Digital Life

You are currently browsing the archive for the Digital Life category.

Facial recognition by machines is out of control. Meaning our control. As individuals, and as a society.

Thanks to ubiquitous surveillance systems, including the ones in our own phones, we can no longer assume we are anonymous in public places or private in private ones.

This became especially clear a few weeks ago when Kashmir Hill (@kashhill) reported in the New York Times that a company called Clearview.ai “invented a tool that could end your ability to walk down the street anonymously, and provided it to hundreds of law enforcement agencies, ranging from local cops in Florida to the F.B.I. and the Department of Homeland Security.”

If your face has ever appeared anywhere online, it’s a sure bet to assume that you are not faceless to any of these systems. Clearview, Kashmir says, has “a database of more than three billion images” from “Facebook, YouTube, Venmo and millions of other websites ” and “goes far beyond anything ever constructed by the United States government or Silicon Valley giants.”

Among law enforcement communities, only New Jersey’s has started to back off on using Clearview.

Worse, Clearview is just one company. Laws also take years to catch up with developments in facial recognition, or to get ahead of them, if they ever can. And let’s face it: government interests are highly conflicted here. The need for law enforcement and intelligence agencies’ need to know all they can is at extreme odds with our need, as human beings, to assume we enjoy at least some freedom from being known by God-knows-what, everywhere we go.

Personal privacy is the heart of civilized life, and beats strongest in democratic societies. It’s not up for “debate” between companies and governments, or political factions. Loss of privacy is a problem that affects each of us, and calls fo0r action by each of us as well.

A generation ago, when the Internet was still new to us, four guys (one of which was me) nailed a document called The Cluetrain Manifesto to a door on the Web. It said,

we are not seats or eyeballs or end users or consumers. we are human beings and our reach exceeds your grasp. deal with it.

Since then their grasp has exceeded our reach. And with facial recognition they have gone too far.

Enough.

Now it’s time for our reach to exceed their grasp.

Now it’s time, finally, to make them deal with it.

I see three ways, so far. I’m sure ya’ll will think of other and better ones. The Internet is good for that.

First is to use an image like the one above (preferably with a better design) as your avatar, favicon, or other facial expression. (Like I just did for @dsearls on Twitter.) Here’s a favicon we can all use until a better one comes along:

Second, sign the Stop facial recognition by surveillance systems petition I just put up at that link. Two hashtags:

  • #GOOMF, for Get Out Of My Face
  • #Faceless

Third is to stop blaming and complaining. That’s too easy, tends to go nowhere and wastes energy. Instead,

Fourth, develop useful and constructive ideas toward what we can do—each of us, alone and together—to secure, protect and signal our privacy needs and intentions in the world, in ways others can recognize and respect. We have those in the natural world. We don’t yet in the digital one. So let’s invent them.

Fifth is to develop the policies we need to stop the spread of privacy-violating technologies and practices, and to foster development of technologies that enlarge our agency in the digital world—and not just to address the wrongs being committed against us. (Which is all most privacy laws actually do.)

 

 

Tags: , , , ,

A few days ago a Twitter exchange contained an “OK Boomer” response to one of my tweets. At the time I laughed it off, tweeting back a pointer to Report: Burying, Cremating Baby Boomers To Generate $200 Trillion In GDP, which ran five years ago in The Onion.

But it got me thinking that “OK Boomer” might be more—and worse—than a mere meme. Still, I wasn’t moved to say anything, because I had better stuff to do.

Then today I followed a link to Not So OK, Boomers, on Pulp. Illustrated by Goya’s horrifying Saturn Devouring His Son, it ends with this:

Goya’s Saturn does not swallow his children whole, but has taken chunks out of the body, chewing off the head and the limbs.

The cannibalism Boomers are inflicting on us appears to be closer to Goya’s vision: deranged, irreversible, and violent. Unwilling to accept a world that goes on without them, they are gluttonously consuming resources.

Their own lives have been extended, but without any appreciable gains in quality of life, and so in their rage, their confusion, they poison the air and water, they raise our cortisol levels.

What do we do with the knowledge that our parents are actively trying to harm us but are incapable of accepting the suffering they’re inflicting?

Our response is going to have to be better than depression memes and the odd glib, ‘OK Boomer,’ if we’re going to survive.

That got 2,200 claps. So far.

So this time I responded, with this:

I like Pulp, perhaps because I’ve been young a long time. But this piece is worse than wrong. It’s cruel and inflammatory.

To see how, answer this: Is there moral difference between prejudice against a race, a gender, an ethnicity, a nationality—and a demographic? If there is, it’s one of degree, not one of kind.

As soon as you otherize any human category as a them vs. an us, you’re practicing the same kind of prejudice—and, at its worst, bigotry.

Try substituting the words “women” or “blacks” for the word “Boomers” in this piece, and you get the point.

Ageism may not be worse than sexism or racism, but it’s still an ism, good only for amplifying itself, which seems to be the purpose of this piece.

Read the closing paragraphs again and ask what kind of action the author calls for that would be proportional to the cannibalism he accuses Boomers of inflicting on his generation.

And then hope it doesn’t happen.

That got 10 claps. So far.

But what the hell, I’ll continue.

If young people want to understand old people (which Boomers are, or soon will be) I suggest this: imagine that if a fifth, then a quarter, then a third, then half, and then most of the people you grew up or worked with in your life—friends, cousins, co-workers, classmates—are now dead. And that meanwhile you’re putting your useful experience and wisdom to work as best you can while you’re still able, knowing that, too soon at any age, you’ll be gone too.

There’s no shit you can give a person like that, sitting on the short end of life’s death row, that can measure up to their intimate familiarity with mortality, and with the work they still face, most of which they’ll never finish. So an “OK Boomer” put-down isn’t going to bother most of them.

But it’s still shit. Or worse.

We can all do better than that.

 

 

A Route of Evanescence,
With a revolving Wheel –
A Resonance of Emerald
A Rush of Cochineal –
And every Blossom on the Bush
Adjusts it’s tumbled Head –
The Mail from Tunis – probably,
An easy Morning’s Ride –

—Emily Dickinson
(via The Poetry Foundation)

While that poem is apparently about a hummingbird, it’s the one that comes first to my mind when I contemplate the form of evanescence that’s rooted in the nature of the Internet, where all of us are here right now, as I’m writing and you’re reading this.

Because, let’s face it: the Internet is no more about anything “on” it than air is about noise, speech or anything at all. Like air, sunlight, gravity and other useful graces of nature, the Internet is good for whatever can be done with it.

Same with the Web. While the Web was born as a way to share documents at a distance (via the Internet), it was never a library, even though we borrowed the language of real estate and publishing (domains and sites with pages one could author, edit, publish, syndicate, visit and browse) to describe it. While the metaphorical framing in all those words suggests durability and permanence, they belie the inherently evanescent nature of all we call content.

Think about the words memorystorageupload, and download. All suggest that content in digital form has substance at least resembling the physical kind. But it doesn’t. It’s a representation, in a pattern of ones and zeros, recorded on a medium for as long the responsible party wishes to keep it there, or the medium survives. All those states are volatile, and none guarantee that those ones and zeroes will last.

I’ve been producing digital content for the Web since the early 90s, and for much of that time I was lulled into thinking of the digital tech as something at least possibly permanent. But then my son Allen pointed out a distinction between the static Web of purposefully durable content and what he called the live Web. That was in 2003, when blogs were just beginning to become a thing. Since then the live Web has become the main Web, and people have come to see content as writing or projections on a World Wide Whiteboard. Tweets, shares, shots and posts are mostly of momentary value. Snapchat succeeded as a whiteboard where people could share “moments” that erased themselves after one view. (It does much more now, but evanescence remains its root.)

But, being both (relatively) old and (seriously) old-school about saving stuff that matters, I’ve been especially concerned with how we can archive, curate and preserve as much as possible of what’s produced for the digital world.

Last week, for example, I was involved in the effort to return Linux Journal to the Web’s shelves. (The magazine and site, which lived from April 1994 to August 2019, was briefly down, and with it all my own writing there, going back to 1996. That corpus is about a third of my writing in the published world.) Earlier, when it looked like Flickr might go down, I worried aloud about what would become of my many-dozen-thousand photos there. SmugMug saved it (Yay!); but there is no guarantee that any Website will persist forever, in any form. In fact, the way to bet is on the mortality of everything there. (Perspective: earlier today, over at doc.blog, I posted a brief think piece about the mortality of our planet, and the youth of the Universe.)

But the evanescent nature of digital memory shouldn’t stop us from thinking about how to take better care of what of the Net and the Web we wish to see remembered for the world. This is why it’s good to be in conversation on the topic with Brewster Kahle (of archive.org), Dave Winer and other like-minded folk. I welcome your thoughts as well.

Go to the Alan Turing Institute. If it’s a first time for you, a popover will appear:

Among the many important things the Turing Institute is doing for us right now is highlighting with that notice exactly what’s wrong with the cookie system for remembering choices, and lack of them, for each of us using the Web.

As the notice points out, the site uses “necessary cookies,” “analytics cookies” (defaulted to On, in case you can’t tell from the design of that switch), and (below that) “social cookies.” Most importantly, it does not use cookies meant to track you for advertising purposes. They should brag on that one.

What these switches highlight is that the memory of your choices is theirs, not yours. The whole cookie system outsources your memory of cookie choices to the sites and services of the world. While the cookies themselves can be found somewhere deep in the innards of your computer, you have little or no knowledge of what they are or what they mean, and there are thousands of those in there already.

And yes, we do have browsers that protect us in various ways from unwelcome cookies, but they all do that differently, and none in standard ways that give us clear controls over how we deal with sites and how sites deal with us.

One way to start thinking about this is as a need for cookies go the other way:

I wrote about that last year at Linux Journal in a post by that title. A nice hack called Global Consent Manager does that.

Another way is to think (and work toward getting the sites and services of the world to agree to our terms, and to have standard ways of recording that, on our side rather than theirs. Work on that is proceeding at Customer Commons, the IEEE, various Kantara initiatives and the Me2B Alliance.

Then we will need a dashboard, a cockpit (or the metaphor of your choice) through which we can see and control what’s going on as we move about the Web. This will give us personal scale that we should have had on Day One (specifically, in 1995, when graphical browsers took off). This too should be standardized.

There can be no solution that starts on the sites’ side. None. That’s a fail that in effect gives us a different browser for every site we visit. We need solutions of our own. Personal ones. Global ones. Ones with personal scale. It’s the only way.

“What’s the story?”

No question is asked more often by editors in newsrooms than that one. And for good reason: that’s what news is about: stories.

I was just 22 when I got my first gig as a journalist, reporting for a daily newspaper in New Jersey. It was there that I first learned that all stories are built around just three elements:

  1. Character
  2. Problem
  3. Movement toward resolution

You need all three. Subtract one or more, and all you have is an item, or an incident. Not a story. So let’s unpack those a bit.

The character can be a person, a group, a team, a cause—anything with a noun. Mainly the character needs to be worth caring about in some way. You can love the character, hate it (or him, or her or whatever). Mainly you have to care about the character enough to be interested.

The problem can be of any kind at all, so long as it causes conflict involving the character. All that matters is that the conflict keeps going, at least toward the possibility of resolution. If the conflict ends, the story is over. For example, if you’re at a sports event, and your team is up (or down) by forty points with five minutes left, the character you now care about is your own ass, and your problem is getting it out of the parking lot. If that struggle turns out to be interesting, it might be a story you tell later at a bar.)

Movement toward resolution is nothing more than that. Bear in mind that many stories, and many characters in many conflicts around many problems in stories, never arrive at a conclusion. In fact, that may be part of the story itself. Soap operas work that way.

For a case-in-point of how this can go very wrong, we have the character now serving as President of the United States.

Set the politics aside and just look at the dude through the prism of Story.

Trump—clearly, deeply and instinctively—understands how stories work. He is experienced and skilled at finding or causing problems that generate conflict and enlarge his own character in the process.

He does this through constant characterization of others, for example with nicknames: “Little Mario,” “Low Energy Jeb,” “Crooked Hillary,” “Sleepy Joe,” “Failing New York Times.”

He stokes the fires of conflict by staying on the attack at all times: a strategy he learned from Roy Cohn, who Frank Rich felicitously calls “The worst human being who ever lived … the most evil, twisted, vicious bastard ever to snort coke at Studio 54.” Talk about character. Whoa. As Politico puts it here, “Cohn imparted an M.O. that’s been on searing display throughout Trump’s ascent, his divisive, captivating campaign, and his fraught, unprecedented presidency. Deflect and distract, never give in, never admit fault, lie and attack, lie and attack, publicity no matter what, win no matter what, all underpinned by a deep, prove-me-wrong belief in the power of chaos and fear.” There is genius to how Trump succeeds at this, especially in these early years of our new digital age, when the entire Internet is one big gossip mill: an ideal environment for what the publicizers call earned media coverage. Meaning the free kind.

Trump’s success at earning the attention of everyone and the loyalty of millions calls to mind The Mule in Isaac Azimov’s Foundation and Empire. (It was by noting this resemblance that I, along with Scott Adams, expected Trump to win in 2016.)

So there we have Fail #1: Journalism’s appetite for stories proves a weakness when it’s fed by a genius at hogging the stage. Fail #2 is that Journalism avoids reporting what doesn’t fit the story format. This includes most of reality.

Alas for journalism, stories are inadequate ways to represent facts and truths. Too much of both get excluded if they don’t fit “the narrative,” which is the modern way to talk about story—and to spin journalists. (My hairs of suspicion stand on end every time I hear the word “narrative.”)

There is a paradox here: We need to know more than stories can tell, yet stories are pretty much all human beings are interested in. Character, problem and movement give shape and purpose to every human life. We can’t correct for it.

That’s why my topic here—a deep and abiding flaw (also a feature) of both journalism and human nature—is one most journalists won’t touch. The flawed nature of The Story itself is not a story. Same goes for  “earned media coverage.” Both are features rather than bugs, because debugging them has proven impossible.

I’ll illustrate my point with the killing fields of Cambodia. Those fields are the setting for a story well-known today, about how Pol Pot and the Khmer Rouge committed genocide on a massive scale, wiping out between one and a half to two million people, or around twenty-five percent of the country’s population.

And yet that story meant close to nothing until it became about one man with a problem rather than a whole population whose lives had ended.

A measure of that close-to-nothing was delivered one evening in the late 1970s, by Hughes Rudd, an anchor at the time of the CBS Morning News. He said, almost offhandedly, that there were reports coming in, saying that perhaps half a million people were dead in Cambodia. Rather than a story, this was just an item: too important to not mention but not interesting enough to say more about. The next morning I checked The New York Times and found the same item mentioned in a short piece on an inside page. It blew my mind: half a million dead, and no story.

What made it not a story was the absence of all three elements. There were no characters, no conflict that was easy to describe, no movement toward resolution. Just a statistic. It hardly mattered to journalistic institutions of the time that the statistic itself was a massive one.

The killing fields finally became a story on January 20, 1980, when Sydney Schanberg‘s The Death and Life of Dith Pran ran in the Times‘ Sunday Magazine. Now the story had all three elements, and pulled in lots of relevant and interesting facts. Eventually it became the movie that gave Cambodia’s killing fields their name.

For journalism, however, what also matters about this is that years went by, with a million or more people dying, before the killing fields became a big story.

And this wasn’t the first or last time that massively important and consequential facts got too little attention in the absence of one or more of a story’s three elements. Consider The Holocaust (six million dead) vs. the story of Ann Frank. The Rwandan genocide vs. Hotel Rwanda. China’s one child policy (untold millions of full-term fetuses aborted or born babies killed or left beside the road to die) vs. One Child Nation. The Rohingya conflict (more than 10,000 civilians dead, 128,000 internally displaced, 950,000+ fled elsewhere) vs. approximately nobody. Heard of Holodomor? How about any of the millions who died during Mao’s revolution in China?

Without characters to care about, or a conflict to focus interest, or movement toward resolution, you mostly just have statistics which become cemeteries of facts. Sure, some of it will be studied by academics and obsessives of other kinds (including journalists who care about the topics and publish what they learn wherever they can). But Big-J journalism will mostly be preoccupied elsewhere, by more interesting stuff. Like it is right now.

_________

*However, if you want good advice on how best to write stories about the guy, you can’t beat what @JayRosen_NYU tweets here. I suggest it also applies to the UK’s new prime minister.

The answer is, we don’t know. Also, we may never know, because—

  • It’s too hard to measure (especially if you’re talking about the entire Net)
  • Too so much of the usage is in mobile devices that vary enormously
  • The browser makers are approaching ad blocking and tracking protection in different and new ways that change frequently, and the same goes for ad-blocking and tracking-protecting extensions and add-ons. One of them (Adblock Plus) is actually in the advertising business (which Wikipedia politely calls ad filtering)
  • Some of the most easily sourced measures are surveys, yet what people say and what they do are very different things
  • Some of the most widely cited findings are from sources with conflicted interests (for example, selling anti-ad-blocking services), or which aggregate multiple sources that aren’t revealed when cited
  • Actors good and bad in the ecosystem that ad blocking addresses also contribute to the fudge

But let’s explore a bit anyway, working with what we’ve got, flawed though much of it may be. If you’re a tl;dr kind of reader, jump down to the conclusions at the end.

Part 1: ClarityRay and Pagefair

Between 2012 and 2017, the most widely cited ad blocking reports were by ClarityRay and PageFair, in that order. There are no links to ClarityRay’s 2012 report, which I cited here in 2013. PageFair links to their 2015, 2016 (mobile) and 2017 reports are still live. The company also said last November that it was at work on another report. This was after PageFair was acquired by Blockthrough (“the leading adblock recovery program”). A PageFair blog post explains it.

I placed a lot of trust in PageFair’s work, mostly because I respected Dr. Johnny Ryan (@JohnnyRyan), who left PageFair for Brave in 2018. I also like what I know about Matthew Cortland, who was also at PageFair, and may still be. Far as I know, he hasn’t written anything about ad blocking research (but maybe I’ve missed it) since 2017.

Here are the main findings from PageFair’s 2017 report:

  • 615 million devices now use adblock
  • 11% of the global internet population is blocking ads on the web

Part 2: GlobalWebIndex

In January 2016, GlobalWebIndex said “37% of mobile users … say they’ve blocked ads on their mobile within the last month.” I put that together with Statista’s 2017 claim that there were then more than 4.6 billion mobile phone users in the world, which suggested that 1.7 billion people were blocking ads by that time.

Now GlobalWebIndex‘s Global Ad-Blocking Behavior report says 47% of us are blocking ads now. It also says, “As a younger and more engaged audience, ad-blockers also are much more likely to be paying subscribers and consumers. Ad-free premium services are especially attractive.” Which is pretty close to Don Marti‘s long-standing claim that readers who protect their privacy are more valuable than readers who don’t.

To get a total ad blocking population from that 47%, one possible source to cite is Internet World Stats:

Note that Internet World Stats appears to be a product of the Miniwatts Marketing Group, whose website is currently a blank WordPress placeholder. But, to be modest about it, their number is lower than Statista’s from 2016: “In 2019 the number of mobile phone users is forecast to reach 4.68 billion.” So let’s run with the lower one, at least for now.

Okay, so if 47% of us are using ad blockers, and Internet World Stats says there were 4,312,982,270 Internet users by the end of last year (that’s mighty precise!), the combined numbers suggest that more than 2,027,101,667 people are now blocking ads worldwide. So, we might generalize, more than two billion people are blocking ads today. Hence the headline above.

Perspective: back in 2015, we were already calling ad blocking The biggest boycott in human history. And that was when the number was just “approaching 200 million.”

More interesting to me is GlobalWebIndex’s breakouts of listed reasons why the people surveyed blocked ads. Three in particular stand out:

  • Ads contain viruses or bugs, 38%
  • Ads might compromise my online privacy, 26%
  • Stop ads being personalized, 22%

The problem here, as I said in the list up top, is that these are measured behaviors. They are sympathies. But they’re still significant, because sympathies sell. That means there are markets here. Opportunities to align incentives.

Part 3: Ad Fraud Researcher

I rely a great deal on Dr. Augustine Fou (@acfou), aka Independent Ad Fraud Researcher, to think and work more deeply and knowingly than I’ve done so far here (or may ever do).

Looking at Part 2 above (in an earlier version of this post), he tweeted, “I dispute these findings. ASKING people if they used an ad blocker in the past month is COMPLETELY inaccurate and inconsistent with people who ACTUALLY USE ad blockers regularly.” Also, “Source: GlobalWebIndex Q3 2018 Base: 93,803 internet users aged 16-64, among which were 42,078 respondents who have used an ad-blocker in the past month”. Then, “Are you going to take numbers extrapolated from 42,078 respondents and extrapolate that to the entire world? that would NOT be OK.” And, “Desktop ad blocking in the U.S. measured directly on sites which humans visit is in the 8 – 19% range. Bots must also be scrubbed because bots do not block ads and will skew ad blocking rates lower, if not removed.”

On that last tweet he points to his own research, published this month.There is lots of data in there, all of it interesting and unbiased. Then he adds, “your point about this being the ‘biggest boycott in human history’ is still valid. But the numbers from that ad blocking study should not be used.”

Part 4: Comscore

Among the many helpful tweets in response to the first draft of this post was this one by Zubair Shafiq (@zubair_shafiq), Assistant Professor of Computer Science at the University of Iowa, where he researches computer networks, security, and privacy. His tweet points to Ad Blockers: Global Prevalence and Impact, by Matthew Malloy, Mark McNamara, Aaron Cahn and Paul Barford, from 2016. Here is one chart among many in the report:

The jive in the Geo row is explained at that link. A degree in statistics will help.

Part 5: Statista

Statista seems serious, but Ad blocking user penetration rate in the United States from 2014 to 2020 is behind a paywall. Still, they do expose this hunk of text: “The statistic presents data on ad blocking user penetration rate in the United States from 2014 to 2020. It was found that 25.2 percent of U.S. internet users blocked ads on their connected devices in 2018. This figure is projected to grow to 27.5 percent in 2020.”

Provisional Conclusions

  1. The number is huge, but we don’t know how huge.
  2. Express doubt about any one large conclusion. Augustine Fou cautions me (and all of us) to look at where the data comes from, why it’s used, and how. In the case of Statista, for example, the data is aggregated from other sources. They don’t do the research themselves. It’s also almost too easy to copy and paste (as I’ve done here) images that might themselves be misleading. The landmark book on misleading statistics—no less relevant today than when it was written in 1954 (and perhaps more relevant than ever)—is How to Lie With Statistics.
  3. Everything is changing. For example, browsers are starting to obsolesce the roles played by ad blocking and tracking protection extensions and add-ons. Brave is the early leader, IMHO. Safari, Firefox and even Chrome are all making moves in this direction. Also check out Ghostery’s Cliqz. For some perspective on how long this is taking, take a look at what I was calling for way back in 2015.
  4. The market is sending a massive message.  That message is that advertising online has come to have massively negative value. Ad blocking and tracking protection are legitimate and eloquent messages from demand to supply. By fighting that message, marketing is crapping on most obvious and gigantic clue it has ever seen. And the supply side of the market isn’t just marketers selling stuff. It’s developers who need to start working for the hundreds of millions of customers who have proven their value by using these tools.

This is a game for our time. I play it on New York and Boston subways, but you can play it anywhere everybody in a crowd is staring at their personal rectangle.

I call it Rectangle Bingo.

Here’s how you play. At the moment when everyone is staring down at their personal rectangle, you shoot a pano of the whole scene. Nobody will see you because they’re not present: they’re absorbed in rectangular worlds outside their present space/time.

Then you post your pano somewhere search engines will find it, and hashtag it #RectangularBingo.

Then, together, we’ll think up some way to recognize winners.

Game?

Tags: