No one can win this. The dollar cost per KIA is going to quickly drop below $20. And this will be for all optical hunter killer drones.
the only way to win is not to play
No one can win this. The dollar cost per KIA is going to quickly drop below $20. And this will be for all optical hunter killer drones.
the only way to win is not to play
How are you going to fix this with a law? The whole problem arises because drones are easily hackable and the technology stack involved is cheap and open. Countries make them, what are you gonna do, sanction them? You could build a drone swarm in your garage and it will be harmless right up to the moment you choose to weaponize it.
Chemical weapons are a terrible parallel because while you don't necessarily need big industrial facilities, it's not that easy to conceal. But there is so much you can do with autonomous systems that's completely innocuous, in much the same way that you can't do a whole lot about cyberwarfare just by declaring it illegal.
I mean look the guy who assassinated Shinzo Abe; Japan is about the most law-abiding country going, guns are rare, and as an island it's super-easy to control imports and exports, so the guy just built one out of parts from the hardware store/garden center. Weapons are just not that hard to make, and if you're inclined toward lateral thinking the possibilities for drones are near endless.
To be frank I think a lot of military drone people are missing a trick because they have unrealistically large budgets to play with so demand these impressive but very expensive models that are designed to be part of an infantry loadout and deliver guaranteed kills from around the corner or through the hole of a pillbox.
You can't disarm America, Russia or China's nuclear weapons like that. The example of nuclear weapons only shows that cat like this won't go back in bags. The only way you'll get countries like this to disarm and renounce a class of weapons is if those weapons were never very good in the first place (e.g. chemical weapons) or were rendered obsolete by new sorts of weapons (battleships.)
Drones are increasingly used by civilians, and we see in Ukraine that consumer drones are being employed in military operations with little more than a software upgrade.
Unfortunately good intentions don't make bad ideas any better. Otherwise, let me propose that we solve world hunger by mailing everybody a free hotdog. A terribly dumb idea, but motivated by the best of intentions trying to solve the worst sort of problem.
An incredibly dismissive straw-man.
Firstly, the inevitability of the disaster is not at all obvious. Drone attacks don't scale in anything like the same way nukes do. Nor do kinetic attacks have the same long-term environmental hangover that nukes and chemical weapons do. Drone swarms are a potentially deadly problem for military elements and potentially political organs, but they excel at narrowly targeted objectives. They're not on the same level of existential risk to entire cities, unless we get into skynet territory with wholly autonomous factories and logistic chains.
What I'm saying is naive is the belief in technical (legal) methods. You write: Bans and controls are effective at limiting quantity but this is clearly not the case. Small arms trafficking is actually widespread despite all the hand-wringing, and efforts at gun control in the US are token at best.
They're ineffective partly because they are asinine and pursued in a childish fashion. Gun control as a policy objective is wholly reasonable, and I can give numerous arguments on its behalf. But in the US, the legal environment is ruled by the 2nd Amendment, whose language about how 'the right to keep and bear arms shall not be infringed' is so plain as to be inarguable. Gun control proponents really won't make any serious headway for the foreseeable future unless they try to repeal or rewrite this. None of them want to do that because it's a political non-starter. (Incidentally, this is not an argument for just giving up on the problem; a small tax on the sale of ammunition, with all revenue reserved for the medical or funeral expenses of shooting victims, is both politically and legally feasible imho).
My argument is not only about quantity, but versatility. I mentioned how easy it is to manufacture weapons. The sad reality is that if you could wave a magic wand and make all guns in the USA vanish immediately (or just privately held guns if you enjoy being pushed around by people in uniform), in a month's time you would have a gun problem again. It would be on a smaller scale - at first. But it wouldn't stay that way. It's costly to manufacture a really well-engineered gun, of course, but it's very very easy to manufacture a crappy gun that's Good Enough. You don't need a fancy workshop or a degree in mechanical engineering, as the Abe assassination showed. If you have some mechanical aptitude or admit the use of cheap and easily available tools, the emerging insurgency in Myanmar is making extensive use of 3d printed guns.
A less spectacular but increasingly widespread use of 3d printing is the manufacture of auto sears, a mechanical part that swapped into the trigger mechanism of a semi-automatic gun to allow fully automatic fire. It's really simple, just a lever with a particular shape about the size of your index finger. They're trivially easy to make and very easy to apply. The ATF charges people for selling them quite frequently, but let's be realistic: they are only finding the suckers who are so stupid/greedy that they advertise more or less openly. This is not such a huge problem in my view because most people who want to use one of these are into the novelty factor of shooting a high powered weapon for kicks. But you'd be foolish to assume that they're not proliferating wildly in certain circles just because of a few prosecutions.
Just as very basic hardware can be repurposed toward weapon manufacture, any computer with a network connection is theoretically a cyberattack vector. It doesn't have to be a super-clever or advanced cyberattack; you can trade off quality against quantity if you are just basically competent in python or shell scripting, and you can even get results with crude readymade tools like Low Orbit Ion Cannon.
So it is with drones. Weaponizing them is trivial; payloads don't need to be huge and the same technology that can save a life by dropping medical supplies can end one by dropping a hand grenade (both of these techniques are being used in Ukraine on a daily basis). Computer vision, maneuver, obstacle avoidance, communications etc. etc. can all enhance the capability of a drone swarm, but all those things have valid purposes too. Like a piece of steel tubing or a Raspberry Pi with Linux installed, it's not the device itself that's the problem but the purpose for which it is configured.
So I think it's childish to think you can legislate away drone swarming, the same way it's childish to think you can just legislate small arms or cyberattacks or drugs or protests out of existence. In the end, legislative solutions are just about making a really nasty example out of the few people you catch doing a Bad Thing and then hoping it scares everyone else into compliance. Once you lose legitimacy or people decide they don't care that much any more, it's useless. I'm emphasizing 'people' here because drone technology is so simple and accessible that it's not limited to nation states or large industrial concerns. Any garage could be a drone lab, any pickup truck can be a launchpad.
This is just completely untrue, as demonstrated by various countries that have such bans in place. The ease of acquiring firearms is the primary driver of many people having firearms. Maybe one person in a hundred here will go to the trouble of acquiring a dumpy 3d-printed gun. Japan does not have a gun problem - one person going to huge effort to create their own gun is not a problem.
In general your thesis seems to be that ease/legality of acquisition & manufacture of weapons will have no impact on the quantity of weapons available, because technical workarounds exist. This is a fairly common point of view among technologists, or at least its analogue is often seen when a new service is announced - "you can already easily do this with [technically-complicated solution], so this new service is not needed". This point of view has a long track record of failure. Of course making something easier lets more people do it! Most people are really not that motivated to do things! Conversely, making something more difficult ensures fewer people will do it.
When we invent a version that will be practical to use, (e.g. small hard to notice drone capable to spray highly poisonous chemical right in enemy barracks) all the "success" in outlawing we had will instantly disappear.
By contrast, precision-guided munitions make the Bomb obsolete (it's great at killing people but it's a comparatively crude means of coercion by comparison), and they can be produced inexpensively from widely available materials.
We should be planning how to preserve what we can from defeat.
There's also disputed allegations of chemical weapon use in Syria. I'm not sure why they've not been seen in Ukraine yet; I think we're back to "not actually that effective in open warfare".
As for why they aren't seeing use in Ukraine? They aren't very useful on the battlefield in most circumstances. Dan Kaszeta (a well regarded expert on the matter) has had a fair bit to say about this.
That is assuming there remains any defense against these drones by new or conventional warfare. This might be a legitimate threat to major military forces themselves at some point.
Yeah, the autonomous weapons cat left the bag decades ago. Acoustic homing torpedoes were first used during WW2, and became very sophisticated during the Cold War. Missiles that can be fired blind and left to find their own targets were also developed during the Cold War, like the (since discontinued) anti-ship variant of the Tomahawk. "Autonomous weapon" describes a lot of systems already in use. Banning them all is DOA in America, let alone in Russia/China/etc. They're too useful to ban.
America won't even sign the Ottawa Treaty banning anti-personnel landmines (because the American military considers them important in Korea.)
The question is whether the American government would accept banning itself from having autonomous weapons. I think they wouldn't; they're too useful and already normalized.
Once the world sees a few wars where the drones win, every country on earth is going to get them. They are cheap, can be made by university students, can be weaponized in a myriad of ways and can be deployed from far away (UAVs with a thousand drones, can deploy them thousands of kilometers) and they are disposable.
Silent drones with autonomous instructions, will be the ultimate booby trap.
Not always. The only counter weapon to a Nuke, is a similar sized or bigger Nuke, better delivery systems for the nuke and a large number of nukes.
Once launched, nuclear weapons are difficult, but not impossible (see ABM defense systems) to counter, but there are many ways to counter their unlaunched effects.
And the Israeli iron dome is in an entirely different class, its hundreds of times too slow. Like trying to use a sling to shoot down a cruise missile.
Actually this is a very good example of why counter-drone systems will likely never deter expansion of drone capabilities. The costs to provide full protection are simply too lopsided.
Why spend a billion dollars to protect against 100 million dollars worth of drones when you could simply buy a billion dollars worth of drones in the first place?
Not just wars. Every conflict should make a pact about RoE to not use autonomous drones. The asymmetry is too high. There is no defense. The cost is asymptotic to zero.
We should outlaw war.
We're lucky people haven't discovered yet how dangerous drone swarms can be. And as commercial delivery drones are allowed to fly the skies, many nameless drones can blend in among them.
Really a very dangerous future. Then again we also have nuclear weapons and the collapse of ecosystems...
All to say, for better or worse, this all could be just as scary RIGHT NOW, but for whatever reason it isn't. No reason to believe "uh oh, here comes potentially scary tech, better try to blanket ban it now."
Sure, but how do you get Putin, Xi, Kim Jong-un and all of their other soul mates into the boat?
So we'll eventually outlaw them once we have a strong edge.