Google's ex-boss tells the US it's time to take the gloves off on autonomous weapons US government should avoid hastily banning AI-powered autonomous weapons and instead step up its efforts in developing such systems to keep up with foreign enemies, according to the National Security Commission on AI. The independent group headed by ex-Google CEO Eric Schmidt and funded by the Department of Defense has …
Banning it should stop everyone doing it. While we are at it we should really look into an international agreement to ban recreational drugs too, we have enough ways to kill ourselves with bad health without taking chemicals for fun!
The box is already open, too late for banning.
With every human advance in technology from the very fist hominid throwing a stone or waving a stick, it has been a fact that any and all new destructive technologies will be taken up, to a lesser or greater degree by any that have the wherewithal.
I am, frankly still amazed that we have so far avoided the MAD scenario, particularly recently with the development of more 'efficient' theatre tactical nukes that many generals consider to 'be far more usable'.
Smart warfare may well turn out to be one of the dumbest things humanity has come up with.
The level of perceptive attention in this report is summed up by its title "Final Report National Security Commission on Artificial Intelligence" and file name "Full-Report-Digital-1.pdf", which could refer to almost anything. The problem is confirmed by 750 pages (more than a quarter of a million words) to say "if our potential adversaries use autonomous weaponry, we should consider doing so too". Duh!
There is currently no plausible scenario where large groups of autonomous robots are marching against each other on a field or so.
Would investing in anti-autonomous waepon technology not be the wiser move?
Develop sensor-decieving technology making humans desappear from sensors, make locating autonomous waepons easier, deploy personal EMPs and other weapons that selectively kill automated gear.
Not quite as cool as a fighting robot, but probably way more effective in battle ...
It's easier to use low tech high explosives to make an enemy soldier disappear than is is to use hypothetical sensor-deceiving weapons to make one of your own soldiers disappear. James Bonds invisible car can't really hide the car from multiple eyepoints.
Ethical, argument goes like this, modified to fit current lingo: they're EVIL, MURDEROUS BASTARDS so they need to be 'eliminated', while we're protecting PEACE, DEMOCRACY, PROSPERITY, OUR WAY OF LIFE, so it's ethical to eliminate them.
By following the Geneva conventions.
If your question is rhetorical based on the concept of killing people being unethical, then good luck with that. Having a potentially bigger stick than the person who wants to bash you is as far as mankind has come, peace strategy wise.
You just need to use every means available in the current media to dehumanise the humans you want to kill, as soon as the consensus is sufficiently large, they are no longer people so you can kill them without troubling yours or the collective conscience of your population.
What happens when the current media -- and populace as a whole -- are split nigh-evenly into thinking the "other half" are inhuman?
New personal working theory: US politics only works because of MAD.
---> Mine's the windbreaker; it's quite breezy up here on the tightrope over the narrowing middle ground. Don't look down.
I think I've heard this argument before. We must purchase that 'machine' gun, and then, it will deter other countries from wreaking havoc! And bonus: WE will be able to wreak havoc upon those silly arabs in those silly -stans, cause won't be able to fire back! Oh now, they already have those drones, many drones! Too late, too late - we need to get more drones, faster drones, smarter drones!!! Now, pay me, gaspadin Putin, I've done my bit for today!
p.s. but honestly, all this talk about banning was pointless (well-meant, but still pointless) when the issue was first floated, several years ago, and it's even more embarrassing, frankly, in 2021, given what happens already on the battlefields (and in more - or less - makeshift labs away from the battlefield).
I'm sorry, I learnt nothing that's not common knowledge, bordering on bloody obvious. Plus, I found it irritating to watch a string of recurring images, loosely on the subject of circling globe / drones / microchip production / pentagon / target acquisition, etc, etc, again and again and again. Other than that, possibly informative as a podcast, for kids. For their future, this business is a-booming :(
...... and Changes You Never Dreamed Possible and therefore Probable.
American should not turn its back on autonomous AI weapons. But the development should be carefully monitored to make sure it abides by ethical policies.
Now it is probably most wise to realise, that would be far too little, far too late .... for the absolute truth is the most one can do is to try and carefully monitor/spy on developments, for autonomous AI weapons, as the moniker attests, have a mind of their own and one certainly can never ever be at all sure they will abide by any man made rules or ethical policies, especially whenever AI/IT/they discover and uncover the fact that what is to be protected by them is long ago corrupted and perverted to server an Abomination of a Global Operating System/Earthly Executive Office Administration which cares not a jot about steadily improving the lives of the many on Earth in favour of just excessively rewarding a few with the fruits of labours of everybody else, just because they can and they be thinking they are never ever gonna be found out, and targeted for special remedial attention.
Was that clearly mentioned anywhere in the over 750 pages of the report? And if not, why not, for it not something you would be wanting to ignore, given what such weapons can so easily do? Don't you know any of that?
Well, now you do. Take care out there with whatever you dare share to win win and never fail with, with an almighty failsafe Global Operating Device.
It is timely that I watched that classic Star Trek episode, A Taste of Armageddon: https://en.wikipedia.org/wiki/A_Taste_of_Armageddon
Once we detach ourselves from the horror of actually performing the killing we sanitise it and make it feel a little more acceptable.
There is no good outcome from that road.
I am reminded of photos and newsreels from the War after the War to End All Wars with the top brass pushing markers around a map to govern troop, ship and airplane movements.
With autonomous weapon systems, the top brass will no longer have to worry that as they move a marker their troops are dying.
Does anyone think that this will reduce the risk of war?
How long before we see another paper arguing for first use of nuclear weapons because of their EMP effects on autonomous weapons?
Saint Petersburg Declaration of 1868
In 1863, the Russian Army had perfected a fulminating musketball that could explode when it hit a hard target and was designed to blow up powder magazines or ammunition wagons. In 1867, they perfected an improved explosive musketball that would detonate on any impact after being fired, even soft targets like people or animals. Predicting the disastrous effect of such a discovery on diplomatic relations with their neighbors, Russia decided to negotiate a ban on the development, creation, and use of such weapons before a grisly arms race commenced.
Only if everyone agrees can it be banned. And there IS the rub. These days, there are too few in power that have seen the horrors of war, experienced the horrors of war, lived with the scars or understand the repercussions of their actions; DAISH/ISIS anyone?
And an item you left out of your argument; they also found that if one dropped said round it could go off injuring or killing the soldier attempting to use the thing. So it would end up probably more lethal to the person using it than the enemy since not every bullet hits its target yet an ammo belt that explodes because the soldier lands on it while diving for cover would happen much more often.
Can this lead to weapons taking over and creating an empire of their own, turning the commanders into their slaves? Machine learning, autonomous weapons? What else will DOD come up with to get more money? ...... Norman123
Yes, Norman123. It can, and therefore does.
Has never been employed by nor worked on a DoD program involving autonomous systems operations. This IS my day job. This person has the morals of a a true Satanist.
So tell me Mr. Village Idiot, is that bus you've detected coming out of disputed territory, been tracking and targeting, a bus full of terrorists or a bus full of nuns? HUMANS make that determination. HUMANS are held accountable for their decisions. Are we going to counts martial a robot? Incarcerate one? Hang one?
Back in your hole you incipiently stupid, knuckle dragging moron.
HUMANS are held accountable for their decisions. .... HammerOn1024
Oh please, you cannot be serious. Where have you been ? The supporters and planners and killers of the likes of a Harry Dunn or a Jamal Ahmad Khashoggi would be able to prove that an absolute nonsense .....
Some humans are held accountable for their actions, some others would think claiming and being granted immunity and safe secure harbour gifts them acting with impunity. You can almost hear them now shouting .... The corrupt system allowed me to do it, your honour. Don't blame me.
And then the system wonders why others exact vengeance and sweet justice via other unauthorised means. What a dumb system that is. Long past its sell by date, methinks, given the self-destructive problems that it creates rather than solves for incipiently stupid, knuckle dragging morons.
"If the time has come for the lion to lie down with the lamb, I prefer to be the lion."-slight paraphrase
Morals are for the living. If we do not give our opponents good reason to worry either that their weapons will be ineffective, or that the response will be more than they want, they do as they please.
And what they please displeases me.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
