Fight, flight, sex, and cocaine
Our brains can't handle the internet's firehose of distraction and threats
Tristan Harris, founder of the Center for Humane Technology, likens social media to the most addictive device ever conceived by humans: the slot machine. Slot machines are so exquisitely addictive because their “reward schedule” is variable: you never know how many coins and pulls it will take to get those lights flashing, bells ringing, and money pouring out. Those sensory inputs trigger a rush of dopamine, the same pleasure chemical released by food, sex, and cocaine, which makes us keep doing it.
Social media works in a structurally identical way. You don’t know when you’ll find a tweet, tiktok, or photo that grabs your attention, so you keep refreshing and scrolling until you get one. That sparks a mini-burst of dopamine, which keeps you looking for the next one. Importantly, unlike ice cream, drugs, and alcohol, there’s no bottom to the social media container. That means your brain never encounters a “stop signal,” which makes it easy, without realizing it, to spend hours scrolling for dopamine hits.
But as we know, social media isn’t all pleasure. In fact, two features of content that drives outsize engagement are 1) the use of words with both moral and emotional resonance, like “shame,” “punish,” and “evil,” and 2) references to out-groups, i.e. talking about liberals to a conservative audience, and vice versa.
As psychology professor Jay Van Bavel and co-authors wrote in 2021:
[S]ocial media may be creating perverse incentives for divisive content because this content is particularly likely to go “viral.” We report evidence that posts about political opponents are substantially more likely to be shared.
Or, as an internal Facebook report from 2018 put it: “Our algorithms exploit the human brain’s attraction to divisiveness.”
It’s the kind of content, in other words, that in some people creates a sense of threat.
Perceiving threats looms large in human evolution, because our ancestors were far more exposed to predators than we are today. Early humans survived this landscape because a part of the brain called the amygdala functions as an early alert system, like a hyper-paranoid lookout.
The instant we perceive a hint of danger – a hissing snake, a movement of grass that might be a lion, etc. – our amygdala sends a red alert to our body so we can escape or defeat the threat.1
The amygdala’s alert bypasses our frontal cortex, the deliberative part of the brain that regulates our behavior, enabling a quick-twitch response that on the savannah could mean the difference between life and death. This is an “amygdala hijack” of our body, in which our deliberative mind, too slow to adequately respond, is cut out of the loop so the threat signal can flood our nervous system with adrenaline and prepare our body for fight-or-flight.
In contemporary life, where deadly threats have been mostly replaced with social and psychological threats, we are often ill-served by our amygdala and its alerts. With social threats we need to be able to stay calm, think clearly, and respond deliberately. But our amygdala sends us the same threat signal it has since hunter-gatherer days, which can lead us to emotional outbursts and other unhelpful responses.
Today, social and often traditional media (cable news) bring us closer than ever to outrages and horrors we may perceive as threats. Media outlets and individuals do this because, as research has shown, it’s the most effective way to win the attention war.
This brings us to a recent David Roberts tweet thread, which makes an important connection between human cognition and the problems of social media.
Roberts says we used to understand that collective discipline is necessary in the information sphere, which may be true, but the more important aspect of the previous era was how simple it was to regulate damaging content. There was far less of it because a comparatively tiny number of entities had the capacity to broadcast2; it was in the economic self-interest of TV stations and newspapers to restrict it; and broadcast licenses were controlled by the federal government.
Roberts is being very honest here that some kind of restriction that can “deny” consumers what they would otherwise choose is the only way to meaningfully shift the harmful dynamics of the internet’s information system. I think he’s probably right. But he doesn’t engage with the question of how that could work, and two issues make the approach unworkable and undesirable, respectively. First, it would be legally and technically difficult, probably impossible. Second, there’s a good chance it would be worse than the status quo, as in, “Before we hear from President Trump, it’s my pleasure to introduce the leader of the Department of Media Regulation, Secretary Tucker Carlson.”
Roberts is right that in a capitalist system people will try to make money every which way, but that was also true from the 1950s through the 1980s. In the 80s, the U.S. had lots of problems, but it wasn’t at risk of descending into civil violence, as we appear to be today. What changed between then and now? Our media ecosystem was overturned by the internet.3 The problems with our information/media ecosystem is primarily an internet issue, not a capitalism issue.
I agree we need more non-market resources for journalism, especially at the local level. But we’ve had lightly regulated capitalism for decades, while the extreme division in our politics today grew hand-in-hand with the internet.
As our cognitive wiring, which evolved to be hyper-sensitive to threats, is confronted with a media ecosystem that bombards us with content that our amygdala perceives as threatening, it’s not a surprise we’re seeing rising anger and paranoia in politics.
I appreciate and respect David’s effort to address this fundamental problem, and to do so in a way that’s honest about what could move the needle. It’s part of why he’s one of the best thinkers around. But I’m pretty skeptical about the feasibility, efficacy, and even desirability of any restrictions regime I can think of. But – I should be honest, too – I don’t see any viable alternative approaches out there. It’s one of the main reasons I’m confident the destabilization of recent years will escalate.
That said, because it’s better than any other ideas I’ve seen, if someone invited me to a brainstorming session on workable internet restrictions, I’d join.
Notes
https://www.nyu.edu/about/news-publications/news/2017/june/messages-with-moral-emotional-words-are-more-likely-to-go-viral-.html
https://www.pnas.org/doi/10.1073/pnas.2024292118
https://www.theverge.com/2020/5/26/21270659/facebook-division-news-feed-algorithms
Humans today have highly sensitive amygdalae because our ancestors did, and our ancestors’ relatives with less sensitive early alert systems were eaten by lions and therefore didn’t reproduce.