Werd I/O

In my Twitter security post I mentioned that you should delete DMs. Ted Han pointed out to me that you can only delete them from your own inbox: the other party can still see them, and they’re consequently still available on Twitter’s infrastructure. It may still be worth deleting them, but only to prevent someone from finding them if they break into your account.

A good idea going forward, though, is not to use DMs and go to another platform. For example, Signal is a great solution for encrypted messaging.

Yesterday I sent a memo to all staff advising them on the situation at Twitter and how it pertains to their own security. I thought it might be useful to share a version of this information with you, too.

Twitter’s Chief Information Security Officer, Chief Privacy Officer, and Chief Compliance Officer all quit on Wednesday night. One can reasonably infer that the team at Twitter is being asked to do things that these people were not comfortable with, and given their roles, it’s reasonable to consider Twitter to be insecure going forward.

At the same time, it’s still where a lot of people find community and reach. You might not want to leave it right now. Here, then, are some suggestions about how to stay safe while remaining on the platform.

Enable non-phone two-factor authentication. Twitter allows you to log in with two-factor auth. Using your phone number leaves you open to having that number leaked - or used for other purposes by the company - in the future. I always recommend using an authentication app. Authy is a good stand-alone app, but this functionality is also built into password managers like 1Password.

Remove your credit card number. If you’ve bought ads, remove your payment details from the system. We know that credit card numbers are stored insecurely on the platform.

Remove sensitive DMs. DMs on Twitter are not encrypted. They could be leaked or mined by the company for other purposes.

Use a password manager to generate your password. Don't try and use a password you've invented yourself. And don't share this password with any other system: it’s just for Twitter.

Use a canary email address, if you can. All Google-powered email addresses can have arbitrary labels added to them using a +. For example, the address ben+twitter@werd.io will still get to me - but if I use that label on my email address in my Twitter account, I'll know my account has been compromised when other entities start using it.

Post via the web or using a third-party app. We know that Twitter tracks very detailed location information from its app users. Web browsers keep you safe, so posting via the web does not carry the same risk.

Two service updates I’m interested in:

Substack has introduced a “bestseller” badge for its top-selling newsletters, as a sign of quality. Newsletters on Substack don’t need to have paid tiers, but this badge is only given out to newsletters that have a lot of paid subscribers. The implication is that non-paid newsletters will experience FOMO - and potentially a drop in new subscribers because they don’t have the badge - and try to climb the ladder. It’s the first time that Substack, to my knowledge, has tied quality visibly to revenue.

Twitter has started selling its blue “verified” badge for $8 a month, replacing the previous system that gave badges to notable individuals in order to protect them from impersonation and abuse. Twitter will also demote tweets from non-verified users, treating them like “email spam”, effectively meaning that you’ll need to pay a minimum of $96 a year to get reach on the platform. Again, this ties revenue to quality: the only people worth being heard, according to Twitter’s new leadership, are people with money. Notably, the feature is only available on iOS in the “US, Canada, Australia, New Zealand, and the UK”, which fundamentally limits the global inclusivity, and therefore the value, of the platform. It’s an odd choice. Based on precedent from Musk, I expect it to be partially rolled-back on a whim, but it’s here for now.

The danger with both plans is that only people with money can be heard. A Substack newsletter by someone with no existing following is unlikely to make it into the bestsellers list; therefore, people who already have privilege and power will find it cemented by the platform. And while $96 might not seem like a lot of money from an upper middle class American perspective, for most people in the world it’s an unjustifiable cost. Without mitigations, we’re likely to find ourselves in a world where voices from vulnerable populations remain underheard or even suppressed.

That’s clearly a problem from an equity and inclusion standpoint - and therefore also in terms of reader experience. The narrower the context and demographics of the people who produce most visible content are, the more homogenous it becomes, and therefore the more boring. We all gain a lot by being exposed to people from contexts and backgrounds different to our own. Not only is building a platform that elevates diversity the right thing to do ethically, it builds a healthier community with more interesting conversations. It’s better all the way around.

I think we’ll see the two platforms diverge in their approaches. Substack will quickly figure out that it needs to elevate different voices to continue to grow: audience diversification is key to it. Twitter, meanwhile, will probably not figure this out in a way that it can act on, not because there isn’t a way for it to be acted on, but because it seems to have lost all semblance of acting on a coherent strategy or mindset under Musk. Substack has some real editorial ethics problems to contend with, but it seems to be adept at the fundamentals of community growth.

I also want to call out that Medium under Tony Stubblebine is figuring out these problems in a very thoughtful and transparent way. As Tony points out in the last link, the quality bar is not up for debate; the question is, who’s in consideration to find an audience to begin with.