Cool

There's a really bleak movie called good kill that captures this in a military context.

And here's the real life version. https://www.nbcnews.com/news/investigations/former-drone-ope...

Anyway, this sounds like a very terrible idea.

Even absent all that, I don't think I'd trust either remote fire control nor AI fire control for the stated scenarios.

But, I remember the Gell-Mann Amnesia effect, and I wonder what the real story behind this article is?

I see two potential problems though: 1) security, 2) sooner or later - in 20 or 50 years - someone will have the brilliant idea of making one of those "autonomous". This would be the beginning of a long nightmare.

The worst mass shooting in history was made by a guy dressed up as a cop, there is no way to tell if anyone is a good guy before they start shooting.

2 to 5 years

A DAO hires gig workers to assemble a "product" in a property which it has rented, the product is then programed by the DAO to attack the city it resides in. The DAO having speculated on shares of competitors of companies within the city, moves ever onwards. Terrorism as a software glider gun.

Its a major theme in my game, any exponential tech advances handed to citizens, become a minority vote mechanism on society. The implications being, that freedom and exponential toys bump into a hard roof, were any clever crank can empty exponential weapons down the hallway. But society as we know it also can not continue without exponentials tapped. A rock and a hard place.

This is called "stochastic terrorism". It has carried out at least two attacks in the United States in the past week. Doesn't require fancy technology, just very effective ideological poisoning.

The ones which had made it through to UK twitter were the Walmart one (not ideological per se, but the workplace massacre is a distinct American phenomenon) and https://en.wikipedia.org/wiki/Colorado_Springs_nightclub_sho... which definitely was ideological.

Is that because the ones we don't think are dumb get called "freedom fighters"?

Whether they get called freedom fighters or terrorists depends more on their race and religion than their capabilities.

Maybe they were terrorists...

(and of course the big outlier in terrorism, 9/11)

Comp science is obvious. Physics is similiar - if you have a working fusion device (ripped from a flying car) or any other device with exponential energy density, it amplifies the problems you can solve and cause.

Biology is harder to grasp, no not really after covid. But in theory one day not far away somebody might take a backyard shed and cook something up, which is not even a disease that has "exponential" castrophe written all over it. Imagine a floating Lema https://en.wikipedia.org/wiki/Lemna, blue, bio engineered to harvest carbon, all good intentions, wrapped around a bubble of methan for buyoncy. Such a plant, engineered by anyone, could block out the sun.

Elysium also offered some interesting scenarios on automated policing.

Something like that. I think there was a similar interaction at the police station. Automated responses, no grey area, computer says no.

1. SERVE THE PUBLIC TRUST 2. PROTECT THE INNOCENT 3. UPHOLD THE LAW

—- 4. CONFIDENTIAL (blink blink blink)

RoboCop: Director's Cut | REMASTERED - ED-209 Malfunction Scene (1080p)

https://www.youtube.com/watch?v=ZFvqDaFpXeM

Hostages or no, the desired result is generally no human casualty, perp or otherwise.

/s just for shiggles; i do appreciate even remote situations may be incredibly stressful - ask a drone pilot/ bomb okay-er or two. Dunno about cops.

It's about risk of loss of life to members of the public or officers is imminent, not for the officer controlling the robot.

when the most well funded and most sophisticated companies in the world are compromised on a fairly regular basis, who tf thinks this is a good idea?

if the engineers who built this aren’t actively speaking out on what a terrible idea this is, then we immediately know either their judgement is entirely flawed with hubris or they just plain can’t be trusted.

Makes me wonder if technologies, like armed autonomous robots, where a single person can easily control an overwhelming military force can potentially lead to erosion of the democracy in the western world.

This view of history needs to account for the nonfree people and how, despite the ambient tech level, they were kept unfree for a century after independence.

Now, that power needs to be distributed properly. If there were a super-villain with an overwhelming military power under direct personal control, (s)he would be able to dictate some terms, but not be able to control all the way down to individual behavior. That would require micromanaging at an impossible level. So, to completely subvert society, you need an army with a hierachy.

You have an historical mistake that supports your argument here.

Bronze weapons were very expensive, due to needing tin which could not be mined locally and needed to be traded for from as far away, such as the balkans, Germany or even Britain [0]. Because of this it was only available for the rich, ie the rulers and their supporters. Bronze age Greece[1] was not democratic, and was based around kingdoms, ruled by those who had access to the bronze weapons.

The later more 'equal' (but by no means equal at all, simply more so than ruled by tyrants, they still had slaves and oppressed lower classes. It's just there was a more numerous upper/middle class) greek states were in the Iron Age. Iron was much easier and cheaper to make as it was available in large quantities locally. This lead to to the iron age weapons and armour being much more widely spread [2] and therefore they had less powerful dictators.

Not all Iron age Greek states were democracies, and those that were would not seem like fair democracies today. But partially because of the more easily accessible weapons, the power in these states was less centralised than it was when the weapons were predominantly bronze.

[0]: https://en.wikipedia.org/wiki/Tin_sources_and_trade_in_ancie.... [1]: https://en.wikipedia.org/wiki/Mycenaean_Greece [2]: https://en.wikipedia.org/wiki/Greek_Dark_Ages

I wonder what the stats are about police needing to draw a weapon to protect someone other than themselves? (Not being facetious, genuinely wondering about the data in order to help form an opinion).

I'm not sure if this kind of rational/utilitarian approach to thinking about policing is useful, because soon you'll end up with a perfect surveillance state and robo executioners for "the greater good". Maybe throw some precrime work in there as well, because statisticians for example have found correlations between certain looks of people and likelihood to commit crime. What if statistically the benefit outweighs the harm, i.e. more harm is prevented than suffering of innocent people being wrongly accused? Should we then lock up people to prevent crimes they haven't even committed yet?

Killing somebody just because they're running from an outstanding warrant would be a new and terrible policy whether or not a robot was involved.

Who guarantees you they wouldn't do it at home if they're now given the same capability on US soil? A terrorist on the run - they could take them out by robot instead of having to go through the arduous process of an arrest and court of law. Maybe he didn't comply with orders given by the robot and ran. You see?

Well it’s illegal for starters.

Same reason they shipped terrorism suspects to Cubs and tortured them there. It would be illegal to do that in the US. Obviously it’s a bit weird that government officials/military can do all sorts of things overseas that would be illegal back home. It’s not something new though, US has been doing this since it first started intervening in other countries back in the 19th century.

The thing is: What could ordinary citizens do about it? The government isn't going to investigate itself, especially not a popular president that wasn't even under much pressure to do so. He brought so much change, just not the good kind: https://upload.wikimedia.org/wikipedia/commons/d/d1/Graph_of...

If you give the police certain powers, they're guaranteed to be abused at some point. Individual officers can be held accountable but what if the robot malfunctions? Killer bots are more than a slippery slope, they're a giant pandora's crate that should never be opened.

While you can make an argument about using deadly force against those types of actors (although, personally, I don't think the answer is clear cut), most concerns I've seen were around using deadly force in situations where it wasn't warranted, against people who did not truly pose an immediate risk to others.

I would think policy makers inside SFPD would be working 24/7 with the politicians in order to bring down the ridiculous amount of crime SF has ( most of it isn't even reported because police now say it outright they wouldn't even bother to respond ).

It feels like when I invent things to do when I really don't want to tackle that hard problem..

They acknowledge "remotely piloted" in the first sentence. But then, back to drama in the second with "granting robots". Granting implies autonomy and choice where there is neither; granting the officer pilot would be more truthful.

The term bot or robot should be reserved for independently thinking entities. The machine here is just a remote manipulator.

San Francisco police using a robot to kill or the use of deadly force, would not be any kind of new ground or blazing new trails for American police.