The Internet has embedded itself in every aspect of our lives. It's how we work, how we play, how we socialize, and even how we date. But the Internet is so much more than our favorite websites and apps. It represents the collective output and knowledge of our species in a way that deeply weaves us together. It interacts with every law of life. The Internet accelerates cultural evolution, creates new cooperative groups, and supports innovation. It will be critical to creating the connections in our collective brain that will lead us to the next generation of fission and fusion and even to spreading the knowledge necessary to disseminate and support these critical technologies. The Internet is also changing us.
Much has been written about what the Internet and social media are doing to us. Panicked polemics point to our reduced attention span, inability to focus, increasing tendency to multitask, and reduced memory for all the things we can now look up in seconds. Social media is said to be destroying our democracies, worsening mental health, and fomenting division and polarization. All of these things may in fact be true, but the most powerful disruptive effects of the Internet and social media derive simply from the fact that they connect us as never before, storing and giving us access to our trove of cultural knowledge.
Our culture is slowly catching up to figure out the best ways to deal with this hyperconnectivity. Older people, for example, are more susceptible to misinformation. How they learned to learn did not prepare them for a world where the signals of authority and authenticity are so easy to fake. Younger people are less susceptible to what, to them, seem like obvious fakes, but have yet to update their social learning strategies to account for people presenting their successes and best selves through posed posts, filters, and selective information or collapsing worth to likes and follower counts. This may be fueling fears of missing out (FOMO) and feelings of inadequacy leading to mental health issues, particularly among young women.
Social media is also disrupting us in ways that may be beneficial in the longer term. We are more exposed to ideas we disagree with. We get angry and frustrated as we doomscroll on ‘hell sites’, but our exposure to even ideas we find abhorrent forces us to deal with the existence of other people who think differently to us. It forces us to hone our own arguments and allows observers to judge who has the best arguments and evidence. We are also more exposed to the inequality that is reflective of broader changes in society. The previous generation of privilege and money learned to not advertise that wealth gap. And if someone flashed the cash, they were often looked down upon as being nouveau riche ‘new money’ in societies where inherited wealth was considered superior. But the next generation are online socially connected to a larger sector of society and often advertise vastly different lifestyles, making the wealth gap more visible. If you want an insight into the lives of children of the ultra-wealthy, watch Jamie Johnson of Johnson & Johnson's documentary Born Rich, featuring a twenty-two-year-old Ivanka Trump.
The Internet also enables us to find those like ourselves and form ad hoc groups targeted at a cause, from Anonymous hacking Russians to open source software projects with contributors who have never met each other in real life, and to academics and entrepreneurs discovering each other's work and finding ways to productively collaborate regardless of geographic distance.
These trends will only continue. For example, the astonishing success of machine translation may melt the linguistic barriers between us, opening up the full Internet, which currently remains somewhat segregated by languages. Machine translation has already increased the efficiency of international trade, increasing exports by around 10%. This will continue to increase, further shrinking our world as the shipping container once did.
The Internet has also sped up cultural evolution. But much of this has happened by accident and happenstance. Through trial and error and A/B testing, small changes – mute lists and blocks, stars changed to hearts, a larger range of emotional expression, revealing or hiding likes or followers, more or less characters to express ourselves – have large multiplied affects across the globe, but the engineers who implemented these changes may have had little understanding of what those changes were likely to do or even what exactly they have done. Because changes on the Internet and social media have such a large instantaneous impact on all users, it is essential that we use our theory of everyone, particularly cultural evolution for evaluation and improvement of these platforms. The low-hanging fruit is engineering social media to better match how people have evolved to socially learn.
The people we interact with unconsciously shape our opinions and what we think most people think. We deploy our social learning strategies instinctively without being aware of the way they write our brain's software. We are a product of what we consume – who and what we watch, read, listen, and speak to.
It's easy to develop a view of the world that doesn't represent most people if we spend a lot of time on the likes of Facebook, Instagram, Twitter, YouTube, TikTok, Reddit, Quora, or 4chan. But the opinions and opinionated posts we see aren't even representative of most of the people on social media platforms, let alone most people in the world. These platforms self-select for people with strong opinions and a strong desire to share those opinions. The data bears out something close to the 90-9-1 rule: 90% of people are lurkers reading the comments made by 9% of people on the content produced by 1% of people.
This isn't unique to social media. We all live in bubbles. We consume media written by journalists and have lives shaped by politicians and political advisors who may watch the same TV, follow the same people on Twitter, live in the same high-income or middle-class communities, send their kids to the same schools, and go to the same social engagements. Through conversations over cocktails and chats at school pick-ups, our views, too, are filtered. And then, as we share those views, they are filtered into others’ perceptions and work. The best we can do to burst the bubble is to expand our networks as much as possible and try to interact with more people who are not like ourselves. That can be difficult to do in real-life physical interactions, which are shaped by neighborhoods and by who we most often bump into. But social media and the Internet can be used to intentionally burst that bubble.
The Internet and social media have enabled us to access knowledge and conversations we never had access to before. You can join groups and listen in on conversations among feminists and men's rights activists, Wahabi Muslims and Orthodox Jews, white supremacists and Effective Altruists. You can hear how insiders talk to their ingroup in semi-public conversations on Reddit, Telegram, WhatsApp, Substack, Discord, forums, or other social media. But many don't.
Instead, we are often passive recipients of newsfeeds produced by algorithms whose only goal is to keep you engaged. But it doesn't have to be designed that way. Social media can be deliberately designed to expand collective brains while keeping people engaged. It can also be engineered to better exploit what we naturally pay attention to.
Many social media platforms show you how many followers a person has – an indication of popularity. Popularity is a signal that our social learning psychology pays attention to, but a social media follower count is a very weak indicator compared to the data we attend to in the real world.
In the real world, we care not only about how many people know someone or follow them, but who those people are. Prestige comes not just from amassing followers but by amassing followers who themselves have many followers. In the language of network theory, we pay attention not just to direct centrality (number of connections) but to eigenvector centrality (connections of connections of connections onwards).
To put it simply, a high-school sports star may have many friends, but if they're all other high-school students, she will be less prestigious than someone with fewer friends but among which are former presidents, celebrities, and successful business people.
High eigenvector centrality has been empirically demonstrated to be a source of influence within communities and a more accurate marker of prestige than follower count. If you've ever looked at who someone's followers are, you are implicitly looking for eigenvector centrality. In fact, Google's success over previous search engines was thanks to its use of eigenvector centrality calculated by PageRank, which ranked web pages based on who linked to them and who linked to the linkers. But this information is nowhere to be found on social media.
Other missing biases include more honest indicators of expertise and success. Social media clout is easily faked. In fact, reputation itself – an important cue for both social learning and cooperation – is often poorly tracked.
I mentioned previously the way in which online reputations can help enhance cooperation for ride sharing or room bookings, but reputation is often not leveraged in other online information.
When we read any article or see a post, we are rarely presented with more context about the author. Who they are, their expertise, what else they've written, how many times they've been right or wrong, what other people think of them and who those other people are. That information exists but is not aggregated, and so one of the most powerful characteristics of a person that we care about – what other people think of them – their reputation – remains untracked and not leveraged.
Instead of individual reputations, most of us use what we might call securitized reputations. Rather than the reputations of individual journalists, most people rely on the reputation of the outlet – New York Times, Washington Post, or other journal, magazine, or newspaper. Journalists who have large enough personal reputations moving to personal newsletter platforms such as Substack are one example of moving away from securitized reputations. But here, too, reputational information is often missing, limiting our ability to find valuable new sources of information.
Reputational information includes past work, expertise, what others think of the author, funding and conflicts of interest, and what people got right and wrong in the past, and which people think what about the author. But in the same way that statements about conflicts of interest can be useful, so too can information about incentives. What we might call incentive audits are useful, particularly when the stakes are high.
An incentive audit is to self-reported conflicts of interest what tax audits are to self-reported tax returns. They make information that may be relevant to a person's beliefs more salient. For example, who they are friends with, their personal circumstances, who own the companies they work with or for, and their sources of income. If you scratch the surface of many people's beliefs, you'll find self-serving incentives. As the old adage goes, ‘It is difficult to get a man to understand something when his salary depends upon his not understanding it.’ You can probably guess who will support which positions if you know enough about the opinion holders.
For example, it should be no surprise that people with more education are more likely to believe in a meritocracy, the upper end of the income distribution are more likely to argue for lower taxes, or that people will signal beliefs that gain their ingroup approval.
Sometimes signaled beliefs are costless to us personally; they simply signal our affiliation or that we are good people as judged by our ingroup. The accuracy of evolution as an explanation for human origins, whether we should defund or defend the police in other people's communities, or whether transwomen athletes should be allowed to compete in women's sports are often topics that people hold strong opinions on, despite them having little bearing on their own lives. We can advocate for and reinforce group affiliation without having a stake in the game or it impacting on our everyday existence. We can continue to enjoy the benefits of evolutionary approaches to medicine, living in a safe community, and not participating in elite sports. One way to reduce the effects is to make signals costly – for the signals to have consequences – revealing and making salient relevant information and potential biases and encouraging a culture that normalizes such disclosures.
Incentive audits are also worth applying to ourselves. It takes humility to admit the limits of our knowledge and that we are incentivized to seek out certain knowledge, be strategically ignorant about what doesn't serve our interests, and when there are competing beliefs, believe what serves us best – and that's not always truth. Remember, some beliefs are acquired through direct experience, but most are indirectly acquired from others, where the causal relationships are complicated or unclear. In an uncertain world of mixed evidence, why not believe what would serve you best if true?
These self-serving biases can sometimes harm us. For example, they may lead us to worry about less important but more controllable decisions. People often worry about the health effects of parabens in shampoo or pesticides in non-organic vegetables, which are decisions that are easier to exercise at the supermarket than actually getting more exercise at the gym or just eating more vegetables, organic or not.
It is also easier to hold self-serving beliefs when the information is socially acquired with a choice of what to believe and when causality is less clear. These are also cases where we tend to copy most faithfully, not knowing which bits of our beliefs are necessary for causality. Religion is an extreme example of beliefs and behaviors that are more indirectly transmitted than directly discovered. As such, religions often have more ritual behavior, with emphasis on exact copying.
For example, in the early 2000s some Catholic priests, rather than baptizing children with the conventional ‘I baptize you in the name of the Father and of the Son and of the Holy Spirit’, began using the more gender-neutral ‘We baptize you in the name of the Creator, Sustainer, and Liberator of Life’. Some priests, wanting to hedge their bets, added, ‘who is also Father, Son, and Spirit’. That may seem like a small change to non-Catholics, perhaps even a preferable one, but it led to great uncertainty about whether children were actually baptized or not. With the destination of souls at stake, the controversy went all the way to Rome and a decision was made that the incorrect wording rendered the baptism invalid. Hundreds of children had to be re-baptized. Although this tendency toward exact copying is common among more devout religious people, this psychology is not unique to religion.
The separation of the natural and the supernatural, the secular and the sacred is a rather recent WEIRD idea. For much of the globe and for most of history, the world was just the world, not a mixture of natural and supernatural. The real psychological separation is between beliefs and behaviors that are more directly or indirectly acquired. Non-religious people also hold indirectly acquired, evidence-free beliefs stated with certainty. Take the US Declaration of Independence: ‘We hold these truths to be self-evident, that all men are created equal’. This idea of equality before the law or before God is not self-evident nor derivable from individual characteristics. It is a bold statement in a world of people such as Usain Bolt, who can run faster than the typical residential speed limit for cars (25 mph), and Simone Biles, who has four gymnastic skills named after her. At best, it is a belief that might be defended as being useful or creating less suffering. It is also not a belief that is practiced as it is written, not even at the time it was penned. The author himself, Thomas Jefferson, owned over 600 slaves. Other American presidents who owned slaves, including while in office, were George Washington (also over 600), James Madison (over 100), James Monroe (~75), Andrew Jackson (~200), John Tyler (29), James K. Polk (56), and Zachary Taylor (~300). John Adams and his son John Quincy Adams are notable for being the only two presidents in the first dozen not to have had any slaves. And yet, the self-evident nature of human rights and equality or the assertion that this is a better belief than the alternative are often asserted as obvious or objective truths. It is indeed a useful belief that can in turn create progress through coordination around the belief, accusations of hypocrisy, and runaway cultural evolution, as people compete over a more authentic and inclusive implementation of equality.
By now, it should be no mystery why some fall for what others call fake news or conspiracy theories. Those who hold these derided beliefs are simply people who have learned what they know from people they think are trustworthy but we think are not. That is to say, none of us is actually prioritizing truth, especially when that truth harms our interests. In any case, few people have the ability to verify first hand the many beliefs they hold about the world, from whether germs and not spirits cause illness to whether the world is round or flat, let alone whether equality as a moral principle is useful. Instead, whether or not we hold true beliefs is not a function of our ability to distinguish truth from falsehood; it is a function of whom we trust. And that's true of scientists as much as it is of anyone else.
Scientists trust and internalize scientific results based on the expertise or success of the person who ran the study, or the prestige of the journal in which it was published, or if the findings are considered good or bad. This is part of why it took decades for the psychological and behavioral sciences to realize that half the literature was untrustworthy. And before you assume this is restricted to these disciplines, after the highly publicized replication failures in psychology, other disciplines also ran replication tests. This should terrify you: more than half of the top cancer studies have failed to replicate.
To trust the science is to trust the people and the process. Rarely are beliefs verified through direct experience or personal replication, but rather their plausibility is checked against everything else we know about the world – or, at least, what we think we know about the world. Many of these assumptions are indirectly acquired from childhood onwards based on the culture in which we happen to be born, the subjects we happened to study, and the people we happen to encounter. And when scientists hold scientific beliefs about subjects outside their own field, then we have no direct experience of even how the theories were developed or the data was collected. Instead, a scientist's belief is founded on a generalized trust in ‘science’ as a process and scientists as honest practitioners of that process.
As cultural learners, we have evolved to focus on who is making the argument more than the argument itself. Among left-leaning Americans, Trump's attempts to close borders during the COVID-19 pandemic were far less well received than Biden's use of the same strategy. The same argument or behavior may be defended when done by those on our team but condemned when done by those on the opposite team.
Those who design popular social media platforms and AI tools have incredible power to shape the software of our minds and therefore our societies. Small changes made by a few people guided by A/B tests and optimizing algorithms deployed to millions can help or harm our ability to cooperate and to innovate. These decisions currently do not reflect a theory of everyone and the ways in which we evaluate who to learn from and what to learn. But by marrying an understanding of cultural evolution and engineering design, we can use the tremendous power of the Internet and social media to enhance the brilliance of our collective brain. The Internet gave birth to the Second Enlightenment, but the future of this Enlightenment and a true Second energy-based Industrial Revolution depends on getting our information architecture right.
With small changes, such as more deliberately exposing people to alternative views and communities who hold opposite beliefs; by providing contextual cues of trustworthiness, honesty, and reputation that we use in the real world; by offering audited incentives that may shape a person's genuinely held beliefs; and by cues of a person's identity and relevance to oneself, we can help trigger a creative explosion and become brighter.