How to Fix the Internet

I’m going to tell you something embarrassing. For a long time, I thought the internet would immediately make society better. How could it not? A world where everyone was more connected would be a world where everyone was more empathetic. Obviously. The gatekeepers would fall. There would be more access, more expertise, more chances for curiosity to be rewarded. 

This is not what has happened. 

There are good parts, of course. We really did get access to many lifetimes’ worth of high-quality information, but it turns out the information doesn’t matter that much. Once we had infinite content, what mattered more than anything was which bits were easiest to pay attention to. 

An entire professional class of content creators emerged whose job was to figure out how to make their little video more engaging, watchable, and unignorable than the next. We entered an attention arm’s race. And we did it with the help of recommendation algorithms designed to amplify even the tiniest difference in the ability of a piece of content to hold your attention. 

As a YouTuber, I too try to capture people’s attention. I make a lot of educational and pro-social content, but our systems are not designed to reward it. Our social media ecosystem can’t consistently measure how true something is. All the algorithms know is whether you click or, on the newer platforms, whether you keep watching what is served to you. 

The issue: paying attention is not the same thing as understanding. In fact, a lot of the time, it is in direct conflict with understanding. The result: a media environment that slices the world into the simplest possible shapes. Heroes and villains. Panic and certainty. Fear and superiority. Danger and outrage. 

When you feed a piece of information into the social media salience machine, it always comes out worse. 

But I have found a bit of hope in an unusual place. 

In the 1890s, cheap printing and easier distribution did not just make newspapers more accessible; they transformed the business model of information. Growing cities, rising literacy, better presses, telegraph wires, rail networks, and an expanding ad market created a world in which papers were no longer mainly rewarded for being useful to a loyal set of readers. They were rewarded for getting bigger. And in a competitive mass market, bigger often meant publishing more stories which were more arresting, more scandalous, more emotionally loaded. What grabbed readers’ attention? Crime, vice, corruption, foreign danger, and social collapse. It was a powerful technology, and one for which we did not have a strong set of social norms. Meanwhile, newspaper barons with names like Hearst and Pulitzer got fabulously wealthy. 

But this era of “yellow journalism” did not last forever. It burned hot, warped incentives, did real damage, and then, over time, consumers began to prioritize credibility. People got tired of yellow journalism. In response, newspapers began focusing on building trusted brands. 

This history reminds us that there is always a market for the truth. There is a market for not being manipulated. There is a market for information that is still useful after the initial jolt of emotion wears off.

I cannot help but believe that this history will repeat itself. Today, people are tired of AI slop and misinformation. The social contract of our digital age is broken. They were told the internet would connect them, inform them, and empower them. Instead, many feel manipulated, polarized, and desensitized.

So when I say I still have hope about the positive power of the internet, I do not mean everything is fine. I mean that I think there is an opportunity.

The opportunity is to build institutions, business models, and platforms where credibility becomes a competitive advantage again. Obviously, organizations like this already exist. But new ones are also being built. To build a better internet, we need buy-in from three groups. The first is those of us creating content. Whether podcasters and comedians or journalists and editors, if you want a business long term, you need to resist the urge to flatten everything into spectacle. Perhaps you want to become a supermarket tabloid, but that seems like a bad business to me. Find a path to create actual value, or suffer the consequences. 

Second, social media platforms must admit that their recommendation systems are not neutral. They are the engines that build our reality. If you design them in ways that reward certainty, conflict, and outrage over accuracy, understanding, and context, you are building a worse world. 

And last, we who love the internet must find ways to enjoy media where you retain choice. Don’t give away your agency entirely to algorithms. Choose for yourself what you want to pay attention to. We all have a role to play, and we don’t need to passively consume what we're fed. 

I am not nostalgic for the old media world. It failed in lots of ways. I do not want to go backward. But I also do not think the current version of digital media is the end of history. Our choices are not old gatekeepers or infinite algorithmic slop. We can build new things. Better things. We can build institutions that are native to the internet but not captive to its worst incentives.

This age of extraction is the beginning of something better, but it’s not going to build itself.