Our phones aren't changing us

Plus: Facebook's AI Jesus; why people really lie to pollsters

Hello there. It’s been a while, right? Between job search and a lot of travel, free time has been scarce. Oh, and I moderated a panel at SF Climate Week. Thank you for your patience.

My family visited the Museum of Modern Art (the one in NYC) a couple weeks ago. That day everyone seemed to be speaking French. My daughter mentioned that the thick crowd made her think that she would not enjoy visiting the Louvre someday, but I assured her that the Louvre would definitely have fewer French people.

Among MoMA’s attractions is Van Gogh’s Starry Night, which is probably one of the 10 most famous paintings in the world. It was easy to find because a crowd of 15-20 people were constantly massed in front of it. As I was taking in Rousseau’s The Sleeping Gypsy, a work of performance art was happening next to me, as new people would join the Starry Night group from the back and work their way to an angle where they could see it. And then most would look at it through their phone screen, take the photo, and leave.

You’ve probably seen similar scenes documented, whether it’s the Mona Lisa viewing, almost any concert, or anywhere on New Year’s Eve. This is a normal and natural act — why wouldn’t we want to capture something that’s special, extraordinary, even unique.

A lot of people think that experiencing life through our phones is something young people do (here’s a particularly annoying example), but it’s absolutely not. Old people like me are as glued to our phone as screenagers are. I’ve observed that the concerts that attract people over age 40 have more phones up, recording longer shitty shaky videos, than those that attract younger crowds. The kids are alright.

But are the kids alright? Steve Jobs allegedly didn’t let his kids have iPhones, and that was well before our entire society reorganized around them.

Among the ways I differ from Steve Jobs is that I was comfortable letting my kids have smartphones as soon as they got to middle school. Well, maybe comfortable isn’t the right word. But it was normal, and I’ve noticed the friends who didn’t get phones got left out of group texts and remained dependent on their parents to see their friends.

Last summer the writer Erik Hoel published a very popular essay called What the heck happened in 2012?, which proposed that year as a downward inflection point for a number of psychological and cultural metrics in the USA and the world in general. It turns out that nonstop entertainment is depressing, and nonstop connection is isolating. Maybe.

Perhaps the Mayans were right and the world ended in 2012. But also, the early 2010s were when smartphones went cheap and mainstream. What we had previously called “mobile phones” were rebranded as “feature phones” or “flip phones.”

I was working at YouTube at the time. In 2012, the majority of YouTube viewing was still on computers, with the second-most on our mobile website, m.youtube.com. 2012 was the start of explosive mobile growth for both YouTube and Facebook, and also this was the year that Facebook bought Instagram for one billion dollars.

An astute Reddit comment in a discussion about 2012 pointed out that YouTube also changed its algorithm from “clicks” to “watch time” that year, leading people down dark content rabbit holes. I forgot that this happened then, which is pathetic because I literally launched this feature. (The Reddit comment links to a Search Engine Watch article that is mostly parsing a long quote by some corporate tool, who strangely has the same name and the same job title I had back then.)

Hoel points out that 2012 was the year when smartphone ownership approached 50%, but he’s careful not to apply causality to it. Jonathan Haidt is not so cautious, as he has been making the rounds over the past few weeks promoting his book The Anxious Generation, which very specifically connects the dots between the psychological decline of young people and smartphone use, specifically social media.

I’m dubious about the moral panic around phone habits. I think we’re all better off shutting down our phones sometimes and touching grass, but I also know parents who lean on the “it’s the phones” excuse to deflect their own responsibility for their kids’ various disappointments.

Phones are also just devices. When people say “phones” in the context of “doing something bad to people’s emotions and brains” they may mean “Instagram” for kids and “Facebook” for olds. They’re probably not talking about texting grandma, reading restaurant reviews on Yelp, or editing their photos from last night.

But phones are also distracting and addictive, and like almost everyone else, I often catch myself grabbing and staring at mine for no discernible reason. I’ve transferred so much on my life to the device, I sweat all the things I couldn’t do if I wanted to turn it off for a few hours. No music, for example. (Maybe I should learn guitar.)

A recent study out of Norway posited that taking kids’ phones away during the school day improves grades and reduces bullying in middle schools. This report got a lot of attention last week, mostly because it puts data to something everyone always expected. My son’s new high school says it’s going to ban phones during school hours this fall. I’m skeptical they’ll pull it off. His middle school tried this with lockable baggies last year, and it hasn’t stuck at all.

Although younger people have grown up with their screens, many of them understand that this is all very new. You only have to watch one old movie or TV show to understand how different human experience was before we had internet access without limits.

I recently saw a video by a man in his 20s who had just watched the new Hulu documentary on Freaknik, lamenting how much more fun it looked in the days before nonstop distractions and communications.

“Our generation looks back at the ‘90s and early 2000s with so much nostalgia, whether it comes to the movies, the music, the fashion... And we’re always trying to throw ‘90s and 2000s parties… But the one thing we constantly miss is that our parents didn’t have cell phones, bro. Like they could focus on being in the moment, and that’s the biggest difference between us and our parents… And that’s why I feel like we will never have as much fun, or more fun, than our parents.”

Now, this prespective isn’t entirely accurate. Cell phones were a luxury good for much of the ‘90s (often supplemented by cheaper alternatives like two-way pagers), but by the turn of the millennium, most adults had them. By the early 2000s, we were even texting, cumbersome and costly as it was.

This was the real dawn of the FOMO era, well before iPhones and Instagram. In 2010, I talked with some graduating college seniors as I attended my 15th college reunion. They were genuinely curious how our college experience differed in the early ‘90s, but they knew “you didn’t have phones” was the most obvious difference. They expressed to me the anxiety they would have felt going to parties or trying to meet up with people, without a reliable way to reach them.

“If you went to a party, and it sucked, how would you find out where the good party was?”

“If you had to meet someone for lunch, and you were going to be really late or couldn’t make it, how would you let each other know?”

“If you were trying to find someone at a concert or party, how would you do that?”

And the answer was: we had clunky ways of leaving and checking messages. But really it was just normal to be at a shitty party sometimes, or to have to work hard to find someone in a crowd, or to be generously forgiving to the friend who got delayed and couldn’t reach you.

This was social culture for the entire history of civilization, until about the early 2000s. And, to be honest, technology has improved this part of life tremendously, and I would never want to go back.

But to kids who grew up able to contact anyone they knew at any time, this sounds risky, even dangerous. Usually this risk was just being annoyed or having a sub-optimal night out, but sometimes the risk felt more mortal.

The young man in the Instagram video goes on to suggest that, maybe, we could replicate that spontaneous Freaknik vibe today with more social events where people voluntarily lock up their phones. Like Norwegian middle school, we’d have better outcomes if people were connected directly to what’s happening in front of them. But he also noted that then we’d likely need more security at events.

But we didn’t have more guards and cops at events back then. We probably had fewer. We certainly never had security at college parties — people ran away when they showed up.

While fewer guns were available back then, the general level of crime and violence was actually significantly higher than today. Our major security measure was “GTFO” — literally just run away. I’d be surprised if a phone-free event today were less safe than one where everyone was still armed with their iPhones and Pixels. In fact, the greatest security value a phone might have at an event is being able to record security people abusing their power.

But also, I get it. Being phone-free can feel unsafe. I went with my family to our local mall about a year ago, and when we arrived I realized I’d forgotten my phone at home. I’m not going to lie, stepping back in time and setting a meeting place inspired all sorts of queasy feelings. If an emergency broke out, what would we do? (Probably meet at the car, it would be fine.)

On the other other hand, my family did attend a phone-free event when we were in NYC — a broadway show. The ushers enforced the rule by walking the aisles and shining flashlights on people who had their screens on.

Now, stage productions are mostly trying to keep people from shooting videos like esteemed Congresswoman Lauren Boebert did, and fortunately we didn’t have any Coloradans vaping and groping and singing and drunkenly ruining the show. But the Hadestown ushers also busted people who were just looking at their phones in their seats, and it’s pretty clear that that enforced focus made the show better for everyone in attendance.

In the end, our phones aren’t really doing anything to us. We’re actually choosing this. As I watched the crowd appreciating Starry Night on that rainy afternoon at MoMA, I saw maybe two-thirds approach the painting with their phones up, snap the photo, and then depart. They never really looked at the painting with their own eyes. But the other one-third stood there and gazed at Van Gogh’s magical brushstrokes, appreciating how he perceived that scene he’d painted, maybe all that had changed since, and what this artwork must have meant to the millions who have enjoyed it over the decades.

Then they whipped out their phones, snapped a photo, and moved on.

When the polling tells us the impossible

As long I’ve been alive, we’ve seen political campaign coverage take the primary form of “horse race journalism.” At the start of each campaign, you can’t really see the horses racing, so media just reports on how the horses are looking to them, how well they did in their past few races, and how they might match up with each other.

Then as people start paying attention, the horses hit their stride. But then people vote, and you realize the horse race was just an illusion. Polls aren’t real.

If polls are to be believed, the 2024 presidential race is essentially tied. After 1988, the Republican candidate for president only won a popular majority one time. A very close popular vote would probably mean an electoral college landslide for Trump, who this year has been found guilty of both rape and a massive financial fraud, and is currently facing 88 felony charges across four different criminal cases.

A lot of people think that “horse race” coverage is actually inappropriate for this race, or at least not helpful. But I think the larger problem is that we can’t see the race: polling is broken.

The poll-readers would have us believe a core narrative, which is that some pollsters are better than others, and a weighted average of polls will give you something resembling reality, like an AI-drawn cartoon of a horse that’s really a composite of every cartoon horse the AI has ever seen.

But the DNA of a poll is the data, and the data requires three wins: (1) you’re reaching a base of people who represent the diversity of the electorate; (2) the people you reach are willing to answer your questions and pay attention; (3) the people answer your questions accurately and truthfully (they don’t lie about whom they’re currently supporting).

People talk about problem 1 a lot. Telephone polls involve calling people, mostly on landlines. It’s almost hard to remember, but back in the landline days, when your phone rang, you answered it. This ostensibly gave us a better population sample. Today most people wouldn’t even answer a call from an unknown number, and imagine the lonely people who do. Hey look, it’s my buddy Scam Likely again!

People started suspecting problem 3 was a real issue following the 2016 election, when Trump wildly outperformed his polling… almost everywhere. It’s impossible to know if problem 3 (truth & accuracy) is actually problem 1 (you didn’t reach a representative sample). But the Trump campaign’s ads and messaging fully embraced the transgressiveness of voting for him. Given his shocking scandals down the stretch (the Access Hollywood tape, Trump University fraud) and reluctance of establishment Republicans to attend his 2016 convention, it’s understandable that a percentage of committed Trump voters would tell a stranger on the phone they were “undecided.”

But today a lot of these polls aren’t on the phone. They’re online. And that’s where problem 2 really comes in — are people even paying attention, or just clicking for dollars?

I vote in YouGov polling every week or so. It’s a fun little diversion, and after a year you might have enough points to take your kids to Buffalo Wild Wings.

Every YouGov poll has a few “are you paying attention?” questions, now standard for online polling. It might take the form of “Which of the following is a day of the week? November, March, Saturday, April.”

Sometimes it gets a little more amusing like this:

They use questions like this to help signal patterns of reckless clicking, but they don’t throw out every survey where someone enters the “wrong” item. Maybe you did patent a new invention last weekend. Your vote counts too!

Pew, who produces some of the most rigorous surveys about American life, recently published a post about the trouble with opt-in online polling, specifically the problem of people clicking for compensation. They reference a 2022 study that showed that the data for opt-in online polling deviated dramatically from reliable, in-person survey panels, often by double-digit percentages.

In one case of the “paying attention” question, 8% people surveyed via online poll said they did something extremely unlikey yesterday, like buying a jet or climbing a Karakoram mountain. Remarkably, this reached 17% in adults under 30 years old, and 28% among all Hispanic respondents.

The funniest check-in question might be “Are you licensed to pilot an Ohio-class cruise missile submarine?” Only a handful of people in the world could truthfully answer “yes,” but 5% of adults affirmed they were. And that jumped to 24% among Hispanic respondents!

Presidential politics aside, a lot of the surveys we see in headlines today are molded from garbage, where they’re not rigorously managing data quality. You may read alarming polls about people denying the Holocaust or the earth being a sphere, or the favorite pizza topping in every state, or that Republicans are going to sweep the elections this fall. In headlines and tweets, these findings are treated as facts, with little scrutiny as to the methodologies.

Or maybe we’re a nation of Ohio-class nuclear submarine pilots. I guess we’ll know in November.

AI Jesus takes over Facebook.

Facebook is rife with AI-gen images of white Jesus. He is often chilling with Asian flight attendants for some reason?

You may have also seen Shrimp Jesus, third-world Jesus, and first responder Jesus. You may have seen thousands of soft-headed people commenting “Amen 🙏” which has been the most common comment on Facebook this year.

Strategic disinformation is expensive, writing a script to generate a zillion AI pictures and spamming them across thousands of Facebook accounts is cheap.

Anyway, I’m more into AI yearbook pages these days. Tag yourself, I’m Broop.

Quick items to take us out

US Air Force confirms first successful AI dogfight test. One thing that always bothers me about sci-fi movies: somehow these civilizations have mastered interstellar space flight, but they’re still manually piloting their spaceships and firing lasers like a WWII movie. Anyway, computers will be flying all our aircraft, even in war, well before we meet hostile civilizations in outer space.

Steak ‘n Shake installs facial recognition at self-order kiosks. You probably didn’t think dystopia was going to take a huge leap at a fast food chain. The most margin-sensitive businesses are going to be the ones to innovate people’s jobs away. California’s bizarre minimum wage carveout just for fast food employees will accelerate this.

Is my girlfriend a time traveler? Young redditor’s girlfriend won’t talk about her childhood, confuses her years, wears antiquated undergarments, has never seen popular TV shows. There’s only one reasonable explanation — she leapt from the 19th century.

Murders are plummeting in American cities. The pandemic-era spike in murders has waned, and people are killing fewer people across America.

Sound of Freedom and the Shape of Modern Propaganda. The sleeper hit of summer 2023 was about a white savior liberating trafficked children in Central America, purportedly based on a true story, starring a Q-anon celebrity. Its success was propelled by a message during the credits asking people to buy more tickets to defeat evil. The linked long-form YouTube video uses this movie to examine right-wing internet culture and the new nature of propaganda. It has the vibe of a rant by a graduate student who’s hanging around in your freshman dorm at 1 am, and I mean that in the most affectionate and positive way I can.

NASA scientists fix Voyager from 30 billion miles away. Voyager has survived outer space for 47 years, and has been communicating from interstellar space for more than a decade. It started glitching, but NASA fixed it somehow. Meanwhile, your new Smart Dishwasher probably won’t make it until 2029.

Kirstie Allie’s parents were killed while dressed as what now? I guess this kind of explains Kirstie’s whole deal with Scientology and her descent into MAGA. Rest in peace, all of them. And may AI Jesus grant them a change of clothes in the afterlife.