I can definitely see these systems being used illegitimately, but this has nothing to do with whether you can face your accuser. A software company telling you that you cannot validate the methodology of their measurement instrument is definitely troubling, but it is not fundamentally dissimilar to the police using speed guns whose internals you cannot inspect to convict you of speeding. Even more problematic than either of these technologies is the fact that the cops jail motorists for DUI on the basis of their own subjective interpretation of those motorists' performances on sobriety field tests.
The use of opaque technology by law enforcement raises significant issues in a free society. But so does the use of opaque non-technological judgement by police officers. Since there are significant benefits to automated enforcement, we should be pushing to improve transparency, rather than pursuing a chimera where we can only be prosecuted on the basis of transparent and reproduceable forensic analysis.
> there are significant benefits to automated enforcement
Could you elaborate on what you think the benefits are?
I'm of the opinion that systems that indiscriminately check whether all people nearby are currently committing a crime should not be things we should be building in our society. I would prefer there be a reason to suspect a person of a crime, and pursue on a targeted basis, rather than simply have panopticons charging people with crimes.
For example- with the system in this article, I would prefer if a person were required to call in and complain "there was a too-noisy car at this location at this time", and then a person reviewed footage from that location at that time and determined whether to send a ticket out or not.
Regarding your points about transparency, I completely agree with you.
> Could you elaborate on what you think the benefits are?
Very simply, I don't want people driving around New York with blown-out mufflers. I have to deal with these people all the time. They are a constant, low-level annoyance that lowers quality of life for everyone in the city. Presently, there are presently no consequences for this sort of selfish behavior, and I'd like that to change by ticketing the perpetrators.
I don't understand what the added value is in making ordinary New Yorkers call the police with a time and cross-street as a precondition for a summons. What are we gaining here? Are we trying to validate the existence of a victim? Of course there is a victim -- there is hardly a block of the city where a loud motorist will not be negatively impacting a resident's life. This place is incredibly crowded. People live almost everywhere. Are we assuming that a manual reviewer is going to be more effective than an automated system? How? At the end of the day, they're going to have to base their judgment on a decibel reading anyways.
The use of opaque technology by law enforcement raises significant issues in a free society. But so does the use of opaque non-technological judgement by police officers. Since there are significant benefits to automated enforcement, we should be pushing to improve transparency, rather than pursuing a chimera where we can only be prosecuted on the basis of transparent and reproduceable forensic analysis.