So you'd trust a human to program a machine, but you don't trust a human to observe the event? Seems laden with hypocrisy. If you believe the cop is fallible and/or biased, why wouldn't the programmer also be fallible or biased???
And since when have camera systems been foolproof?
Machines typically don't have emotions, or bad days.
A human can be a good person, or a bad person, at different times during the same day, and you can't tell the difference until you see their actions play out.
The programmer has a specific job to do:
Make this gizmo do this specific thing when it sees a vehicle drive down the road.
If it sees the vehicle doing x, y, or z, then execute these specific commands.
Then, the gizmo is tested in a variety of locations and conditions.
If the gizmo shows reliability, and accuracy, then the gizmo is distributed, and used.
If the programmer decides to add in code that if the vehicle is red, or if the vehicle has a Carolina Squat, then execute different code, and that code is determined to be illegal, then the innocent civilian can pay someone to expose that code in court.
It's still made really difficult to prove that a police officer is anything other than a model, upstanding example of law enforcement, and he has the full backing of his department and union rep in almost all circumstances.
It's typically easier to get justice against a gizmo that is operating outside of its design specifications than it is to get justice against a police officer, police department, or justice system that is operating outside of legality.
I would much rather refuse any testing that is subjective, keep my mouth shut, and let myself and my fluids be tested by devices and laboratories that are definitive.