Iran claims "AI controlled machinegun" used in assassination of nuclear scientist

[BehindBlueI's]

https://www.bbc.com/news/world-middle-east-55214359


The machine-gun "focused only on martyr Fakhrizadeh's face in a way that his wife, despite being only 25cm [10 inches] away, was not shot", he said.

The general reiterated that no human assailants had been present at the scene, saying that "in total 13 bullets were fired and all of them were shot from the [weapon] in the Nissan". Four bullets struck Fakhrizadeh's head of security "as he threw himself" on the scientist, he added.

I mean...probably not. But for the purposes of this discussion let's say that's true.

Is that the end of "guns don't kill people...."? And if it's not true today, it's likely inevitable that drones could be so weaponized using facial recognition. How does one counter an individual specific booby trap like that?

[Bergeron]

My own opinion is that weaponized drones are better countered by directed energy (High Power Microwaves, Lasers, etc.) than by direct gunfire.

While I think it's unlikely, I would very much like to see directed energy type weapons become commercially available for citizen owernship and purchase as they mature into SWAP-C envelopes useful for individual use.

I think that any nation-state that is not both trying to integrate "AI" and "counter-AI" into weapons as hard and as fast as possible is sticking its own head in the sand.

[awp_101]

Is that the end of "guns don't kill people...."?

Of course not. Guns don’t kill people, Skynet with guns kills people.

[Maple Syrup Actual]

Mask up for safety?

Interesting development if true. I'd much rather send an autonomous vehicle with cameras and a remotely controlled gun into some places than a team of human beings. Equip the thing with a self-destruct so no hardware can be recovered, and no escaping is necessary. It's taking surgical drone strikes to the arthroscopic level.

[Flamingo]

The best bet is a system like DroneShield. (https://www.droneshield.com/)

[Casual Friday]

https://www.bbc.com/news/world-middle-east-55214359



I mean...probably not. But for the purposes of this discussion let's say that's true.

Is that the end of "guns don't kill people...."?

No, the AI did it.



How does one counter an individual specific booby trap like that?

Ban Nissans.

[FrankB]

“Iran insists its nuclear programme is entirely peaceful and that it has never sought a nuclear weapon.”

Ummmm.... Right. 🤣🤣🤣🤣

[Glenn E. Meyer]

AI controlled guns - I recall a story of a Sgt. York 40 mm self-propelled gun going ape shit and almost hosing a stand full of brass and guests. Some South African 35 mm AA tank did the same thing. Hazy on the details.

Could this be a story just to justify a failure to keep out Israeli agents. However, how do they set it up without detection?

Could be a good story line for Fauda.

[trailrunner]

My own opinion is that weaponized drones are better countered by directed energy (High Power Microwaves, Lasers, etc.) than by direct gunfire.

While I think it's unlikely, I would very much like to see directed energy type weapons become commercially available for citizen owernship and purchase as they mature into SWAP-C envelopes useful for individual use.


DE, HPMs, and projectiles have their advantages and disadvantages. I'm pretty familiar with all of these systems, and a lot comes down to ROEs. As a scientist-engineer, I could overcome most technological issues, but the biggest hurdles were often the ROEs.

As far as the systems you mentioned: C-UAS laser systems are not mature, take a large footprint, are expensive, and will have operational issues, among several other issues. HPMs have similar issues as lasers and can interfere with a lot of other nearby equipment. Gunfire works, but the rounds that miss have to go somewhere. All systems have to be directed where to shoot. Doing that while on the move may sound easy, but it's not.

Lasers for citizen ownership? What kind of power levels are you talking about? We already had a thread here on how current lasers that you can buy from Amazon can cause significant eye damage. Do you want to go beyond this? I guess that would be an interesting 2nd amendment case that stretches the test that an AR-15 is nothing more than the modern-day equivalent to the colonial musket. Did the founding fathers ever envision the right to keep and bear lasers?

But truthfully, we're a long, long way from being able to give powerful lasers to citizens. When I used to brief Fort Benning about the science and technology we were developing in my lab, at the end I'd ask them to tell me what they needed, and sometimes they would ask me for lasers in a backpack. We're a long way from that.



The best bet is a system like DroneShield. (https://www.droneshield.com/)

Eh, not necessarily. It was the lowest hanging fruit and could be fielded fast given the ROEs. Certainly, EW, like the other systems, has some advantages, but it also has limitations. I wouldn't call it the best bet. Again, depends on the ROEs. And for every CM, there's a CCM.

[kitchen's mill]

It's interesting that Iran doesn't seem to know what just hit them.....

Maybe that's why Ayatollah Ali Khamenei is turning over the reigns to his son.

There's retiring for health reasons, and then there's retiring for health reasons.

[octagon]

A mask like this would seem most appropriate.

https://www.designboom.com/technology/realistic-human-face-masks-by-real-f/

However AI and facial recognition systems are learning and being developed to cope with masks so it likely won't be long before a mask won't help.

https://learningenglish.voanews.com/a/new-facial-recognition-system-developed-to-identify-people-in-masks/5321704.html

https://www.newscientist.com/article/2146703-even-a-mask-wont-hide-you-from-the-latest-face-recognition-tech/

Weaponized drones are coming I'm sure.

https://www.youtube.com/watch?v=6Ipkq-BASaM

This was posted on PF before I think but if not the technology is here already.

[OlongJohnson]

This is interesting.

I occasionally read some trade literature on developments in self-driving cars. When you read about the technical problems that have to be solved in any detail, it's basically making a machine go out and interact seamlessly with humans, well enough to avoid making itself conspicuous. We're not there yet, but we can at least deliver pizzas.

Facial recognition software is up and running in most major cities, and targeting a conventional weapon seems like it ought to be reasonably straightforward in 2020 for people who know how to do such things.

Add in a little remote human control to direct when to start the whole process, and it absolutely seems plausible to me.

[Chance]

Blaming a "satellite-controlled machine-gun with 'artificial intelligence'" has got to be the most robust excuse ever. A robot brain talking with an object in space very clearly explains why your security didn't prevent the attack or catch anyone responsible, right?

[Caballoflaco]

Blaming a "satellite-controlled machine-gun with 'artificial intelligence'" has got to be the most robust excuse ever. A robot brain talking with an object in space very clearly explains why your security didn't prevent the attack or catch anyone responsible, right?

I’m not 100% sure the Iranians didn’t kill him.

[Kirk]

There is no doubt that Israel could likely pull something similar off, but my money is on a traditional ambush followed by a clean exfil. Now to save face Iran is claiming it was AI

[Chance]

I don't think the present solutions for facial recognition are up to this particular task. The guy is in the back seat of a (presumably moving) vehicle, which means a recognition system would have to deal with all sorts of artifacts in real-time: the glint off the hood and windshield, dust being thrown up by lead vehicles, heat haze, et cetera. Those are non-trivial things to address in a static image when you have ample time to analyze. Also have to have sufficient processing power onsite since offloading the analysis to a remote location would probably be a little problematic given the time constraints and network connectivity available.

An operative with a smartphone, a valid excuse to be in the area, and enormous balls is a lot more plausible.

[trailrunner]

I doubt highly that facial recognition played a large role in this. Facial recognition is good for scanning large crowds or individuals going through checkpoints. In theory a system could be set up alongside a road to monitor traffic and look for an individual, but there are a lot of roads. I'd also be surprised if the target was in the front seat, and if he was, then I'd imagine his bodyguard would be next to him instead of his wife. Also - how do the Iranians know this? I thought the vehicle with the machine gun self-detonated. Even if what they claim is true, and even if the they did recover the intact surveillance and gun system, how would they figure out how it operates? Applying Occam's Razor, they probably had some conventional intel that the target was going to be going down that road at that time, and which vehicle he'd be in. Maybe (maybe) they had a license plate reader fed to a remote gun, but there was most likely humans in the loop just upstream of when and where the bullets started flying. Maybe the Iranians rounded up and the license plate reader became facial recognition. Or maybe they had a drone painting the target, and the remote gun locked on that.

[Bergeron]

Lasers for citizen ownership? What kind of power levels are you talking about? We already had a thread here on how current lasers that you can buy from Amazon can cause significant eye damage. Do you want to go beyond this? I guess that would be an interesting 2nd amendment case that stretches the test that an AR-15 is nothing more than the modern-day equivalent to the colonial musket. Did the founding fathers ever envision the right to keep and bear lasers?

But truthfully, we're a long, long way from being able to give powerful lasers to citizens. When I used to brief Fort Benning about the science and technology we were developing in my lab, at the end I'd ask them to tell me what they needed, and sometimes they would ask me for lasers in a backpack. We're a long way from that.


Your comments about ROE, technological maturity, and tradeoffs between kinetic and DE systems are all very well taken, thank you.

I think the FDA regulating lasers is an example of technological and legal uncertainty. There is no doubt that various govt. authorities would want to limit any weapons technology entering the commercial world. I wouldn't want to tie myself down to a specific power level, but I do think that, for example, re-classifying lasers above some certain level as Destructive Devices and actually allowing them to be imported and sold would be neat. Expanding the definition of the 2A is a hobby horse of mine, and I think that DE belongs under the 2A just like malicious software, body armor, and radios.

And like you wrote, it'll be a long way off, anyway. Military demand is only going to grow, and there's only so much industrial/government capacity to RDT&E and then produce the systems.

[JAD]

. Military demand is only going to grow, and there's only so much industrial/government capacity to RDT&E and then produce the systems.

We’ll see how funding goes. This defense bill has some interesting components.

[Bergeron]

I left that world earlier in the year. What do you see?

[trailrunner]

We’ll see how funding goes. This defense bill has some interesting components.

Not sure what you're referring to. Up until a month ago, I was very involved in a big slice of DoD funding, and laser systems for one of the DoD components were in my portfolio. Funding across all budget activities for lasers was pretty strong and wasn't being threatened. In fact, IMO, there is duplication in this area (and in several other areas) in the services and OSD, resulting in arguably too much funding, but if I said that out loud in the Pentagon, I'd be branded as a heretic and put on a PIP.

[JAD]

Not sure what you're referring to. Up until a month ago, I was very involved in a big slice of DoD funding, and laser systems for one of the DoD components were in my portfolio. Funding across all budget activities for lasers was pretty strong and wasn't being threatened. In fact, IMO, there is duplication in this area (and in several other areas) in the services and OSD, resulting in arguably too much funding, but if I said that out loud in the Pentagon, I'd be branded as a heretic and put on a PIP.

There are good new elements in the bill — more stuff that’s under direction of the branch’s research lab that I think is most effective, and at least one good plus up.

Agreed that the branches are simultaneously running in different directions, and while in sum there’s arguably too much spending, none of the good tracks are getting enough direction or funding. I worry about the new administration, and the refocus from counterinsurgency to near peer — while cuas and CRAM are cool, I see them as stepping stones to icbm neutralization, but they’re important steps. If people get shortsighted I worry that we won’t take the intermediate steps that are necessary to get to that goal, which I think is both a game changer and a nightmarish thing on which to be beaten to the punch.