I think it's legitimate for companies to implement automated systems, such as CSAM and spam filtering, to limit the amount of unwanted material on their networks. I don't have any problem with Apple, Google, and Microsoft, checking the hashes of files I upload (or attempt to upload, in Apple's case) to their servers against ICSE. I would have an issue if employees of those companies had unfettered, unaudited access to users files.
Outside of giving my opinion of a specific proposal I don't know what you expect me to say.
Perhaps you could describe your own "limit" to how much avoidable suffering is acceptable to you before you would support automated scanning of uploads. I don't personally believe it's possible to precisely explain an overarching "limit" in situations that balance competing moral and philosophical concerns.
---
I'm being rate limited now due to downvotes, might not be able to respond further
Sorry you got rate-limited. On HN, new accounts are rate-limited by default because of past abuses by trolls. I've turned that off now for your account (I'm a mod here).
It's probably timing that is limiting your responses. AFAIK downvotes won't do that. And I didn't down-vote you.
I accept your answer, and acknowledge that different perspectives are valuable.
My own opinion is far less relevant as I probably will not be implementing or executing systems that target information from large swathes of a population for inspection by my organisation.
Nevertheless, let me give it a shot. I asked you about the social cost of false-positives. You reciprocated by asking me about false-negatives. It's tough - on both sides of the equation.
I try to apply weighting. If the dragnet would be targeted to a limited number (say 100) then it could easily be justified, since the relative horror of one over the other is surely in that ball park. Maybe even 1,000.
The problem for me is that mass surveillance such as the subject of this discussion is not numerically constrained. It's trawling the entire ocean floor for the one or two species that may be legally caught.
I think it's legitimate for companies to implement automated systems, such as CSAM and spam filtering, to limit the amount of unwanted material on their networks. I don't have any problem with Apple, Google, and Microsoft, checking the hashes of files I upload (or attempt to upload, in Apple's case) to their servers against ICSE. I would have an issue if employees of those companies had unfettered, unaudited access to users files.
Outside of giving my opinion of a specific proposal I don't know what you expect me to say.
Perhaps you could describe your own "limit" to how much avoidable suffering is acceptable to you before you would support automated scanning of uploads. I don't personally believe it's possible to precisely explain an overarching "limit" in situations that balance competing moral and philosophical concerns.
---
I'm being rate limited now due to downvotes, might not be able to respond further