Make Twitter great again — a guide for Elon Musk
I for one am optimistic about Twitter, he’s a smart guy with not a lot of respect for conventions.
Some people are concerned that a single person would control such an important medium, but when I think of it — no one asked me how to run Twitter before Elon either, so not much of a difference from that perspective.
Anyway, even titans of industry could use a hand from the simple folk. So I’m writing this guide for Elon Musk, to make sure he does indeed make twitter great again.
First, some observations on human nature:
- There are things that will only be said with anonymity guaranteed. Whether it is too personal, or too embarrassing or may have dire consequences due to political oppression. We are an embarrassed species, and there’s a lot of thoughts and feelings we all share that we can’t talk about in the open.
- People have different moods which require different content. They go into the same website (e.g. twitter, reddit) at different times looking for different things, or broadcasting on different frequencies. These moods are not always as explicit as subreddit communities. Sometimes they are more general or subtle than that — e.g. a desire to be entertained vs a desire learn something or a desire to simply connect to other humans. Different posts will be a hit for one mood but a miss for another mood, for the same person.
- People associate social media accounts to certain topics. Even though there are complicated human beings behind this account, accounts typically converge to a certain type of communication that their followers expect.
- People can be manipulated by certain UI motifs. Inconsistent rewards is one famouse example, a notification dot is another. It’s easy to control people’s dopamine using a well crafted UI, but it does not correlate to long term well being.
Let’s add to that a couple of observations about technology:
- It is incredibly cheap to transmit text between humans. This means that business models of twitter are not so important. It’s enough that it’s financially viable to keep the servers running, and a small team of dedicated engineers running it.
- The true democratization of social media is a serverless scenario. As long as there is a centralized place where servers reside, or an organization that owns the app, then it is vulnerable to governmental intervention in the form of retribution, censorship or propaganda. In western democracies this is more of a theoretical pain, but in other places, it’s a matter of life and death. If the app is distributed it makes things a lot safer, and a lot harder to stop. In essence it is the application of the fundamental idea of democracy to the software industry.
So what should we do with this? First of all, we can’t have everything right away, so don’t expect that. Poor Elon has his work cut out for him. But there are some things we should have in the immediate future to get one step closer:
- Make it really easy to privately create and operate multiple identities under the same account. This is not easy enough today in Twitter. To make it easier we should not require a different email for each identity. We should allow consuming the feed without committing to a specific identity, and not require a strict switch. You should be able to comment using two different identites without too many clicks. It should be technically impossible (or at least really hard) to link two different identites, unless the owner chooses to… reintegrate?
- Keep making it easy to collect explicit feedbacks on a post. The rise of the negative signal in social media is probably something I should write more about, but in broad terms, it is essential that users can swallow the click bait, then go back to their feed and signal that it was not worth their time. Today it is done via the “show less content like this” button (one extra click to reach it though). This allows people to conserve their willpower and avoid teaching the algortithm the wrong thing.
- Feed algorithm should be open source. Elon got that one right already. This will make it is harder to manipulate the feed, and easier to understand what the feed is maximizing on, so public criticism can address it, and PRs can fix it.
- Start researching truely p2p social media technologies. Ultimately this is the future and a prerequisite for freedom, because it means the app itself is not controlled by any single entity. We can’t have this immediately, and it will take some time to imagine how to do well. Obviously we should not sacrifice eco-friendliness for this as bitcoin has.
- Do not abuse the notification dot anymore. Use the notification dot only for what users expect — mentions, replies on their threads and the likes of that. Don’t turn it into a recommendations feed or any other explorative/ML driven notifications. The way it works now it causes a serious alert/notifications fatigue, e.g. what goes on Facebook. Leave the notification dot alone!
- Allow people to communicate their mood by making the feed non linear. E.g. if I go left, I want a more humuristic time, on the right, tech geekiness.
- Allow an unlimited length of posts. Who the hell thought this is a good idea to limit it anyway. We need more in-depth conversations.
- Allow editing. Figure out the UX, you are smart enough to make it easy to use and hard to abuse.
- Audio spaces are great, keep working on them. Smaller rooms are where the magic happens.
- Bots/automated accounts are not necessarily bad, but should be curbed. We need to start getting used to the idea that bots and humans are not so easily distinguishable. However for the time being and the political climate, we should prevent individuals or organizations from flooding us with their messaging. This should be done by making it too expensive to be viable, by a combination of captchas and correct feed editing (hence the open source’s importance).
- Twitter works because it’s simple, keep it simple. Remember the old FB UI? So distracting and colorful, so many menus and submenus, so many colors. It’s vital to keep Twitter (and any other product for that matter) doing a minimal set of things.
- Figure out when a user has spent more time in their feed than they would have wanted to and help them prevent it from recurring. A simple way to do that is to ask after the session has ended, and potentially throttle the amount of posts that can be visible in one session.
- Keep censorship to what is minimally required by law. Another one that Elon got right. Ultimately we want people to say whatever they want. Not saying something doesn’t make it unthinkable. As long as Twitter is a centralized organization with servers and all that, it needs to adhere to governing laws. Making it fully distributed will allow the next level of democratization, but meanwhile censorship should be kept to a minimum. In general censorship is not the right way to fight dangerous ideas. Ideas behave a lot like viruses, and as covid-veterens we should know it is more effective to immunize people than to quarantine them, hoping that they are never exposed.
Good luck Elon!