During these boundaries people can die. Consider the advent of yellow journalism and the connection with the Spanish-American war 1898: https://en.m.wikipedia.org/wiki/American_propaganda_of_the_S...

Outside of the transition to a large city, virtually everything you've mentioned happened in the last 1/2 century. Even the phone was expensive, and not widely in use in under 100 years ago.

That's massive fast change, and we haven't culturally caught up to any of it yet.

Just because we haven't yet destroyed the human race through the use of nuclear weapons doesn't mean that it can't or won't happen now that we have the capacity to do so. And I would add that we developed that capacity in less than 50 years of creating the first atomic bomb. We're now living on a knife's edge and at the merge of safeguards which we don't give much thought to on a daily basis because we hope that they won't fail.

That's how I look at where we're going with AI. Plunge along into the new arms race first and build the capacity, then later figure out the treaties and safeguards which we hope will keep our society safe (and by that I don't mean a Skynet-like AI-powered destruction, but the upheaval of our society potentially as impactful as the industrial revolution.)

Humanity will get through it, I'm sure. But I'm not confident it will be without a lot of pain and suffering for a large percentage of people. We also managed to survive 2 world wars in the last century--but it cost the lives of 100 million people.

I tend to think the answer is to go back to villages, albeit digital ones. Authentication only enforces that an account is accessed by the correct "user", but particularly in social media many users are bad actors of various stripes. The strongest account authentication in the world doesn't help with that.

So the question, I think, is how do we reclaim trust in a world where every kind of content can be convincingly faked? And I think the answer is by rebuilding trust between users such that we actually have reason to simply trust the users we're interacting with aren't lying to us (and that also goes for building trust in the platforms we use). In my mind, that means a shift to small federated and P2P communication since both of these enable both the users and the operators to build the network around existing real-world relationships. A federation network can still grow large, but it can do so through those relationships rather than giving institutional bad actors as easy of an entrance as anyone else.

> Each of these transitions caused big problems. None of these problems have ever been completely solved. But each time we found mitigations that limit the impact of any misuse.

This a problem with all technology. The mitigations are like technical debt but with a difference. You can fix technical debt. Short of societal collapse mitigations persist, the impacts ratchet upward and disproportionately affect people at the margin.

There's an old (not quite joke) that if civilization fell, a large percentage of the population would die of the effects of tooth decay.