Can decentralized social media save us?

Can decentralized social media work?

Welcome to The Perfect MVP. If you’re joining the party for the first time, welcome! We meet here once every two weeks. I write my reflections on how we can build tech to be better for all of us. You can subscribe by clicking the purple button below.


I’ve been thinking about social media lately. When the history books are written about the 2010s, it will be impossible to omit Facebook, Twitter et al. For better and for worse, these apps have become dominant news sources for billions of people. They’ve enabled activists to create strong movements across borders and they’ve facilitated the spread of fake news at an unprecedented scale. They’ve given us open-source tools like React, Presto and PyTorch that software engineers around the world use everyday. And they’ve failed to properly safeguard our data hack after hack.

Each platform has its own resume and societal rap sheet. But the kinds of problems they face, issues of privacy, election fraud, doxxing, physical violence, self-harm, suicidal content and violent radicalization are problems with no clear silver-bullets. There are better and worse solutions, of course, but rarely do you hear a single solution that balances all the different tradeoffs in a meaningful way.

Recently, I’ve been hearing more people tout decentralization as one option to improve social media. Twitter in particular is very fond of the idea. Jack funded a small team to explore a decentralized social media protocol.

What does decentralization mean in the context of social media? And how might it help fix some of the challenges herein?


What is decentralized social media?

❗️TLDR (too long, didn’t read)

A decentralized system is one where there is no single body in control. So a decentralized social media network is one where there are multiple networks, each with their own rules, values and enforcement mechanisms. In this world, there are 10,000 “Twitters” instead of one. Each network is owned by the community, and you the user get to choose which “flavour” you want. Maybe you want a network where you own all your data. Or one that’s private by default with zero ads. Or one where you choose what kind of “funny” tweets are recommended to your timeline. Maybe you want to define which issues or values you care about so you see less or more (choice is yours!) about these.

The choice here is yours. There is no central body. No single attack surface. And no single intellectual orthodoxy. These sites already exist today (see Mastodon, Minds)

📚TSWM (too short, want more)

Decentralized systems are an extension of the idea that “people should be trusted to choose what is good for themselves as long as they don’t harm others”. This idea is reflected in some but not all of society’s machinations today. Consider toothpaste and tobacco. We know toothpaste is required for good oral care, and tobacco kills nearly 500,000 people a year in the US. Yet, we allow both to be sold in stores. Sure, we tax tobacco products to reduce smoking but we don’t outright ban them. The personal liberty argument is that adults should be free to decide for themselves if they want cigarettes or not. Even though it costs $300 billion each year in the US healthcare costs, we allow adults to choose.

So we insist adults have freedom in the offline world. Freedom to do as they wish barring they don’t break laws. Even if these things are harmful, or addictive (see tobacco or gambling). We use personal liberty and freedom to explain our rationale there. How then should we think about adults on social media? Should adults have more freedom with their social media? What if someone wants to use Twitter but feels limited by its recommendation systems? Do they only have two options- use Twitter as is, or leave? Like we have multiple toothpastes, can we have many other kinds of Twitter with customized community values? Can we trial other versions of Twitter with different governance models? Can we build Twitters that will uphold different values? What might these look like?


How might decentralization help social media?

❗️TLDR (too long, didn’t read)

It’s a lot harder to hack into or spread fake news in 10,000 places than one place so there is the benefit of resilience and security. Fake news will still persist of course, but users will know the moderation approaches of the group they join. So if you’re in a network with “loose” moderation and you disagree with the handling of some content, you can leave for another one (and possibly carry over your data and profile assuming the systems are inter-operable) by clicking a button. Moreover, more social networks means that you can choose to join ad-free, privacy-focused networks that will employ different(better?) tactics than today’s social media giants. And if you stay in the ad-supported ones, at least you explicitly did that because that’s your heart’s desire.

📚TSWM (too short, want more)

🧠Intellectual orthodoxy

If Einstein was alive today, would he be popular or even respected on Twitter?

Today’s social media is not a place for intellectual debate. The format encourages shorter snippets that provoke and outrage rather than illustrate and elucidate. Despite this, people still try (and fail) to have regular debates on the platforms. It’s gotten so bad that people are genuinely surprised whenever a debate ends in a somewhat civil fashion. There’s also an intellectual orthodoxy problem. Whenever there is a centralized platform, therein lies an economy of incentives that dictate what kinds of ideas or content will flourish. Some think their timelines are too leftist. Others think they’re not progressive enough. I bet Twitter would look and feel quite different if it was founded and run by conservatives. But regardless of who owns, funds or runs these sites, there will always be outsiders. Some outsiders should be outside - all views aren’t good faith views so they don’t all deserve consideration or amplification. But good ideas often come from the fringes and I often consider how Einstein would be treated on social media for his views on gravity and space-time. He would be ridiculed, dunked on and laughed at. I mean that’s what happened a century ago, and I worry we’re making the same mistake with other good-faith, evidence-based, thoughtful, non-orthodox ideas today.

🌍 Splinternet

More networks means an ever-widening splinternet. That is, the further fragmentation of the internet across nationalist, political and religious interests. I think this is a foregone conclusion. I don’t see a way back. I worry that tomorrow’s decentralized social media will fracture the notion of collective truths. This could easily lead to deeper, tribal divisions between us in the analog life.

We already struggle with notions of shared reality and objective vs subjective truths while being on the same networks. Tomorrow’s Twitters will have self-proclaimed “oracles” and “orators” who will manipulate their networks into indoctrinated, online armies. I worry about these echo chambers and feel we need new systems that can connect people across social networks in a non-judgemental, human manner.

What do we become when we can’t even agree on what happened yesterday?

❣️Experiments welcome

Some might argue that Twitter is “passing the buck” of moderating social networks to decentralized social networks. While I think they should do more to fix their existing product, I think this work makes sense. Twitter today is a melting pot of tribal groups yelling at each other, and taking occasional breaks to laugh at the meme du jour. We’ve seen what happens when tribal groups in the same place interact all day, and it’s not necessarily a good outcome. I think we need more experiments. Twitter needs more anti-harassment tools built into its product. Twitter needs to find out the unintended consequences of its existing recommendation algorithms. And then act on them in meaningful ways. And they should trial decentralized social networks.

There’s so much to see:

  1. Will people even leave the default social network? Or will they just stay where their friends are?

  2. What kinds of user-controls will exist? How easy will it be to move from one network to another?

  3. What kinds of social networks will emerge and what behaviours will those networks incentivize?

  4. Will the popular networks be pseudonymous?

  5. Given regulators struggle to deal with a few large social networks. How will they deal with thousands of them? Will they be able to effectively handle social networks where illegal activity flourishes?

  6. If Parler was on Twitter’s decentralized platform when the Capitol Riots happened, would they have de-platformed them?

  7. What would the monetization strategies look like?

There are many unanswered questions in this domain. But I’m curious to see how this develops.


✨Talk to Tobi

What do you think about decentralized social networks? Are you excited to use them? Are you concerned? Tell me in the comments!

Share

❗️❗️If this post resonated with you, consider sharing this with your friends.