2 ms·
I don't know what has been posted before this all I see is [flagged]. >posting publicly and voluntarily with the expectation of moderation I find this stateme
by EMCymatics 3y ago
I don't know what has been posted before this all I see is [flagged].
>posting publicly and voluntarily with the expectation of moderation
I find this statement a bit difficult when it is perhaps a duty to participate in the public square or whatever square is used to embody the desires of common or public will. If people are deterred from this arena without cause and abuse then perhaps we have failed. This entire ecosystem rests on the belief of decency or fairness. I have yet to see it.
>The issue here is the public's expectation of privacy in certain situations. In fact, no one (or very few people, anyway) dispute the need for states to surveil sometimes. The question is when and what oversight should be required. And in this case the the argument is that oversight was sidestepped.
That is incorrect. During the oversight committee they identified there was a consistent mis-interpertation of the law (now corrected) from one side to the other under 702 which perhaps enabled unauthorized use (which perhaps was appropriate under other sections). It was up to (not just due to mis-interpertation) 30% of searches in error in the potential to be in the millions of incorrect searches. I dont see much oversight for other codes. We have possibly 10,000 people with access to obtain information and we are not close to consistent oversight given moving codes and authorizations. Congress has failed the bodies that keep us safe but not without some significant non-compliance. [1]
Private access to information probably has great advantage over public ones. I believe we should consider carefully how we want to yield the power the public sector has given us. The potential for private and public conflict on this matter is significant.
We should not shoo people away with differing opinions. If language models are the future for communication and teaching while they are learning with data off of 'public' and 'accurate' discorse we are very much in danger of having a future led by biases inherent in these models (from moderation and 'weeded garden' environments) that 'isnt a problem' to the people creating the future.
[1] https://www.youtube.com/watch?v=pIEqAjRKVyE https://www.youtube.com/watch?v=pIEqAjRKVyE