When customer support teams stop using Twitter, that's when you know a key component of Twitter's former value is finished.
Pre-Elon Twitter often showed off what airlines could do with customer service through their platform. Now that Elon owns Twitter, they're jumping ship.
Still, there's a big need for customer service to utilize microblogging -- and I suspect these airlines will migrate elsewhere. Maybe not now, but eventually.
https://gizmodo.com/twitter-airlines-air-france-klm-air-travel-1850418687
potential seems unlimited for corporate #fediverse apps innovating re the customer service.
You would run into privacy problems since all Fediverse content is public so there's no way for a person to communicate with customer service without their business being broadcast.
Visible only to the admins of their servers (just like Twitter). You're not suggesting that admins go snooping around in people's DMs are you?
Visible to admins of every server the message passes through, which is definitely not just like Twitter, as that unknown collection of individuals won't be in a position where they can be held legally accountable for breaking agreements made with the company.
A company can sign an agreement with Twitter to provide privacy standards. In a system like this there is no point of contact to make that agreement with, and the system is designed from the foundation to broadcast publicly, so such an assurance isn't really possible anyway, even if there was someone to grant it.
What are you on about? If I DM you—and only you—the message goes from my server directly to yours, not all over the Fediverse. Moreover the message is transferred via an API over HTTPS, an encrypted secure link. My admin and your admin are the only people other than you and I that could see it—and I know my admin has better things to do with his life.
Well the ActivityPub protocol is a little more complicated than that, but even just based on what you describe, well it's great that you trust the admin of your own instance, and that you trust his ability to trust everything about his setup, but you don't know my instance admin. You don't know what my guy might be doing with that DM.
And imagine you want to delete some content. That's great that you trust your admin to honor the delete request, but when he sends it over to my instance, my admin isn't under any obligation to honor the delete request.
Heck, my admin might for whatever reason think he is doing the world a favor by maintaining those records.
And that's just talking about the simple case of a DM style post. Anything with any larger audience only compounds those issues of trust.
These privacy issues matter. Or at least, users need to be aware that these weaknesses exist, and legal implications need to be considered as well.
That's true. There's a one sided relationship of trust between users and admins that can't be mitigated.
I manage that by never posting anything here that's truly private. If I need real privacy there are better solutions.
That's the key. I'm just really afraid that so many users of this system don't realize and post private things they wouldn't post if they knew how it worked.
I really hate to think about users being mislead, with that false sense of security.
It's very difficult to achieve real security with online messaging. Take WhatsApp for example. With commercial E2EE the app is closed source and they invisibly control distribution of the public keys used in a conversation. You trust that the "ends" your messages are being encrypted for and sent to are the only ones visible in the app. But you don't actually know that.
Right but one issue is that users aren't aware of how insecure this platform is. I know, I've come across a few people who are really surprised when I point it out.
Yes, security is hard, but users need to know what they are getting into here, regardless of how hard it might be to design a more secure system.
