The attention economy. Attention is all you need. The politics of attention. Attention has never been more central to our lives. It’s precious, easily hijacked, and dominates how we receive information. There’s never been a more important time to rethink how we use it.

My career hacking attention

Attention on deck!

You get used to hearing this as a plebe. Any time an upperclassman, professor, or officer enters a room, whoever is closest to the door shouts it, and everyone snaps to attention. It’s thus only slightly ironic that so much of my career has focused on attention.

With arcade games, it was the attract mode and that first quarter. Especially in the late ’90s, how did you get a 14-year-old heading for SF2HF to pause and consider your weird game with trackballs? Better have big dragons and dramatic camera movements. For Road Rash, even though it was sneakily open world—you could go infinitely off track in any race if you wanted to—it was about making sure when you played it at a friend’s house, the first level had you in and that multiplayer had you laughing and yelling. Plus a commercial of course.

Attention reinforcement

X (then Twitter) and Meta (then Facebook) had different problems from early games—much closer to Second Life—near infinite content, low signal-to-noise. Twitter solved it by continuously driving down the cost of attention, the cost of checking and scanning through a ton of posts. This scaled for a while, but eventually broke catastrophically. Facebook had the crucial insight that the only hope was to deliver an algorithmic feed, one prioritized for you. Originally hand-tuned, the company turned ranking over to ML models just before the mobile transition in 2012.

Feed Machine Learning models—which prioritize both content and advertising for you in feeds and feed-like interfaces—are broadly about attention reinforcement. Oh you liked that, here’s some more? Oh, people who generally liked X and Y, also liked Z. Like all AI, they are prediction machines, operating off of user actions like views, likes, and shares.

With everything that has happened over the intervening 15 years, it’s easy to forget how transformative an experience News Feed was in the early days. You reconnected with friends, saw that picture you would have been sad to miss, and suddenly had a broad web of weak ties that brought joy to hundreds of millions of people.

It was awesome. While Twitter was superior for eavesdropping on celebrities and keeping track of live events, in most ways the algorithmic feed was far superior. It scaled, respected your attention, and — with the changes I introduced to mobile advertising in early 2012 — monetized spectacularly.

Why did this all work? To a large degree, you were likely to engage with content that was high quality for you. Those clicks, likes, and shares ensured you saw more things like it. Even better, amateur social content gave ML systems signals that were good at predicting quality: does it have an image, how many words are above the fold, does it have a headline.

You won’t believe what happened next

Over the next decade, two powerful trends started taking over. The first, US political polarization accelerated. The second, everyone figured out how to game attention with outrage. Whether the goal was growing your personal brand, advertising, or spreading disinformation, hijacking attention reinforcement algorithms with outrage became the winning strategy.

My friend Hank — who’s built a remarkable career and brand online over the last 15 years — goes into the details and economics of it:

The crux is that it appeared that no one actually listened to my words on the video, or if they did, they didn’t care. They just wanted to be outraged.

The Hate Machine requires us to be so in order to function.

It feeds on grievance, on umbrage. In so many areas of our life. And we willingly feed it those things, at all hours of the day. For me, Instagram is, normally, a haven away from this. But in this case look what happened: I posted something that turned out to be controversial, and I was rewarded for it.

All those likes. All those new followers. All those views. If all I cared about were metrics, I could post that sort of thing all the time, be internet famous, revel in my influencer status.

The platform rewarded me. This is the fundamental trap of attention reinforcement algorithms in a world of outrage. Outrage drives attention, attention drives clicks, clicks make everyone richer.

Feeds and news

It’s even worse for news. Attention reinforcement feeds didn’t cause polarization—after all, we’re 30 years into large-scale, partisan news as business—but they are easily hijacked by it. Despite spending significant time and resources trying to improve news in News Feed, Meta has deprioritized news content. X has gone in a different direction, embracing the polarization and seen many users from across the political spectrum return.

The long-term data is complicated but looking from the outside at how large platforms have moved attention away from covering news, it’s clear that in a highly polarized world, news is too easily weaponized to generate outrage and hijack attention. Great for short-term revenue, horrible for long-term retention.

Using quality to respect attention

Part of what drew me to SmartNews is its Mission:

Delivering the world’s quality information to the people who need it

Quality information. What a compelling framing to think about news. Today – thanks to an incredible set of publisher partners – we are fortunate to have more quality content than anyone could possibly read in a day. This puts us in a position of deep responsibility to consider how we rank, prioritize, and deliver that news. Both our Mission and our first Core Values –- ”for the common good” –- put tremendous pressure on us not only to be opinionated about news and quality, but to respect people’s attention.

If there’s news you want to pay attention to, how do we bring you the best, most complete view of the story? We don’t want to bombard a reader with repeating similar stories, nor fall into the trap of a single, narrow editorial viewpoint.

And in an America where political polarization so often frames—and rewards—the presentation of news, we must recognize this is often the least informative way to understand a story.

For me, this goes back to the emotional core of the early days of social media.

Quality news is something you would regret not knowing

It’s time to respect that idea.