Comment by iamnothere
1 day ago
Since you don’t consider that “good faith” I have rephrased it for you, in a format you may prefer:
> EUR02a. In the interest of protecting children, some politicians are calling for the automatic searching of all personal electronic mail and messages of each EU subject in the search for dangerous, illegal child pornography. Suspected cases will be notified to the police. An advantage of this could be that more offenders are caught and children protected. According to activists who defend child pornographers, police reports indicate that a few innocent people may be mildly inconvenienced due to unreliable processes. Please place yourself in the position of a law enforcement official trying to catch these evil people, who is currently obstructed due to false questions of “rights” and “privacy.” What is your opinion?
I jest, of course, I’m sure you would prefer something more straightforward and less manipulative like “EUR02a. Do you support child pornography?”
I understand that you're trying to dunk on me but I don't get what the point is supposed to be. It's certainly possible to come up with other bad ways to ask the question. If someone was interested in genuinely understanding the public's opinion, rather than including a demand to take some particular perspective, they would ask things like "Do you support online platforms scanning all personal messages for child pornography?" or "If you had to weigh the two, would you consider detection of child pornography or the right to privacy to be a higher priority?" And of course for the latter you’d randomize the order in case people are more or less likely to pick the first option.
I suppose the gold standard would be to present detailed arguments from each side with evidence (if any), for context. Barring that, the original question did seem to provide a rough summary of each side’s position. It may be weighted towards the anti Chat Control side, with the formulation “pro says this, but anti says this, and imagine that you are affected”. So perhaps they could have asked a reverse formulation 50% of the time to be more fair. But the poll was commissioned by Breyer, and of course he wanted to bolster his position.
Your context-free formulation on the other hand provides no information for voters to weigh. Privacy or child porn detection? Well I guess I’ll pick child porn detection. Oh, you wanted to do what to my privacy? Never mind!
Even your slightly longer formulation doesn’t really explain what scanning means and how that might affect people and society. Most people aren’t familiar enough with technical and legal details to dig into the implications without added context.
You could just write the simple effect of the law: "EUR02a. Facebook employees read your private messages. If they see child pornography, they will call the police."
It's important not to phrase it as "read your messages to detect child porn..." because that implies they won't do anything else with the messages, and Europe is a place where they still assume that if someone says they do X to do Y, they're only allowed to do Y and severely punished for doing anything else.
What scanning means, and how that might affect people and society, are precisely the issues in dispute. A poll which tries to provide arguments and evidence about contested issues cannot yield meaningful results. To pursue this line of reasoning, for example:
> Your context-free formulation on the other hand provides no information for voters to weigh. Privacy or child porn detection? Well I guess I’ll pick child porn detection. Oh, you wanted to do what to my privacy? Never mind!
you might ask questions like "How comfortable would you be with your personal mobile provider scanning your chats for child porn?" or "What would you consider to be an acceptable accuracy rate for such a scanner?". Then you could reasonably infer that respondents who say "not comfortable" or "90-100%" oppose Chat Control, because Chat Control will perform such scans and they're not 90% accurate. (Someone else who thinks the scans are 90% accurate might respond that your interpretation is wrong, but this dispute would not call the poll results themselves into question.)
1 reply →