> My view is that there are hard to quantify risks that it is rational to have some fear about and see as problems to be solved.
Heavily agreed. But those are not irrational fears.
They become irrational fears when learning more about the risks and learning more about the benefits and downsides of different mitigation techniques doesn't change anything about those fears one way or another.
We all form beliefs based on incomplete information. That's not irrational. It is irrational for someone to refuse to look at or engage with new information. If someone is scared of the potential for encryption to facilitate CSAM because they're working with incomplete information, that's not irrational.
If someone is scared of encryption because they have incomplete information, and they refuse to engage with the issue or to learn more about the benefits of encryption, or the risks of banning it, or what the stats on child predators actually are -- at that point, it's an irrational belief. What makes them irrational is the fact that they are no longer being adjusted based on new information.
A rational person is not someone who knows everything. A rational person is someone who is willing to learn about things when given the opportunity.
> Why do you mention this again? Nobody is arguing for a solution designed to assuage people’s fears.
I guess I don't understand what you are arguing for then.
Let's look at your "reasonable people who are reasonably afraid" camp. We'll consider that these people have doubts about encryption, but don't hate it. They are scared of the potential for abusers to run rampant, but are having trouble figuring out what that looks like or what the weak points are in a complicated system. They are confused, but not bad-faith, and they have fears about something that is legitimately horrific. We will say that these people are not irrational, they recognize a real problem and earnestly want to do something about it.
There are 2 things we can do with these types of people:
1) We can educate them about the dangers of banning encryption and encourage them to research more about the problem. We can remain open to other proposals that they have, while making it clear that each proposal's social benefits have to be weighed against their social costs.
or
2) We can offer them some kind of compromise solution that may or may not actually address their problem, but will make them feel like it does, and which will in theory make them less likely to try and ban encryption.
You seem to be suggesting that we try #2? And this apparently isn't designed to assuage their fears? But I'm not sure what it does then. Presumably the reason they'll accept your proposal is because it addresses the fears they have.
My preference is to try #1. I believe that if someone is actually in the camp you describe, if they have reasonable fears but they're looking at a complex social problem, openly talking to those people about the complex downsides of banning encryption is OK. They'll listen. They might come up with other ideas, they might bring up their own alternative solutions. All of that is fine, none of us are against reducing CSAM, we just want people to understand the risks behind the systems being proposed.
But importantly, if someone is genuinely reasonable, if they aren't irrational and they're just trying to grapple with a complex system -- then talking about the downsides should be enough, because those people are reasonable and once they understand the downsides then they'll understand why weakening encryption isn't a feasible plan. From there we can look at alternatives, but the alternatives are not a bargaining chip. Even if there were no alternatives, that wouldn't change anything about the downsides of making software more vulnerable. First, people must understand why a proposed solution won't work, and then we can propose alternatives.
To me, if someone comes to me and says, "I'm not interested in hearing about the downsides of banning encryption, come up with a solution or we'll ban it anyway" -- I don't think that person is reasonable, I don't think they're acting rationally, and certainly I'm not interested in working with that person or coming up with solutions with that person.
> If we don’t offer an alternative solution we aren’t offering them anything at all.
Where I fall on this is that I am totally willing to look for alternative solutions; but encryption, device ownership, privacy, and secure software -- these are not prizes to be won, conditional on me finding a solution.
We can look for a solution together once we've taken those things off the table.
Because if someone comes to me asking to find a good solution, I want to know that they're coming in good faith, that they genuinely are looking for the best solution with the fewest downsides. If they're not, if they're using encryption as some kind of threat, then they're not really acting in good faith about honestly looking at the upsides and downsides. I have a hard time figuring out how I would describe that kind of a person as "reasonable".
> I personally think you would be very disappointed
> Why would you think this? Did I say anything anywhere about convincing people who are arguing for weakening encryption?
Let me be even more blunt. I think that you could come up with a brilliant solution today with zero downsides that reduced CSAM by 90%. And I think you would be praised if you did come up with that solution, and it would be great, and everyone including tech people like me would love you for it. And I also think it would change literally nothing about the current debates we're having. I think we would be in the exact same place, I think all of the people who are vaguely worried about CSAM and encryption (even the good faith people you mention above) would still be just as worried tomorrow. You could come up with the most innovative amazing reduction strategy for CSAM ever conceived, and it would not change any of those people's opinions on encryption.
I'm not just talking the irrational people. It would not change the opinions of the reasonable people you're describing above. Because why would it? However good your solution is, if encryption is genuinely not worth preserving, then it would always be better to implement your solution and ban encryption. I don't say that derisively, if the benefits of banning encryption really did outweigh the downsides, then it would genuinely be good to get rid of encryption.
The only reason we don't get rid of encryption is because its benefits do heavily outweigh its downsides. Not because this is some kind of side in a debate, but because when you examine the issue rationally and reasonably, it turns out that weakening encryption is a really bad idea.
> My argument is that the best way to make things better is to make better options available.
This is another point where we differ then.
As far as I can tell, any reasonable person who is convinced that encryption is a net negative is always going to be interested in getting rid of encryption unless they understand what the downsides are. Any reasonable person who is on the fence about encryption is going to stay on the fence until they get more information. I don't see how proposing alternative solutions is going to change that.
So I believe that the only way these reasonable people you describe are going to change their minds are if they're properly educated about the downsides of making software vulnerable, if they're properly educated about the upsides of privacy, and if they're properly educated about the importance of device ownership.
And maybe I'm overly optimistic here, but I also do believe that reasonable people are willing to engage in good faith about their proposed solutions and to learn more about the world. I don't think that a reasonable person is going to clam up and get mad and stop engaging just because someone tells them that their idea to backdoor software has negative unintended side effects. I think that education works when offered to reasonable people.