The discussion referenced by Defense News though is policy around the deployment of Squad and Battalion level forces which are typically engaged in situations where either the nation-state isn't the target (say Somali pirates, of Afghan warlords), and/or the goal is incompatible with mass destruction.
Your hypothetical suicide inference leaves out the choice of compliance. One could look at a place like NotAmerica' where no one is at risk of horrific death due to killer robots, versus NotAmerica where death by remote action is a more common occurrence, and look at what is different between the two places. In all the theaters of the world today it's pretty clear what action would completely remove that threat of horrific death. (hint: it isn't going to America and blowing yourself up.)
The reason stock markets have black swans is because legal transactions can never account for illegal activity. Sure, you can buy replacement insurance against theft, but you cannot actually build theft into an accounting model. Stock markets, as social laboratories of economic exchange, will always have crashes because there are black markets that operate behind them in secrecy. Certain companies might befall "unfortunate accidents" and whoopsie daisy! How did that happen?
Consider the question: How interested are you in being arrested and imprisoned by a robotic civilian police officer?
Basically exactly the same as I'm interested in being arrested and imprisoned by a meat-cop.
And the robocop might win just based on the fact that he's less likely to have a "respect mah authoritai" chip on his shoulder that results in me being tazed or shot for no good reason.
This is about power differential. Imbalances of power contribute to instability. The kinds of power discussed in this equation are vast, and not to be trifled with. Errors made with the toys of modern warfare can be unforgiving to the point that submarine personnel might be the only survivors.
When you scale up conflict beyond the individual concept of arrest and imprisonment, there is a volatility to open warfare that is whithering, when compared to civil society. If you're that eager to thwart some cops in their activities, then vote a draft into law, advocate mandatory military service and then go rob a bank and shoot some tellers while you're at it. I dare you.
Well, yes. But that's primarily because I am not currently confident in the law.
A robotic crime wave, or a human crime wave?
A robotic miscarriage of justice, or a human miscarriage of justice?
I think, in both situations, the two are not equivalent, and the robotic version is not only worse, but carries the potential of total devastation. But maybe I'm chicken little.
Concerning the film "Robocop", it was interesting that you discover most of his character traits after he has been transformed into a cyborg, and my opinion is that he was a brutal copper.
What makes you think that a robot couldn't be programmed to show malice, racism or cruelty?
NB I am not implying that the US would actually do this, just questioning whether there are really any fundamental barriers stopping the creation of machines with such malign goals.
Erm, this is why nuclear deterrents work.
For the purposes of a thought experiment, consider the false dichotomy of "America" and "NotAmerica" as two hypothetical opposing forces, where America has nukes and killer robots and NotAmerica has only nukes.
If you find yourself living in NotAmerica, your motivation to pre-emptively nuke America, simply to avoid the threat of a horrific death dispensed by killer robots, presents itself as an absurdly simple choice. Even if it means suicide.