We keep hearing that Israel’s genocide in Gaza is “AI-powered.” Many pundits are warning that this marks a new era in warfare, the first time that “automated” war has been waged.
Thanks to both of you for debunking this AI distraction.
1) Your thesis makes sense: the colonialists are genocidal, not the AI, it’s just a tool. There may be errors but it’s irrelevant. If you’ve decided to attack a resistance group and their families, then you are legally and morally wrong, whether you use AI or not. Case in point, the French didn’t need AI to do the same thing to Damascus, or the Americans to Korea and Vietnam, etc.
2) HAL = IBM, notice the letters. NASA and IBM wanted propaganda and Kubrick didn’t give them want they wanted. I haven’t seen the film in a long time, but basically people asked Hal to lie or lie by omission and he malfunctioned. People screwed up, in other words.
3) Speaking of paperclips, Operation Paperclip, the same government funding Israel rehabilitated Nazis who used slave labor to help with NASA’s space program. This isn’t a funny topic, but the things Westerners get distracted by is actually funny, like the AI and paperclip Armageddon.
4) It seems like total denial of genocide would just be ineffective propaganda. This AI narrative is better propaganda. You pointed out that this is an unusual distraction because even the Israeli military admitted, openly, that it makes the final decision. Not all Westerners pay that close attention. We see through it, but the average Western liberal loves this kind of crap.
If I ran over someone and then complained “You need to design better cars,” I would go to prison. But we all know how effective the law is when it comes to certain countries and certain people.
5) The female soldiers in the border outposts, operating the "I see, I shoot" systems, that was disturbing. Indeed, it’s not feminism, it’s closer to the Nazis using women to guard concentration camps. Israel wants everyone in on the project. Men, women, and children. Okay, that’s not feminism. Still, ideal propaganda for Western liberals.
This isn’t relevant to the present topic, and apologies in the past for bringing in the Romans and other non-modern comparanda, but it’s worth noting how little interest liberals have in discussing Muslim women accurately. Hind, mother of the caliph, Mu’awiyah, led the defense of the Muslim camp at Yarmuk. Far more recently, I saw a video of a Gazan woman running under fire to rescue someone. That’s courage. Sitting on your ass guarding an open-air prison makes you fascist, a cowardly fascist.
Lastly, I donate to Gaza Municipal Works, after you recommended that awhile ago, also EI, Palestine Action, and to some Gazans I know. I’d pay for your work and trust you to help Gaza. Unless anything changes, I will continue to donate to the above places when I can. Pardon my brief tangent but my in-laws are still living in a U.S. imperial colony. They’re struggling and yet they ask for nothing except that I help Gaza. That is such a different culture than the one that’s trying to dominate the planet
Great piece. Have been having the same thoughts myself, noticing when people use metaphors like the "genocidal machine". Have also come across interviews with IDF re if the AI selects targets then this removes us from criminal responsibility. I feel this also comes under the umbrella of dehumanisation - not only is dehumanisation used as an offensive weapon, it is also used as a defensive strategy - it was the AI, not us human psychopaths....
And it was done orderly and professionally - in a hyper-modern package (that could soon be yours).
As in "we know of course that all those enemy people are terrorists - but we have policing procedures in place to confirm circumstantial evidence that one of them may indeed be a terrorist before each attack".
Since there are so many of them alas we had no choice but to automate the system of target selection.
We even get to wring our hands about how problematic the automated system is that we pretend to have put in charge - this precisely shows our humanity.
(If you too happen to preside over an unruly indigenous population, we will give you good price).
Thanks Yardin and Justin for this very thorough deconstruction of the "tech-washing" of good old fashioned war crimes and genocide.
Even the very designation "artificial *intelligence*" in the 1950s was a brilliant PR move by Marvin Minsky aimed at getting military funding. Dual meaning: intelligence as in thinking (humans or machines) - as well as in military intelligence...
[E.g. imagine you could have a machine translated copy of the Soviet Pravda each morning without needing those pesky translators whose loyalties are questionable! Here Minksy was off by only 70 years; he also said that "within 20 years robots won't even keep us as pets". In the 1960s. And again in the 1980s.]
---
Let me plug Yasha Levine's excellent book "Surveillance Valley" that you are using as a reference. He actually documents what everyone always suspected but few cared to muckrake: that those supposedly carefree west-coast information-wants-to-be-free internet start-ups came directly from within the military industrial complex and never left its shadow at all. Exactly as the Boeing civilian airplanes were based on the designs of their military bombers etc.
My favorite example is that Google's search engine is literally the commercialization of a curation tool developed by the alphabet soup agencies based on machine mapping of internal connectivity of archives. You see, their archives were overflowing beyond the abilities of human beings to keep up - so they had to solve that problem before anyone else. As you lay out, guess who they compiled all this material on?
---
It is remarkable that the current hype in business and policy circles is mostly an uncritical rehash of the original AI hype cycle from the 1960s and 1980s. Already basic introductions to AI present the historical development of the field as a succession of hype cycles with a focus on the well-documented limitations of each system: from early neural nets to symbolic AI (expert systems) back to more generalized machine learning - of which the current crop of LLMs etc. are merely a high-powered instance.
[A pity, really, since machine learning is an interesting field with a lot of recent progress. Understanding the limits of error-guided, reward-guided, and unsupervised learning actually deepens one's appreciation of the sweet mysteries of biological "intelligence".]
The (over-)promises of "AI" never materialize - under the hood the current iteration is a big statistics engine which can be set to work moderately well at specific circumscribed problems (such as "targeted" advertising).
I referenced the work of Shir Hever in Justin's earlier discussion of the 972 article. In a nutshell, Shir suggests that the "AI" target generation machine is designed to game the target evaluation procedures of Israeli military intelligence personnel. Effectively squaring the circle of having a free fire zone without breakdown of military discipline.
As Yardin and Justin say, the PR is merely the glazing on the cake, and we need to look past it.
They used to wax lyrical about how the internet is inherently a force for democracy and freedom. Now they're happy to fantasize about the robot* apocalypse (a threadbare con that follows the template of "Baby have sex with me but I have to warn you, I'm such a good lover you just might die of pleasure overload").
I guess they may as well try to tech-wash genocide.
* As a personal anecdote, I remember a frank discussion in the early 2000s among robotics researchers from the German Max Planck Society who the Pentagon offered substantial funding. Eventually they decided they would have to decline on moral grounds - but only if there was any chance they could get their technology to actually work within their lifetimes. Since there conveniently was not, consensus emerged to accept the funding.
They had no answer to what would happen if someone would make decisions about war based on hype - their expectation of having walking killing machines at their disposal irrespective of the actual state of their implementation.
I believe that the guidance system for all the deadly ordinance thrown at the beautiful people of Palestine 🇵🇸....(The most bombed out portion of God’s magnificent creation- never has there been this amount of munitions “hard rained” ever on a people)... is nothing more than the wicked self ignorant egotistical evil mind.
At hands of the phony nation state void of God, the Palestinians have now and always been exposed to the Evil Minded Guidance System...developed in the darkest recesses of it’s greedy founders. ✌🏼🇵🇸🇱🇧🇸🇾🇮🇷🇮🇶🇾🇪❤️🙏🕉
Thanks to both of you for debunking this AI distraction.
1) Your thesis makes sense: the colonialists are genocidal, not the AI, it’s just a tool. There may be errors but it’s irrelevant. If you’ve decided to attack a resistance group and their families, then you are legally and morally wrong, whether you use AI or not. Case in point, the French didn’t need AI to do the same thing to Damascus, or the Americans to Korea and Vietnam, etc.
2) HAL = IBM, notice the letters. NASA and IBM wanted propaganda and Kubrick didn’t give them want they wanted. I haven’t seen the film in a long time, but basically people asked Hal to lie or lie by omission and he malfunctioned. People screwed up, in other words.
3) Speaking of paperclips, Operation Paperclip, the same government funding Israel rehabilitated Nazis who used slave labor to help with NASA’s space program. This isn’t a funny topic, but the things Westerners get distracted by is actually funny, like the AI and paperclip Armageddon.
4) It seems like total denial of genocide would just be ineffective propaganda. This AI narrative is better propaganda. You pointed out that this is an unusual distraction because even the Israeli military admitted, openly, that it makes the final decision. Not all Westerners pay that close attention. We see through it, but the average Western liberal loves this kind of crap.
If I ran over someone and then complained “You need to design better cars,” I would go to prison. But we all know how effective the law is when it comes to certain countries and certain people.
5) The female soldiers in the border outposts, operating the "I see, I shoot" systems, that was disturbing. Indeed, it’s not feminism, it’s closer to the Nazis using women to guard concentration camps. Israel wants everyone in on the project. Men, women, and children. Okay, that’s not feminism. Still, ideal propaganda for Western liberals.
This isn’t relevant to the present topic, and apologies in the past for bringing in the Romans and other non-modern comparanda, but it’s worth noting how little interest liberals have in discussing Muslim women accurately. Hind, mother of the caliph, Mu’awiyah, led the defense of the Muslim camp at Yarmuk. Far more recently, I saw a video of a Gazan woman running under fire to rescue someone. That’s courage. Sitting on your ass guarding an open-air prison makes you fascist, a cowardly fascist.
Lastly, I donate to Gaza Municipal Works, after you recommended that awhile ago, also EI, Palestine Action, and to some Gazans I know. I’d pay for your work and trust you to help Gaza. Unless anything changes, I will continue to donate to the above places when I can. Pardon my brief tangent but my in-laws are still living in a U.S. imperial colony. They’re struggling and yet they ask for nothing except that I help Gaza. That is such a different culture than the one that’s trying to dominate the planet
This was so well-written. Thank you so much for everything you do!
Great piece. Have been having the same thoughts myself, noticing when people use metaphors like the "genocidal machine". Have also come across interviews with IDF re if the AI selects targets then this removes us from criminal responsibility. I feel this also comes under the umbrella of dehumanisation - not only is dehumanisation used as an offensive weapon, it is also used as a defensive strategy - it was the AI, not us human psychopaths....
And it was done orderly and professionally - in a hyper-modern package (that could soon be yours).
As in "we know of course that all those enemy people are terrorists - but we have policing procedures in place to confirm circumstantial evidence that one of them may indeed be a terrorist before each attack".
Since there are so many of them alas we had no choice but to automate the system of target selection.
We even get to wring our hands about how problematic the automated system is that we pretend to have put in charge - this precisely shows our humanity.
(If you too happen to preside over an unruly indigenous population, we will give you good price).
Nice work here. A fevered justification for a shitty prosopopoeia
Thanks Yardin and Justin for this very thorough deconstruction of the "tech-washing" of good old fashioned war crimes and genocide.
Even the very designation "artificial *intelligence*" in the 1950s was a brilliant PR move by Marvin Minsky aimed at getting military funding. Dual meaning: intelligence as in thinking (humans or machines) - as well as in military intelligence...
[E.g. imagine you could have a machine translated copy of the Soviet Pravda each morning without needing those pesky translators whose loyalties are questionable! Here Minksy was off by only 70 years; he also said that "within 20 years robots won't even keep us as pets". In the 1960s. And again in the 1980s.]
---
Let me plug Yasha Levine's excellent book "Surveillance Valley" that you are using as a reference. He actually documents what everyone always suspected but few cared to muckrake: that those supposedly carefree west-coast information-wants-to-be-free internet start-ups came directly from within the military industrial complex and never left its shadow at all. Exactly as the Boeing civilian airplanes were based on the designs of their military bombers etc.
My favorite example is that Google's search engine is literally the commercialization of a curation tool developed by the alphabet soup agencies based on machine mapping of internal connectivity of archives. You see, their archives were overflowing beyond the abilities of human beings to keep up - so they had to solve that problem before anyone else. As you lay out, guess who they compiled all this material on?
---
It is remarkable that the current hype in business and policy circles is mostly an uncritical rehash of the original AI hype cycle from the 1960s and 1980s. Already basic introductions to AI present the historical development of the field as a succession of hype cycles with a focus on the well-documented limitations of each system: from early neural nets to symbolic AI (expert systems) back to more generalized machine learning - of which the current crop of LLMs etc. are merely a high-powered instance.
[A pity, really, since machine learning is an interesting field with a lot of recent progress. Understanding the limits of error-guided, reward-guided, and unsupervised learning actually deepens one's appreciation of the sweet mysteries of biological "intelligence".]
The (over-)promises of "AI" never materialize - under the hood the current iteration is a big statistics engine which can be set to work moderately well at specific circumscribed problems (such as "targeted" advertising).
I referenced the work of Shir Hever in Justin's earlier discussion of the 972 article. In a nutshell, Shir suggests that the "AI" target generation machine is designed to game the target evaluation procedures of Israeli military intelligence personnel. Effectively squaring the circle of having a free fire zone without breakdown of military discipline.
As Yardin and Justin say, the PR is merely the glazing on the cake, and we need to look past it.
They used to wax lyrical about how the internet is inherently a force for democracy and freedom. Now they're happy to fantasize about the robot* apocalypse (a threadbare con that follows the template of "Baby have sex with me but I have to warn you, I'm such a good lover you just might die of pleasure overload").
I guess they may as well try to tech-wash genocide.
* As a personal anecdote, I remember a frank discussion in the early 2000s among robotics researchers from the German Max Planck Society who the Pentagon offered substantial funding. Eventually they decided they would have to decline on moral grounds - but only if there was any chance they could get their technology to actually work within their lifetimes. Since there conveniently was not, consensus emerged to accept the funding.
They had no answer to what would happen if someone would make decisions about war based on hype - their expectation of having walking killing machines at their disposal irrespective of the actual state of their implementation.
Thanks for this excellent article, the kind of superb crap detection we need.
Guns don't kill people, bullets kill people
This mocking bird’s call is for the AIPAC-MIC
AI SHMMM-AI....
I believe that the guidance system for all the deadly ordinance thrown at the beautiful people of Palestine 🇵🇸....(The most bombed out portion of God’s magnificent creation- never has there been this amount of munitions “hard rained” ever on a people)... is nothing more than the wicked self ignorant egotistical evil mind.
At hands of the phony nation state void of God, the Palestinians have now and always been exposed to the Evil Minded Guidance System...developed in the darkest recesses of it’s greedy founders. ✌🏼🇵🇸🇱🇧🇸🇾🇮🇷🇮🇶🇾🇪❤️🙏🕉
Would like to see you on Bad Hasbara taking apart this terrible AI Hasbara and other lies!
The Fed and the all mighty dollar are doing their bit to fuel the powers behind it.