So then how would people reach me? Ignoring snail-mail, all means of reaching me involve either my phone number or email address.

Perhaps you'd give them single-use invites, or invites that included some form of certification path? I've noticed that Discord makes me go out of my way to hand out a reusable/non-expiring invite URL instead of a 1-day use-up-to-10-times link; obviously Discord's implementation relies on their hosting all the servers, but one could imagine a cryptographic equivalent that worked in a federated protocol.

I don't use (or want to use) chat apps like Discord and such, though. It would also be a real challenge convincing the people I interact with routinely to start using such apps.

Sure; my point is that we can imagine email/phone-like systems that would work the same way.

My company recently switched to MSFT Teams, and the other day I was thinking how difficult it would be to get spam through teams.

By design, only people part of our organization can communicate with each other. There (AFAIK) is no endpoint for my organizational teams account visible to other teams users outside my organization. For spam to occur, someone's account would need to be compromised (which would be found quickly) or a fake account somehow created by the org admin (which seems unlikely).

With email, the IT department is commonly spoofed, or the president, HR, etc. Most people can spot the differences, but with teams they would virtually never get any fraudulent messages.

Yes, though internal spam is a thing too. Chat tools are the prime driver of it

I suppose “social networks” are the replacement, you’re exchanging getting spammed by random actors you don’t know to a single actor (the social network).

Aaaaand here we go reinventing the wheel of "trust graph." Key signing party anyone?