Twitter won't be killed by Elon or by a clone

Twitter is a textbook case of how potent network effects are. The platform won’t die even though the widespread feeling is that using Twitter feels worse now than a year ago. At this point, I don't see anyone defending Elon's management of Twitter — he is alienating both the hardcore users and the developer ecosystem. He's capable of acknowledging the problems that Twitter faces, but at every turn chooses the wrong option. I get a similar feeling seeing how Zuck handles Meta: they are very clever people capable of great things, but they lack the basic social intuitions to innovate and manage a social product.

Twitter is focusing on things like cutting costs, increasing revenue, and reducing spam. These are lofty goals, but a social structure is a delicate thing and needs to be manipulated with care and empathy. One big decision is whose voices you amplify. Twitter has become a pay-to-win, on top of the previously existing say-something-crazy-to-win model (and a CEO that's great at it). Another mistake is to turn their back away from the grassroots movement that supported the platform from the beginning.

Towards the end of 2022, developers saw an opportunity to build and promote alternative social networks. In different waves, each app tried to pull away as many Twitter users as possible; each attempt clashing against the insurmountable wall of network effects.

The clones only differed on the underlying protocol, usually a new decentralized spec: ActivityPub, Farcaster, AT, Lens, Nostr... Threads is the latest attempt. With Meta's power behind it, the app reached 100M users in 5 day becoming the fastest growing app in history. But its usage is already declining.

Why is it that no application is capable of breaking through? Twitter has never been that big either. With less than 500M monthly active users, it's far smaller than the likes of Facebook, Instagram, Whatsapp, Youtube or TikTok. Telegram, in direct competition with Whatsapp, has more MAUs. It seems that the network effects seen on a Twitter-like network are much stronger than that for a chat app.

Tomas Pueyo points to the fact that any standalone app that was a clone of a social app failed:

https://twitter.com/tomaspueyo/status/1676372381416161282

He claims a clone is only able to convince a small part of the user base to move to the new app but to succeed, they would need to convince everyone. That's not going to happen; that's how strong network effects are. Instead, big companies found success cloning as features inside their existing popular apps, (e.g. stories, reels, and shorts).

This only works if you have an existing social network with hundreds of millions of users. But if you’re not there yet, should you give up? Unless you're really determined, probably — as is the case with any other ambitious project. If you are determined though, there are some things you can try.

First, we should examine how each successful social network made it. Keep in mind they aren’t only competing with existing social networks but with all other attention-grabbing applications like games or streaming services. The only path to longstanding growth is by finding a format that encapsulates information in a novel way — at least in a way that no existing platform has already built cemented network effects around.

It's not enough for the format to be novel. Since the inception of social networks, there have been dozens of attempts at different ways of presenting text, images, videos, and more recently, audio. Most failed like Beme, Path, and Vine, and the jury is still out with Clubhouse. Others found success like Snapchat, TikTok, or Pinterest. It's a mystery why some formats work and others don't. New attempts like BeReal try to fill in the hole left by existing networks of a more intimate connection with your close friends. Will they work long-term? Only time will tell. The only learning we can extract is that we need to start with a format that's not already controlled by an existing giant.

There are 3 ways to find a format that can captivate people's attention. One option is to build new creation tools that make an old medium more accessible. TikTok and Snapchat made it really easy for people to create short videos, Twitch brought broadcasting to the masses and Instagram helped people take better pictures with their phones by using filters. Another course of action is finding a creative way to use an existing medium like how Twitter popularized microblogging or how Snapchat added AR filters. The third path is to embrace a new technology. Instagram wasn't possible until everybody had a camera in their pockets. Generally, we see platforms making use of all 3 methods. The older networks surfed the waves from Web 2.0 and mobile. Newer ones came from a maturing of the mobile space that led to more creativity by using the platform natively while also embracing AI.

These past few years we've discussed the upcoming intersection of 3 different technologies: AI, Crypto, and XR. These tools can be used for different parts of the stack. AI can aid as a creation tool. Generative AI can create a whole message in any format (text, video, audio, 3D) from a small idea. Decentralized protocols — not necessarily crypto-based — can create an open platform that brings people together while avoiding the limitations of closed platforms. Shortcomings like depending on a single company to provide creation tools, deciding on the parameters for an algorithmic feed and offering a viewer UI that's enticing. Crypto can also help discover new monetization methods. XR devices are helping commoditize 3D as a format. They are a natural interface to both create 3D content and view it. 3D is unexplored territory as a communication format, so I expect different applications to try different ways to make the most of it in the next few years.

Will the Twitter killer be a mix of these 3 new technologies? Not necessarily. It will be a short-form content network that chips away from the time spent on a quick scroll. There's also a lot left to explore with text. For example, I'm a big fan of Plexus that's playing around with AI and how to structure different messages to build on each other's ideas. I've actually been using Plexus to write initial drafts for my last several posts. Subconscious is doing something similar but with an approach that stems from a tool of thought that aims to become a big human brain. They are using AI to help you evolve your ideas and a decentralized protocol to combine your ideas with other people's.

With the fall of existing networks and the learnings from issues like privacy, attention hoarding, mental health, and mindless scrolling, together with the advent of new technologies, it's a great moment to think about how we can create better content and more meaningful connections. We communicate for many different use cases, and each format, and each implementation of these formats is good for different goals. Twitter is great to hear many people's thoughts about a variety of topics from a simple timeline that one can read from anywhere at any time. It could be a while until a new platform is able to take away our current Twitter time. But I'm confident it won't be a Twitter clone but a new way to communicate with short-form content.