to attend, and spend time thinking and talking about books with and around other book-minded people. I look forward to the next year’s book festival and the sessions! It’s always easy to attend when it’s convenient to get to and the weather is beautiful. Now I’m inspired to tackle even more of the books on my shelves…


Tips for live tweeting an event

If you use Twitter and are attending an event that you want to share with your twitter followers, you can live tweet it as it’s happening. While you can live tweet basically any event, these tips focus mainly on talks that you might attend as part of a conference, a meetup, a sponsored speaker series, or another presentation.

I’ve live tweeted several conferences (two as part of a job, such as #SUMIT14), talks, and series of talks as @smorewithface.

First, the basics on live tweeting an event, then some pro tips and best practices to follow before and during the event.

Continue reading

2015 Resolutions, 2016 Music, and 2017

In 2015 I made some resolutions. I haven’t followed up on them since. In 2016 I made no resolutions, but I listened to a lot of music.

How the 2015 resolutions fared in 2015 and 2016

I did okay.

1. Stay off Twitter more, read fewer articles on the web, and create more.

I’ve continued to use Twitter over the past couple years. My use of it waxes and wanes depending on the news. I periodically delete it from my phone to get a break. My Pocket queue and a change in my commute mean that I’ve certainly shifted my web-based reading habits. I read 73 books in 2016 (that I bothered to add to Goodreads, so let’s round to 75). I’ve taken a few vacations without my laptop, and have spent a bit more time working with my hands—whether at the gym, climbing, or making things like jewelry.

2. Learn JavaScript

I still have not learned JavaScript. I gave up on it. Part of this is because semicolons are rude, and part of this is because I joined a company where Python is the primary backend language. I realized Python would be easier to learn and my interests shifted more toward data analysis (and maybe some digital mapping?) and further from interactive web content, so I worked on learning Python instead. I ran into the same issue learning Python as I did JS though… it requires a lot of time, and a lot of practice. Side projects are hard to maintain, especially when they’re similar to your day job. I still have some tabs open about learning Python, and I took more Python tutorials, so I have at least reading-level knowledge of both languages.

3. Read something huge, and write something huge.

Still haven’t read Gödel Escher Bach: An Eternal Golden Braid. Still haven’t written up a magnum opus on the reification of geographical and political and linguistic borders on the web. I’m hoping to break it apart and start publishing bits and pieces here this year. We’ll see.

Goals for 2017

I’m meeting with friends this weekend to work through some. I’m setting some in terms of reading material, to better improve my knowledge of key (also depressing) moments in history. I can’t be a history nerd without being fascinated by dictators, oppression, and systemic discrimination. So I’m branching out to different geographical dictators and oppression in history this year. Beyond that, we’ll see what goals I end up with.

2016 in Music

2016 was a music-filled year, after a multi-year hiatus of drifting away from it being as central in my life as it deserves. These songs stuck with me for various reasons. They helped me regain some hipster credibility, branch out into less-explored genres, and reminded me how important it is to support artists you care about.

Songs that stuck

This year I bought the CHVRCHES album, the Bon Iver album, the Frank Ocean album, and The Little Prince soundtrack. In addition to a bunch of one-off songs, because shuffle is how I roll.

I listened to Jason Derulo a lot, and listened to a four song playlist of Carly Rae Jepsen, Ariana Grande, Ingrid Michaelson, and Adele for longer than I’d like to admit.

Shows I saw

I saw the following artists play live:

  • Justice and Sebastian (he might still go by SebastiAn)
  • Broods and Two Door Cinema Club
  • Daughter
  • Still Flyin’, Annie Hart of Au Revoir Simone, and Slow Club
  • Tallest Man on Earth and The Head and the Heart
  • Cold War Kids

If I had to rank them, CHVRCHES would be first.

I started the new year by binging on SoundCloud recommendations from The Burning Ear and the related tracks that SoundCloud plays. A great way to fall into a rabbit hole of discovery, and a totally different experience from Spotify’s more carefully-constructed and curated experience of the Discover Weekly and other playlists.

Stats that were gathered for me, passively by last.fm

Last.FM tells me things about last year in music too.


My listening increased after I moved. CHVRCHES was a continual favorite. I listened to a lot of different types of music, but mostly stuck to indie. Hey Rosetta! – Trish’s Song is a great song to listen to if you’re trying to fall asleep.

2017 in Music

In 2017 I have tickets to see these artists in concert…

  • Less Than Jake
  • Mike Doughty and Wheatus
  • Jens Lekman
  • Radical Face
  • Matt Pond PA
  • Gibbz
  • Knox Hamilton and Colony House

And that’s just the first three months. City life suits me. Forging ahead into 2017 suits me. Here’s to more reading, more music, more learning, more blogging, and more self care in the year ahead.

Netflix, Amazon, and Hulu: Watching and Sharing on the Web

I like to share with my friends. I’ll send them links to articles in Pocket, book recommendations on Goodreads, and music recommendations in Spotify. I’ll also post SoundCloud songs on Facebook, quote articles on Twitter, and email other articles to friends. All of these sharing methods augment word-of-mouth, letting me seamlessly share my experiences with others within or outside of an application.

What’s missing is the ability to easily share recommendations in the apps we use to watch TV and movies online. The word-of-mouth augmentation that social media provides is pretty minimal when it comes to TV and movies.

Continue reading

Physical technology and inherited meeting places

Digital technology is central in our social and personal lives. Laptops, smartphones, and the apps installed on them allow us to communicate more frequently & at greater distances than ever before. This capability is reshaping those communications. This centrality of the role of tech mirrors the importance that churches once had in peoples’ lives. The Internet Archive could be aiming to bring this back, in a more physical manner.

Continue reading

The Politics of Crisis Communication

Some mentions in Zeynep Tufekci’s post The Politics of Empathy and the Politics of Technology got me thinking about crisis communications for incident response.

Facebook will have to decide which incidents are “serious and tragic” versus which ones are “ongoing crises” where Safety Check would not be useful. Iraq is not officially at war, but suicide bombings there are almost horrifically routine. Their new policy raises many important questions that should be carefully considered. Will Baghdad bombings be considered endemic? How many in a year to declare something endemic or chronic? Are we just acknowledging that people in the regions of the world suffering from chronic crises have no way to feel “safe”? Who gets to check in? “Useful” as defined as useful to whom? Would you not want a “Safety Check” everyday if your loved one were trapped in a region with a dangerous and fast-moving epidemic like Ebola?

When your platform takes the role of defining a crisis, that’s putting the media in social media.  By enabling Safety Check for “unnatural” disasters, it stops being a nifty feature and becomes an essential tool for communication during a crisis.

Activating Safety Check constantly would lessen its value as a signal. Right now, it functions as a forceful push, and you get a “notify” on your phone when a friend in the affected area checks in as safe. Getting hundreds of these notifications per day would reduce its efficacy. However, not getting the notification when you were worried about someone would also be a problem. This type of system requires decisions to be made about when to activate, and when to hold back.

Avoiding alert fatigue is key when it comes to crisis communications. Facebook is both defining the crisis and communicating about it.  It’s easy to send a message to millions of people when they log into your service on a daily (or even hourly) basis. It isn’t easy to do it well.