A suicide drone can be used to take out any individual anywhere on the planet, fully remotely and untraceably with minimal collateral damage, while radar, missles and explosives can't. The scenario depicted in the film below is getting closer to reality:
Modern guided missiles are not essentially different that these drones. I guess the main difference is that these drones are intended to be cheap and disposable, while having a much shorter range.
I love this analogy. While Turkey can't spend hundreds of millions of dollars to buy smart missiles, it innovated to protect herself.
Its allies used to rent drones for surveillance against terrorism, then used it as both a carrot and a stick. Now, Turkey has its own drones that regularly take down targets effectively, even some air defenses.
After Syrian war started and "modern western liberal democracies" failed to help with the refugee crises, those borders became even more treacherous.
As for refugees, they weren't forced. The border police just didn't stop them. I agree Erdogan coming on TV and yelling it out openly was a below-the-belt tactic and after that people started transporting refugees by the bus to the border.
Being on the side of Turkey in the Syria conflict cannot be the right side of history. Especially because they are now in control of Kurdish occupied lands in Syria. Saying that the Turkish government harbors a century-long paranoid animosity toward the Kurds is the understatement of the year.
Turkey's government is just the big despotic fish in the small geopolitical pool that is the Caucasus / Northwest Middle East. Finding cheap ways to kill is not a defensive strategy for the gov't.
Drones are consumer level technology. This isn't very different from a high-end Mavic drone, except that someone mounted an explosive instead of a camera to it. On the other hand there is absolutely nothing that a regular consumer can buy that would reassemble a guided missile.
I agree. Technology trickles down the bank account spectrum. At first only nation states could do drones. Then smaller nation states could do drones. Then bigcos could do drones. Now we've got drones in consumers hands. It should come as no surprise that some nation states decide to use consumer quality and price drones for nation state purposes (their police probably patrol the streets with the same handguns consumers can buy). Just because it's trickled down doesn't mean it's not useful.
oh you should definitely checkout some of the more serious "hobby" rockets and launches. All it would need is some extra kick and some guidance. Hezbollah builds and launches even more basic missiles all the time, and hurts Israel all the time.
You can buy a jet engine and build a 700mph drone that can carry a payload. You can also buy a laser gyroscope, or use open source code to build a SLAM system for guidance.
I wonder what chance anyone has of stopping something like this. It can certainly carry an explosive and flying 50 or even 200 km is probably doable with a bit of a scaleup.
It is great when the use of a weapon is well thought out. The risk of collateral damage forces people to think twice before deploying these weapons though.
It's an interesting thought experiment as to whether there is a point of inflection where the risk of collateral damage is so little that it drives the opportunity for extra judicial killing too high.
> A suicide drone can be used to take out any individual anywhere on the planet, fully remotely and untraceably with minimal collateral damage, while radar, missles and explosives can't.
This just sounds like you're trying to persuade people on why drones are definitely the future of war.
They are going to be, doesn't matter what I say. The cost/kill is significantly lower with a small fully autonomous suicide drone carrying a small explosive. All it takes is a small drone (30$), a few grams of explosives and software to kill someone (anyone). I'm sure the software will get there.
Drones don’t have unlimited range. The ones in the article have a range of 6km. If there was really a war situation I imagine an EMP could disable them or a MOAB could just blow everyone up instead.
The nuclear bomb brought nuclear reactors, providing affordable electricity in many places. Drone swarms also have the potential to turn into useful applications.
Are you suggesting that nuclear reactors could have never been developed independently, and that the nuclear bomb was a necessary precursor? It sounds almost as if you're giving these deadly, weaponized drone swarms a pass because it's possible to do non-deadly things with drone swarms as well.
Not op but: I wouldn’t say they couldn’t have been invented, but it’s unlikely they would have for economic reasons. Controlled Fission almost had to be done by government due to the resources required and uncertainty of the payout. It seems unlikely any private company would have done what the Manhattan project did just to get a small cost savings over burning fossil fuels
> Are you suggesting that nuclear reactors could have never been developed independently, and that the nuclear bomb was a necessary precursor
An uncontrolled nuclear reaction is significantly simpler than a controlled one. We still don't have fusion reactors, while fusion bombs were developed 10 years after fission bombs.
That's a very unconventional use of the words "uncontrolled" and "simple". The reaction inside a nuclear bomb (fission or fusion) has to be incredibly tightly choreographed down to the microsecond in order to work as intended, whereas a fission reactor is (egregiously oversimplifying) put some fissile stuff next to some other fissile stuff for a while and let it get warm.
The Chicago Pile came way before Trinity. Rock formations have even been known to come together accidentally to form natural nuclear reactors, but I think it will be a very long while before the first natural nuclear bomb.
To a degree they would not have been invented with out the weapon. For example thorium reactors are designed in the 60s. The US ignored them because we couldn’t make weapons from their waste. The only reason thorium even got started was we thought it could be a weapon.
Your calls are in vain. That would only make sense in more liberal countries like western ones. When it comes to tech and innovation Turkey has become quite forward-looking but they still retain their nationalist culture.
In case of war I expect no difference between the "liberal" countries and the "nationalistic" countries. Remember that Hiroshima and Nagasaki were civilian cities.
> During World War II, the Second General Army and Chūgoku Regional Army was headquartered in Hiroshima, and the Army Marine Headquarters was located at Ujina port. The city also had large depots of military supplies, and was a key center for shipping.
As long as there's at least 1 actor (state or not) doing this, it's very useful to develop such technologies. Powerful military is damn nice deterrence.
I'm very skeptical. We can't even get police to properly investigate many crimes in the US, let alone without "making a mess". The idea that we're going to magically invent AI that "detects crime" and carries out lethal summary judgment seems completely detached from reality. By the time we have general AI that can do that, we won't even be in charge anymore and whatever happens from there will be up to the AIs.
I don't think you need a general AI for something like this.
You can detect what the other person is carrying and that works good as a filter. A person carrying knife in a crowd or school outside of canteen doesn't make any sense by general norms.
Second would be to detect the intent - now that's a much harder problem but you can whitelist few situations where the person has already stabbed someone or shown sufficient physical evidence of criminal intent.
Not sure we need general AI, we can already detect undesirables and shadowban them on Twitter, all you need is to throw some facial recognition into the mix and you can link a person to an account that has a "wrong" ideology.
"Linkedin implements a system to identify to-be criminals by personal history and behaviour it tracks", in a report on Friday from the executives, Linkedin has decided to label offenders on its platform to reduce their reach and stop them from getting a job before it's too late. The CEO explained that this is just one of the policies that linkedin has implemented to serve justice to our society.
Mark Zuckerberg says "justice shall be served. Social media platforms have the duty to govern society". Facebook has decided replace profiles of people breaking its policy with support social justice banners and links to donation box of variation NGOs. Additionally, they have decided that this policy will effectively block all offenders accounts on any platform they control including instagram, whatsapp, and integration which uses Facebook sign in functionality. They will share this data publically and encourage others such as partnered advertisers to block those offenders.
After years of toxicity on Twitter, jack Dorsey says they are finally taking steps to serve justice to our society. New twitter update will make it mandatory to attach your credit card at all times and social security number as well as sensitive personal information. Offenders will be charged and fined on the platform and their accounts will be handed out to the victims for exposure therapy.
Sundar Pichai thinks google must fix people's wrong perception and biases. Google representative told us that they would soon manipulate search and ad results of people it deems problematic in an attempt to "fix" them. Racist people will start seeing more anti-racist content in the coming days.
Financed by states, sure. But they can and most definitely do operate without states. They can just turn to crime, drug smuggling, or if popular enough can finance themselves.
Human losses could be prevented if all parties in a conflict have access to this tech. Then there's also the matter of trust. You stopping work on it doesn't mean your adversary has stopped work on it.
Yeah go work for Google instead so you can suppress the freedom of speech for millions of unnamed data points on the internet... at least this southern wolf bears it's teeth before it bites.
unfortunately money and cutting-edge tech are in this kind of applications
the amount of money put into may-be-probable piece of tech is astronomical in defence indistury, because it may really be game-changer. (eg. lasers, rail-guns etc)
every piece of tech we use today started as a defence project. You may say ML/AI is just research stuff. If you think AI is being developed to help you with groceries, you are being lame... (no harm intended but my opinion)
Pretty sure all major powers either have or will soon have the capability but don't feel the need to blab around about it.
The issue I have is when this gets actually exported to countries that may want to use against civilians (to squash opposition) though it was know for a long time this must happen.
> The issue I have is when this gets actually exported to countries that may want to use against civilians (to squash opposition) though it was know for a long time this must happen.
Turkey wouldn't need to export it to anyone for it to be used against civilians, that's a pretty common thing for Turkey.
The issue I have is non-state bad actors acquiring or building them on their own to create flying IEDs. It would be pretty easy to fly a drone into a crowded stadium or protest and kill hundreds if not thousands of people with a weaponized drone. Getting the explosives is the hard part and many groups have already shown they can do that.
Because that is an escalation that nobody wants, for now. Even piracy was tolerated for a while until it became economically too disturbing. I don't think it all disappeared but it is not that bothersome.
If anybody started threatening commercial civilian flight, countries would fall on them like bricks. Quarantine for a few months was devastating on many fronts. Now, think if all flights were grounded for months...
The only thing I can think of is that the places where it's relatively easier to get explosives (war zones like Iraq, Afganistan, or Syria) doesn't overlap much with places where it's easy to get powerful drones (developed countries).
Like the nuclear agreement, where only a small subset of countries rushed into having nuclear weapons and restricted others from having any, and now it is extremely unfair to other countries to not have the power but others do and enjoy having the weapon superiority.
No international agreement will survive war, they are just used to suppress and bully countries.
International agreements are not perfect but they are still necessary. They are more or less holding in the chemical-weapons area, which has really nothing to do with restricted technology.
Humans can (and must) agree to be better than frightened animals.
>They are more or less holding in the chemical-weapons area,
Tactical inconveniences that limit utility are what's preventing chemical weapon usage. Nobody uses chemical weapons because we have better, more effective ways to kill each other making them not really useful except in limited circumstances like terrorizing civilians (a static target that probably doesn't have gas masks) in areas where you have air superiority (e.g. Syria now, Iraq decades ago, etc.). Chemical weapons were mostly abandoned after WW1 because they just didn't work well enough for the effort and resources required to use them.
What's the reason to regulate them? As weapons go, they're extremely clean. A step up from almost anything else, in terms of side-effects and collateral damage. Scary - as hell. But they're meant to kill people so yeah.
These weapons are so cheap that the cost of starting a country just fell by an order of magnitude. I could see a hacker warlord of the future giving everyone who is a citizen some sort of transponder to prevent drone attacks and have the drones kill everyone else in a given area.
It's so obvious this is the next stage of conflict. China could have thousands of these sleeper drones in the US. To preempt before a real conflict, it could simultaneously take out most of the heads of state, causing chaos.
It's worse than that. If you undermine chip fabrication to insert backdoors at the silicon level, you could disable...almost everything with an arbitrary signal. The only problem with this is that you, the attacker, need to be careful not to install the altered chips in your own infrastructure.
The likely future is the other way round: frog boiling and border pushing, like at the border with India. China doesn't want to have a real conflict, it wants to keep poking and pushing until it gets what it want.
"China could have thousands of these sleeper drones in the US." --
Yeah, and 911 was an inside job!
EDIT: If you think China has thousands of sleeper drones in the US you are on par with people who think 911 was an inside job, end of story. Down vote all you want but be sure to don your tin foil hats afterwards.
>China could have thousands of these sleeper drones
Why is China your example and not X? There are other countries that use killer drones, do secret killing and kidnapping operations on other countries territories etc.
I have the impression that a few years back you would have said Russia, then N Coreea and today China, like the big bad enemy is changing each few years depending on what politics need.
How did it shift from Russia to China, the same guys are still in power, the only guy that changed is Trump(I think he
has mentioned China a lot but no idea if he caused this shift of attention or just using it).
It's foolish for a state actor leave the remnant pieces of an advanced weapon (like an autonomous drone) at the scene of a crime during an assassination. First because the uniqueness of the technology makes the actor identifiable, second because the country targeted will react more cohesively to this novel kind of cassus belli (e.g. news coverage of 1 drunk driving death vs 1 autonomous vehicle death)
The good old fashion alternative is just a gunman. It's more plausibly deniable - e.g. still nobody is exactly sure what was going on with Lee Harvey Oswold and Jack Ruby - and it doesn't put the targeted country immediately on a war footing.
500 isn’t a lot but I imagine having 10,000 flying suicide drones could be very effective in battle. Probably an EMP would bring them down. I wonder if something as simple as smoke bomb would render them useless if they couldn’t see properly.
Swarming drones and AI is the scariest tech and we are going to see the consequences. Because it will be so cheap and ubiquitous. This isn’t even the atomic bomb — where the work was just too sweet to stop. This is basic technology with small incremental improvements tons of people can do.
Drones are the first robots that are among us (overhead). Self-driving cars are at least very expensive to produce. But drones can be mass-produced and deployed anonymously.
Drones with live munitions is terrifying. Imagine an anonymous swarm descending on Union Square in SF.
Even under normal operation, drones can fall and hit someone by accident. Let alone a swarm.
This is one technology I should not want to be in anyone’s hands, but if it is, it should be in the hands of the State and not everyone. Just like bombs.
PS: With cheap cameras being ubiquitous and people can be identified by their gait, heartbeat etc. consumer level tech can also include an AI that can cross correlate everywhere you’ve been. Businesses would use it to fine you when you park your car for too long and don’t go into the store.
But governments and police departments can concoct tons of believable stories with parallel construction and put anyone they want away. In fact AI can also be adopted in other places in the policing and judicial process (as with Palantir) leading to a Kafkaesque experience for sentencing.
And with deepfakes it’s worse than that. Anyone could create a believable video of a rape-caught-on-video or a damaging “speech at a private event” or whatever. Reputational attacks were cheap already (see CIA handbook) but with a swarm of sleeper twitter bot accounts and other bots you could introduce anything.
In short I am worried we are going to be entering a dystopia soon. This video nails the drones issue:
USA has been operating drones since last decade in conflict zones in the Middle. The same drones that they use to "neurtalize" terrorists and dictators. Is there still time to oppose that or have we casually accepted them as a necessary for the war on terrorists?
The collective trauma experienced on 9/11 largely made anti war movements DOA. Combined with the sheer volume of conflicts and proxy wars were involved in, were probably too far gone.
not only did it stop anti-war movements, it also stopped anti-capitalist and anti-globalist movements. Seattle in 1999 to 2001 had huge protests and riots [1,2]. After 9/11, protesting the west almost meant supporting the terrorists. When globalisation used to mean the exploitation of the developing world by the west - now the west was under attack and so to attack imperialism via violent protest was to be sympathetic to the terrorist.
It's only now, 20 years later that the children of these protestors are repeating the same things their parents did. And it also explains why certain politicians try to associate protest with terrorism these days, that it's a different generation protesting explains why this association falls flat.
Today's protestors were not born when the towers fell.
Turkey is slowly becoming a regional power. It is interesting how well they seem to be playing the political game with EU, US, Russia and Israel. I am not a fan, but I can't help but to admire that level of skill to balance competing interests.
Turkey has a unique geographic location of immense strategic value. There is a reason NATO bends over backwards to include a country that is culturally completely incompatible with the rest of the mostly European members and even openly attacking other NATO members like Greece. What you see now is heavy influence of China that wants to take complete control over Europe under the guise of something called the "Belt and Road" initiative with the heavy support of some European industrial leaders.
Assume that the leadership in Turkey is nothing more than a puppet regime fullfilling those interests.
to me it looks like they are overstretching their hand and will eventually snap. if AKP stays in office eventually turkey will be the next Iraq without oil
Can't drones be jammed? If so what does it take to jam their signals, fry their circuits with some sort of electrical guns? Having these war flying tools advanced so much in the last decade, isn't there an incentive to develop some good defense against them? Or maybe it is happening but it's not made public.
They don't have full autonomy yet, there appears to be a human operator who takes key decision remotely.
But, even if that was so, is there no way to have some defense against such attacks, having that they're electric and rather slow? Tesla ray guns come to mind but I am not very knowledgeable in this field.
While its true that they have human operators currently, they don't necessarily need a lot of autonomy to attack someone jamming their signal. Most of the drone signal jammers I've seen take the form of a handheld directional radio "gun", so it would be pretty easy to just have your bombdrone just direction find the jamming signal and bomb the operator, freeing the field for the n other drones.
Dean Ing wrote a SF novel 'Butcher Bird' exploring this. Why is humanity stuck in the same story? We make something cool. We then commoditize it. We then militarize it. I think every tool/discovery/thing we invent/produce should also be introduced with its own antidote/kill switch. Clearly a malevolent state would be able to excise the the suicide switch in the chip, but at least this is a start. Deploying an EMP device at scale is a lot harder than activating a paralysis/suicide switch.
After all, not everyone can deploy their own custom cpu. And the REAL worry is when a script kiddy gets one of these and deploys them at the next protest. I dont want to say the 'T' word. By making commodity hardware fragile and defeatable, we protect ourselves against our dark side. And no matter how good an AI gets at face recognition/inference/etc, we will still be able to turn it off. Did we not learn the lesson of 'Colossus: The Forbin Project'
Well, that was a chilling and terrifying look into the future.
But, to be honest, I’m not really worried. It doesn’t really change the tactical situation of today. It just makes the attack vector cheaper.
The USA already uses drones that loiter from 30 miles away, and has a high resolution camera that can see you. Then you’re targeted and a Hellfire missile is locked on, and shot at the target.
Same effect. The target is killed.
However, this technology increases the critical retaliation magnitude.
If the attacker chooses to assassinate the political leadership of the enemy country by suicide drone, then the victim country can retaliate with an all out nuclear assault.
So, we are now basically at the same situation we were at before. MAD principles still apply. Nothing has changed.
To play devil's advocate against the panic I'm watching unfold in this thread: I don't think there's an enormous categorical difference here from weapons that already exist.
We've had long-range (longer range than most drones) guided missiles for decades. We've had drones that can launch them, for decades. Those technologies already have the ability to wipe out targets anywhere, anytime, en masse, with little to no risk. Yet we don't see people (as one commenter suggested) wiping out entire plots of land so they can start a new country. We don't see (most) countries turning such weapons on their own citizens. We don't see (very many) heads of state being assassinated in this way. What this tells me is that there are sociopolitical forces that make such things un-advantageous or infeasible, for one reason or another.
The only true differentiator I can see with suicide-drones is a potential increase in precision, though that's not really true right now because these in the article look to be basically flying hand-grenades. That's not really better than a missile; in fact it's probably a lot easier to intercept or shoot down. Now, many of us have seen that (fictional) video with the micro-explosive assassination drones. Maybe that would change the game in some way, but I still don't think it's a drone apocalypse we'd be looking at. Nation states have so many ways of killing people at this point. When they don't, it's usually because it doesn't make sense to, not because they lack some key capability.
You are relying on the presumption that they won't become autonomous machines for wiping out "criminals".
A drone can be used for more than a singular purpose. There might be more efficient ways to kill someone we already know like you said but what if we don't know them yet?
What if they don't exist yet?
Imagine you have a crowd and there is one violent person inside it, how would you immobilize such a person without harming the others?
They could already wipe out people they want to designate as "criminals" by busting into people's houses and shooting them in their sleep. The reason they don't (very often) is that, as we've seen in recent weeks, that's how you get riots. The current unrest resulted from only a handful of unjustified killings. Imagine what would happen if they started sending drones to kill people en masse?
That's a very US centric view. It doesn't apply to many other countries. The point is still that it will be cheaper to manufacture large number of autonomous justice drones than to hire policemen to do the dirty work.
Why pay for their pension years after they have retired?
Why pay for their healthcare, training, damage, etc?
You really think lack of money for hiring thugs with guns to shoot helpless citizens is the limiting factor for any ill-willed nation-state?
Even the most insidious state relies on its citizenry to some extent. You're not ruling anything if your people are constantly in revolt, or if they're all dead.
And you really don't think those inhabitants would riot and swiftly overwhelm whatever enforcement existed if their friends and neighbors started getting picked off by assassination drones? If anything it sounds like India's enforcement should be more scared of the populace than they are in the U.S. because of how outnumbered they are.
Playing the devil's advocate since you think cops are thugs with guns who can't work without making a mess, why don't you believe in replacing them with autonomous drones?
In an ideal future, autonomous drones would make less mistakes since they don't feel threatened or care about dying. They won't kill someone just because it looked threatening. They won't turn off the body cam because of reasons. They will have a more accurate aim and will be more trained since training doesn't cost anything for them. You can invest the other funds in improving rehabilitation or de-escalation services.
Humans aren't good in stressful situations and their aim isn't great. A ground view isn't necessarily helpful in situation where the criminal might be inside the group ring.
You can only do that for specific situations. What if the criminal runs inside the building or grabs a person to shield himself from the sniper? Are you going to surround the person with snipers from every direction?
Also sounds like the cost for those snipers would be pretty high than manufacturing few drones to lock in a target and swarm under their head.
Cost of drones is only going to decrease while the cost of sustaining human life remains more or less the same if not increase.
A bomb drone would have to be absolutely tiny to satisfy not hurting others. I don't want slaughterbots to be science fact as much as anyone else, but these drones currently aren't small enough.
What if LEO don't have LoS?
What if that violent person is in the middle of a massive crowd?
Article says this drone can fly at 90 mph, which means it can get to the threat muuuuch faster.
For a gun, you also need personnel on the ground, which puts them at risk and they can cover only so much area. A drone at a good enough altitude with a streaming camera can send it back home for constant, automagical scanning and threat-profiling.
Absolutely not. That and the "rooftop sniper" idea are how you get Bloody Sunday. Firing into crowds is indiscriminate mass murder. As is detonating bomb drones in the crowd.
I largely agree at this point in the proposed capabilities. As you say these are just hand grenades that fly in cool looking formations. Independently guided aerial buck shot.
But the game changer is when these cheap independent systems can communicate and coordinate an intercept task against SOTA weapons systems - namely jets and nuclear subs.
As we know from birdstrikes, just a collision at the speed jets are moving at will probably neutralize it. To do this, the swarm will create several layers of buckshot that the jet can't avoid every one of them. If the jet is able to evade this first cloud, that cloud will now have very accurate targeting data to relay on to the next cloud.
And there will no longer be five billion dollar, 100-person manned subs to intercept other subs, all we'll need is an omni-present net of autonomous buoys that can "collapse-in" on any detected threat.
Depends. How fast are the drones? How fast is the SOTA plane they're trying to intercept? Trying to build a build a cloud in front of the jet is difficult when it's 20 times faster than the elements of the cloud.
It's even more difficult when the plane can fly on the deck or at 60,000 feet, or anywhere in between. Now you have to build a cloud 60,000 feet tall in front of something 20 times faster than the cloud. That is non trivial.
[Edit: Oh, yeah, what's the range of the drones? I read 6 km; is that accurate? If so, good luck getting them in front of anything.]
Agree. Even the murderdrones black mirror scenario posits a two-stage system: ~5 mile transport/breech-stage + 100-foot-range intercept unit. All to attack a stationary school building. So yeah, intercepting jets is going to need a another stage. And I think the intercept stage might look more like a guided mortars than a quadcopter.
The problem posed in the slaughterbots video wasn't the technology itself being inherently dangerous (although that was a considerable factor), but instead it focused on what happens when it gets compromised or otherwise falls into the wrong hands. The small, mass producible nature of the slaughterbots made it very easy for this to happen, and was the worst case scenario for fully autonomous weapons in general.
That said, I think we are right to be concerned about these "suicide drones", because they are a lot smaller than the fixed wing drones with the missiles, and as such would be harder to detect. I'm picturing these things dropping out of the back of a small aeroplane, one by one, nearly invisible to radar as they fall and land themselves in waiting for troops on the ground to collect them. Alternatively, they fall from the aeroplane and take flight to a GPS location of some small, but bombable fixture like a radio mast or portacabin.
This is true, there is very little stopping someone strapping a grenade to an off the shelf drone and flying it into a crowd, other than regulation on the explosives themselves, but that's not something that can easily be mass produced. Anyone with enough skill and determination can produce a weapon by themselves, I just think mass production makes it far easier for ill-willed people to obtain a ready-to-go weapon.
I still think the bigger risk, currently, is from the people buying these legitimately. These are not much more than precision bombs in their current state, and the risk is not from terrorism but war, and specifically how close to home that war can get. When the Salisbury poisonings happened it scared a lot of people, so imagine the panic if something like this was used to attack a van or a small building in a populated area.
Most of the hand-wringing about this strikes me as pretty far-fetched. While I have no doubt that attack drones will be used extensively on the battlefield, I think their use beyond that will be isolated. There are few advantages to using drones as a weapon that aren't already available to both terrorists and states. If you want to hunt specific people down and kill them, you don't need a drone. And if you want to cause mass casualties, there are much better options today than a swarm of hundreds of fragile one-time-use drones. Why don't we have more terrorist attacks using guns or truck bombs? Why don't we have drone IEDs right now? Why aren't drones taking down airplanes and helicopters all over the world? Why don't we have SAM attacks at every airport in the world?
The nightmare scenario most people seem to be worried about is the "slaughterbots" one from that video, but it doesn't make any sense to me. In some mythical future, we'll have the ability to use a tiny explosive charge to blow a hole in a specific person's head? I guess the person who came up with this missed the memo about the 300 million devices we have in this country that can use a shaped explosive charge to put a hole in someone's head.
And who is the entity that's releasing these drones to kill half a city? If it's our government, they could clearly do this now if they wanted to. Why don't they? If it's another government, that's no different than any other act of war they could currently carry out. What advantage does this offer? If it's a terrorist organization, again, they have better and cheaper options available today. And it's not like we couldn't stop them: who was flying that giant plane? Who were those guys releasing these swarms from the van? Where did they buy these drones? Where did they get the explosives? Just like many other threats, we'd track them down.
I'm not saying there's no risk, just that I think it's completely overblown. Every new technology carries a risk that a malicious actor will use it for harm, but the same technological advancements also enable us to better defend ourselves. So maybe there is some future dystopia where huge swarms of anonymously-controlled "slaughterbots" roam the skies looking for targets to murder, but if we have the ability to do that, we also have the ability to maintain defensive swarms, personal EMPs, chaff and smoke emitters, laser sentries, broad spectrum signal jamming, automated hacker entities, or whatever else we come up with.
The problem I have with it is this part of your statement:
> And it's not like we couldn't stop them: who was flying that giant plane? Who were those guys releasing these swarms from the van? Where did they buy these drones? Where did they get the explosives? Just like many other threats, we'd track them down.
I think that might not be true. Sure, the slaughterbots video showed a swarm released from a van with operators present. Seems like somebody might notice. But it's easy to imagine a scenario where the operators leave the drones in some out-of-the way place, inactive, without arousing suspicion, and they activate on a timer to carry out their mission. How would you track down who committed the act?
A bomb can go off on a delay, but requires the killer to go to the target site to place it. A gun can kill from a great distance, but only with line of sight, the killer needs to be involved at the time of the shot, and even with a silencer it makes a fairly loud noise, drawing attention from those near the shooter. The fictional-for-now slaughterbots scenario would enable killing people at an event without ever going to the event site, without even being in the same city at the time of the event, and at low cost.
Remember the scramble to identify and track down the Boston marathon bombers? They attended the race, set down bags with bombs in them, and then left. Fortunately they were both caught on camera and positively identified by surviving victims. What if they didn't have to go to the race at all, but instead went a week before to an abandoned building, inactive construction site, etc and set the hypothetical slaughterbots in place, set to launch later? Would they ever be caught?
The combination of low cost, low skill needed by the deployment crew, and low risk of being caught (with corresponding deniability by state actors) is the frightening thing.
The other end of it is that you could catch somebody if you had near universal surveillance -- you trace the drones back to their launch site based on reports/footage of them flying midair, and you pull footage from cameras that were watching there too. Or you stop it before it happens with massive data collection online, trying to detect the planning, discussion, or purchasing of equipment. Neither is going to be 100% effective but they'll be enough to get bills passed authorizing more surveillance after a couple attacks.
So you could have a situation where terror attacks happen more frequently, or a situation where everybody is subject to a surveillance state granted even more power than the one we have today, or a little from both columns. Not thrilled about that.
Obviously the tech is not there yet for the bots to execute attacks with autonomy like depicted in the video. But it seems very possible within this century.
Seen one reference to slaughterbots already, but even without AI on board these are terrifying. These could be small enough to have little to no radar return, so just dropping them one by one out of an aeroplane would be a viable delivery method for stealth killers.
145 comments
[ 3.2 ms ] story [ 176 ms ] threadI don't really disagree with you, just asking.
https://www.youtube.com/watch?v=9fa9lVwHHqg
Although the above is overdramatised, it's not unrealistinc.
Its allies used to rent drones for surveillance against terrorism, then used it as both a carrot and a stick. Now, Turkey has its own drones that regularly take down targets effectively, even some air defenses.
After Syrian war started and "modern western liberal democracies" failed to help with the refugee crises, those borders became even more treacherous.
Heard that before?
...and it is laughable to suggest EU will lift another finger for Greece after taking their wallet and vacationing with that money too
As for refugees, they weren't forced. The border police just didn't stop them. I agree Erdogan coming on TV and yelling it out openly was a below-the-belt tactic and after that people started transporting refugees by the bus to the border.
Turkey's government is just the big despotic fish in the small geopolitical pool that is the Caucasus / Northwest Middle East. Finding cheap ways to kill is not a defensive strategy for the gov't.
So yes, a consumer can build a guided missile.
I wonder what chance anyone has of stopping something like this. It can certainly carry an explosive and flying 50 or even 200 km is probably doable with a bit of a scaleup.
This just sounds like you're trying to persuade people on why drones are definitely the future of war.
Dunno about "untraceable" though, there's a long history of tracing munitions from fragments.
Private individuals will soon be able to afford automated drone swarms. Set it and forget it.
An uncontrolled nuclear reaction is significantly simpler than a controlled one. We still don't have fusion reactors, while fusion bombs were developed 10 years after fission bombs.
The Chicago Pile came way before Trinity. Rock formations have even been known to come together accidentally to form natural nuclear reactors, but I think it will be a very long while before the first natural nuclear bomb.
Probably in https://en.wikipedia.org/wiki/Human_radiation_experiments the section about USA is bigger because it is more open, but it is scary anyway.
> During World War II, the Second General Army and Chūgoku Regional Army was headquartered in Hiroshima, and the Army Marine Headquarters was located at Ujina port. The city also had large depots of military supplies, and was a key center for shipping.
So, a civilian city, but not just that.
Next thing you know there is an AI detecting crime and killing the criminal without making a mess.
looks up how an EMP works
You can detect what the other person is carrying and that works good as a filter. A person carrying knife in a crowd or school outside of canteen doesn't make any sense by general norms. Second would be to detect the intent - now that's a much harder problem but you can whitelist few situations where the person has already stabbed someone or shown sufficient physical evidence of criminal intent.
You can even think up a dystopia where Facebook does the policing.
"Linkedin implements a system to identify to-be criminals by personal history and behaviour it tracks", in a report on Friday from the executives, Linkedin has decided to label offenders on its platform to reduce their reach and stop them from getting a job before it's too late. The CEO explained that this is just one of the policies that linkedin has implemented to serve justice to our society.
Mark Zuckerberg says "justice shall be served. Social media platforms have the duty to govern society". Facebook has decided replace profiles of people breaking its policy with support social justice banners and links to donation box of variation NGOs. Additionally, they have decided that this policy will effectively block all offenders accounts on any platform they control including instagram, whatsapp, and integration which uses Facebook sign in functionality. They will share this data publically and encourage others such as partnered advertisers to block those offenders.
After years of toxicity on Twitter, jack Dorsey says they are finally taking steps to serve justice to our society. New twitter update will make it mandatory to attach your credit card at all times and social security number as well as sensitive personal information. Offenders will be charged and fined on the platform and their accounts will be handed out to the victims for exposure therapy.
Sundar Pichai thinks google must fix people's wrong perception and biases. Google representative told us that they would soon manipulate search and ad results of people it deems problematic in an attempt to "fix" them. Racist people will start seeing more anti-racist content in the coming days.
Putting your head in the sand won't save you.
the amount of money put into may-be-probable piece of tech is astronomical in defence indistury, because it may really be game-changer. (eg. lasers, rail-guns etc)
every piece of tech we use today started as a defence project. You may say ML/AI is just research stuff. If you think AI is being developed to help you with groceries, you are being lame... (no harm intended but my opinion)
The issue I have is when this gets actually exported to countries that may want to use against civilians (to squash opposition) though it was know for a long time this must happen.
Turkey wouldn't need to export it to anyone for it to be used against civilians, that's a pretty common thing for Turkey.
If anybody started threatening commercial civilian flight, countries would fall on them like bricks. Quarantine for a few months was devastating on many fronts. Now, think if all flights were grounded for months...
Maybe kamikaze drones would be more apt.
No international agreement will survive war, they are just used to suppress and bully countries.
Humans can (and must) agree to be better than frightened animals.
Tactical inconveniences that limit utility are what's preventing chemical weapon usage. Nobody uses chemical weapons because we have better, more effective ways to kill each other making them not really useful except in limited circumstances like terrorizing civilians (a static target that probably doesn't have gas masks) in areas where you have air superiority (e.g. Syria now, Iraq decades ago, etc.). Chemical weapons were mostly abandoned after WW1 because they just didn't work well enough for the effort and resources required to use them.
Edit: https://acoup.blog/2020/03/20/collections-why-dont-we-use-ch...
We’re in a new arms race. Better drones. Better anti drone systems. Hopefully better pizza delivery.
Yeah, and 911 was an inside job!
EDIT: If you think China has thousands of sleeper drones in the US you are on par with people who think 911 was an inside job, end of story. Down vote all you want but be sure to don your tin foil hats afterwards.
Why is China your example and not X? There are other countries that use killer drones, do secret killing and kidnapping operations on other countries territories etc.
I have the impression that a few years back you would have said Russia, then N Coreea and today China, like the big bad enemy is changing each few years depending on what politics need.
The good old fashion alternative is just a gunman. It's more plausibly deniable - e.g. still nobody is exactly sure what was going on with Lee Harvey Oswold and Jack Ruby - and it doesn't put the targeted country immediately on a war footing.
Gee, chaos, that would be terrible.
https://www.quora.com/How-much-would-it-cost-to-build-an-EMP...
Similar to tanks vs. machine guns, but the machine guns needed human operators.
I'd think IR (e.g. https://www.sparkfun.com/categories/139) could see through smoke, otherwise lidar keeps getting cheaper (https://www.sparkfun.com/categories/tags/lidar).
We'll need defensive drone swarms.
Drones are the first robots that are among us (overhead). Self-driving cars are at least very expensive to produce. But drones can be mass-produced and deployed anonymously.
Drones with live munitions is terrifying. Imagine an anonymous swarm descending on Union Square in SF.
Even under normal operation, drones can fall and hit someone by accident. Let alone a swarm.
This is one technology I should not want to be in anyone’s hands, but if it is, it should be in the hands of the State and not everyone. Just like bombs.
PS: With cheap cameras being ubiquitous and people can be identified by their gait, heartbeat etc. consumer level tech can also include an AI that can cross correlate everywhere you’ve been. Businesses would use it to fine you when you park your car for too long and don’t go into the store.
But governments and police departments can concoct tons of believable stories with parallel construction and put anyone they want away. In fact AI can also be adopted in other places in the policing and judicial process (as with Palantir) leading to a Kafkaesque experience for sentencing.
And with deepfakes it’s worse than that. Anyone could create a believable video of a rape-caught-on-video or a damaging “speech at a private event” or whatever. Reputational attacks were cheap already (see CIA handbook) but with a swarm of sleeper twitter bot accounts and other bots you could introduce anything.
In short I am worried we are going to be entering a dystopia soon. This video nails the drones issue:
https://m.youtube.com/watch?v=9CO6M2HsoIA
It's only now, 20 years later that the children of these protestors are repeating the same things their parents did. And it also explains why certain politicians try to associate protest with terrorism these days, that it's a different generation protesting explains why this association falls flat.
Today's protestors were not born when the towers fell.
[1]https://en.wikipedia.org/wiki/1999_Seattle_WTO_protests [2]https://en.wikipedia.org/wiki/Seattle_Mardi_Gras_riot
Assume that the leadership in Turkey is nothing more than a puppet regime fullfilling those interests.
Only if they're remote controlled. Nothing to jam if they're autonomous and using the right navigation sensors.
> fry their circuits with some sort of electrical guns
EMPs are painful for everyone involved, not just the target.
> incentive to develop some good defense against them
Far as I can tell, autonomous drone swarms are the Nash Equilibrium.
But, even if that was so, is there no way to have some defense against such attacks, having that they're electric and rather slow? Tesla ray guns come to mind but I am not very knowledgeable in this field.
After all, not everyone can deploy their own custom cpu. And the REAL worry is when a script kiddy gets one of these and deploys them at the next protest. I dont want to say the 'T' word. By making commodity hardware fragile and defeatable, we protect ourselves against our dark side. And no matter how good an AI gets at face recognition/inference/etc, we will still be able to turn it off. Did we not learn the lesson of 'Colossus: The Forbin Project'
But, to be honest, I’m not really worried. It doesn’t really change the tactical situation of today. It just makes the attack vector cheaper.
The USA already uses drones that loiter from 30 miles away, and has a high resolution camera that can see you. Then you’re targeted and a Hellfire missile is locked on, and shot at the target.
Same effect. The target is killed.
However, this technology increases the critical retaliation magnitude.
If the attacker chooses to assassinate the political leadership of the enemy country by suicide drone, then the victim country can retaliate with an all out nuclear assault.
So, we are now basically at the same situation we were at before. MAD principles still apply. Nothing has changed.
We've had long-range (longer range than most drones) guided missiles for decades. We've had drones that can launch them, for decades. Those technologies already have the ability to wipe out targets anywhere, anytime, en masse, with little to no risk. Yet we don't see people (as one commenter suggested) wiping out entire plots of land so they can start a new country. We don't see (most) countries turning such weapons on their own citizens. We don't see (very many) heads of state being assassinated in this way. What this tells me is that there are sociopolitical forces that make such things un-advantageous or infeasible, for one reason or another.
The only true differentiator I can see with suicide-drones is a potential increase in precision, though that's not really true right now because these in the article look to be basically flying hand-grenades. That's not really better than a missile; in fact it's probably a lot easier to intercept or shoot down. Now, many of us have seen that (fictional) video with the micro-explosive assassination drones. Maybe that would change the game in some way, but I still don't think it's a drone apocalypse we'd be looking at. Nation states have so many ways of killing people at this point. When they don't, it's usually because it doesn't make sense to, not because they lack some key capability.
A drone can be used for more than a singular purpose. There might be more efficient ways to kill someone we already know like you said but what if we don't know them yet?
What if they don't exist yet?
Imagine you have a crowd and there is one violent person inside it, how would you immobilize such a person without harming the others?
Why pay for their pension years after they have retired?
Why pay for their healthcare, training, damage, etc?
Even the most insidious state relies on its citizenry to some extent. You're not ruling anything if your people are constantly in revolt, or if they're all dead.
Yes, there is a money problem for employing enough cops.
Reference - India.
https://en.m.wikipedia.org/wiki/List_of_countries_and_depend...
In an ideal future, autonomous drones would make less mistakes since they don't feel threatened or care about dying. They won't kill someone just because it looked threatening. They won't turn off the body cam because of reasons. They will have a more accurate aim and will be more trained since training doesn't cost anything for them. You can invest the other funds in improving rehabilitation or de-escalation services.
Hmm...a gun?
Humans aren't good in stressful situations and their aim isn't great. A ground view isn't necessarily helpful in situation where the criminal might be inside the group ring.
Also sounds like the cost for those snipers would be pretty high than manufacturing few drones to lock in a target and swarm under their head.
Cost of drones is only going to decrease while the cost of sustaining human life remains more or less the same if not increase.
A drone is $30k
But the game changer is when these cheap independent systems can communicate and coordinate an intercept task against SOTA weapons systems - namely jets and nuclear subs.
As we know from birdstrikes, just a collision at the speed jets are moving at will probably neutralize it. To do this, the swarm will create several layers of buckshot that the jet can't avoid every one of them. If the jet is able to evade this first cloud, that cloud will now have very accurate targeting data to relay on to the next cloud.
And there will no longer be five billion dollar, 100-person manned subs to intercept other subs, all we'll need is an omni-present net of autonomous buoys that can "collapse-in" on any detected threat.
It's even more difficult when the plane can fly on the deck or at 60,000 feet, or anywhere in between. Now you have to build a cloud 60,000 feet tall in front of something 20 times faster than the cloud. That is non trivial.
[Edit: Oh, yeah, what's the range of the drones? I read 6 km; is that accurate? If so, good luck getting them in front of anything.]
That said, I think we are right to be concerned about these "suicide drones", because they are a lot smaller than the fixed wing drones with the missiles, and as such would be harder to detect. I'm picturing these things dropping out of the back of a small aeroplane, one by one, nearly invisible to radar as they fall and land themselves in waiting for troops on the ground to collect them. Alternatively, they fall from the aeroplane and take flight to a GPS location of some small, but bombable fixture like a radio mast or portacabin.
Control on explosive substances (and on aircraft themselves, to a degree) is what. These drones don't change that equation.
I still think the bigger risk, currently, is from the people buying these legitimately. These are not much more than precision bombs in their current state, and the risk is not from terrorism but war, and specifically how close to home that war can get. When the Salisbury poisonings happened it scared a lot of people, so imagine the panic if something like this was used to attack a van or a small building in a populated area.
The nightmare scenario most people seem to be worried about is the "slaughterbots" one from that video, but it doesn't make any sense to me. In some mythical future, we'll have the ability to use a tiny explosive charge to blow a hole in a specific person's head? I guess the person who came up with this missed the memo about the 300 million devices we have in this country that can use a shaped explosive charge to put a hole in someone's head.
And who is the entity that's releasing these drones to kill half a city? If it's our government, they could clearly do this now if they wanted to. Why don't they? If it's another government, that's no different than any other act of war they could currently carry out. What advantage does this offer? If it's a terrorist organization, again, they have better and cheaper options available today. And it's not like we couldn't stop them: who was flying that giant plane? Who were those guys releasing these swarms from the van? Where did they buy these drones? Where did they get the explosives? Just like many other threats, we'd track them down.
I'm not saying there's no risk, just that I think it's completely overblown. Every new technology carries a risk that a malicious actor will use it for harm, but the same technological advancements also enable us to better defend ourselves. So maybe there is some future dystopia where huge swarms of anonymously-controlled "slaughterbots" roam the skies looking for targets to murder, but if we have the ability to do that, we also have the ability to maintain defensive swarms, personal EMPs, chaff and smoke emitters, laser sentries, broad spectrum signal jamming, automated hacker entities, or whatever else we come up with.
And in the end, even those jobs were automated.
Remember, how poorly people are treated now and in the past is how we are treated while the 1% needs something from us (our labor).
Do you honestly expect better treatment when we are useless eaters?
> And it's not like we couldn't stop them: who was flying that giant plane? Who were those guys releasing these swarms from the van? Where did they buy these drones? Where did they get the explosives? Just like many other threats, we'd track them down.
I think that might not be true. Sure, the slaughterbots video showed a swarm released from a van with operators present. Seems like somebody might notice. But it's easy to imagine a scenario where the operators leave the drones in some out-of-the way place, inactive, without arousing suspicion, and they activate on a timer to carry out their mission. How would you track down who committed the act?
A bomb can go off on a delay, but requires the killer to go to the target site to place it. A gun can kill from a great distance, but only with line of sight, the killer needs to be involved at the time of the shot, and even with a silencer it makes a fairly loud noise, drawing attention from those near the shooter. The fictional-for-now slaughterbots scenario would enable killing people at an event without ever going to the event site, without even being in the same city at the time of the event, and at low cost.
Remember the scramble to identify and track down the Boston marathon bombers? They attended the race, set down bags with bombs in them, and then left. Fortunately they were both caught on camera and positively identified by surviving victims. What if they didn't have to go to the race at all, but instead went a week before to an abandoned building, inactive construction site, etc and set the hypothetical slaughterbots in place, set to launch later? Would they ever be caught?
The combination of low cost, low skill needed by the deployment crew, and low risk of being caught (with corresponding deniability by state actors) is the frightening thing.
The other end of it is that you could catch somebody if you had near universal surveillance -- you trace the drones back to their launch site based on reports/footage of them flying midair, and you pull footage from cameras that were watching there too. Or you stop it before it happens with massive data collection online, trying to detect the planning, discussion, or purchasing of equipment. Neither is going to be 100% effective but they'll be enough to get bills passed authorizing more surveillance after a couple attacks.
So you could have a situation where terror attacks happen more frequently, or a situation where everybody is subject to a surveillance state granted even more power than the one we have today, or a little from both columns. Not thrilled about that.
Obviously the tech is not there yet for the bots to execute attacks with autonomy like depicted in the video. But it seems very possible within this century.