Top Israeli spy chief exposes his true identity in online security lapse | Hacker News Reader

Top Israeli spy chief exposes his true identity in online security lapse

theguardian.com

284 pointsphilk10259 comments

> The embarrassing security lapse is linked to a book he published on Amazon, which left a digital trail to a private Google account created in his name, along with his unique ID and links to the account’s maps and calendar profiles.

I'm interested to know more about the nature of the lapse here. Is there a bug in Amazon where self-publishers expose their account email addresses? Or was it more like, he shared Google Map and Google Calendar content in the book, not realizing that those features exposed his real email?

From that line and a few others in the article Amazon doesn't seem to be involved at all. It seems like there was a Gmail account created to receive feedback, corrections, reviews, and media outreach that was also supposed to be relatively anonymous but was registered with his actual name which would be visible to anyone that he responded to at the very least, but may also show up through other account associations.

The article also said he created that email specifically for the book to deal with queries. So it's possible there's more to it.

> Contacted by the Guardian, an IDF spokesperson said the email address was not Sariel’s personal one, but “dedicated specifically for issues to do with the book itself”.

I figured it wasn’t “small nation state bug discovered in supply chain” but rather “yea, he used his real name”.

It's always a small detail that gets you... You only have to fail once for your OPSEC to crumble into nothing.

> Email Id with unique id leaked

What's the unique id here ? SSN number / military id,...?

His name

The book is published under a pseudonym made up or his rank and initials and the blurb on Amazon reads:

>[pseudonym], an expert analyst, technology director, commander of an elite intelligence unit, and winner of the prestigious Israel Defense Prize for his artificial intelligence based anti-terrorism project

The defense prize projects & unit conducting the projects are published in The Times of Israel. The person's rank & name is publicly known from other historic news articles. With that context, the google account and Amazon weren't really important to linking the General to the unit in question.

Where did you find that this person even exists ? If the email is yossi.sariel@gmail.com then you still need some kind IDF directory that lists his name, along with his current position. A name alone doesn't mean shit. I think ?

Israel is a very small country. It has the population and size of the Bay Area.

Everyone is basically a 2nd or 3rd level connection.

It has it's pros (eg. VC and Entrepreneurship is much easier because the barrier to entry is lower) but also it's cons (gossip flows very quickly, and it's fairly easy to unmask someone).

This is also why the IDF is so ferocious in Gaza - everyone in Israel and the Diaspora either knows someone or is related to someone who died either on 10/7 or during the deployment (eg. I'm from the west coast yet an elementary school classmate of mine died in the Nova massacre, and a couple former coworkers were mobilized to Gaza because they had an infantry MOS back in the day) so there is no appetite for reconciliation.

Also, most of 8200 basically leaves and joins the private sector after a couple years, and the younger guys (post-2010) seem to have sloppy opsec compared to the older ones, for example publicly listing that they are ex-8200 on LinkedIn or social media.

The IDF, Shabak, Aman, and Mossad have all deteriorated severely since the 2000s, because the best and the brightest in Israel now have private sector options that pay way more and give you way more fame, instead of working as a relatively underpaid bureaucrat in a country just as expensive as the Bay Area.

A similar trend is happening in Singapore as well, as almost all my A*STAR friends and classmates broke bond and naturalized in the US instead of returning. Only those whose families owned a house or a car returned (iykyk).

> The IDF, Shabak, Aman, and Mossad have all deteriorated severely since the 2000s

Absolutely true. Never forget that the vicious attacks by Hamas were never leaked to the Israelis. The supposed spookiest spooks in the whole wide world didn't know about an attack being planned within their own borders.

Actually, it's even worse than that.

The lower levels of the Intelligence Services had a vague idea an attack like this would happen [0] but dismissed it as unrealistic, and Egypt even warned Israel 3 days before the attack [1] but most likely it was ignored/procrastinated due to the Judical Protests and Yom Kippur holiday season.

There was a systemic failure that is very unsurprising due to the relative lack of long term institutional knowledge as Intelligence Community and Policy members left to work for American MNCs like Microsoft or Google, or start their own massive cybersecurity companies like Palo Alto Networks, SentinelOne, Wiz, etc.

Before the economic reforms in the 1990s-2000s, the only options for the best and brightest in Israel was to become a careerist in the Israeli Government or emigrate to the US to found companies like PANW. This meant a large subset of Israel's top talent remained in the government, but all that fell apart due to economic liberalization because people had better choices that paid more and had better work hours.

It's the exact same story in Singapore and South Korea today, and a similar brain drain happened in the former Soviet Union and India in the 1990s.

> The supposed spookiest spooks in the whole wide world didn't know about an attack being planned within their own borders

Imo, the reason Israel's intelligence community was a top player in the 1970s-90s was because most Israelis were 1st-1.5 generation refugees from Arab countries. Mizrahi Jews (especially Yemeni) are fairly overrepresented in careerist Military and Police roles in Israel due to a relative lack of career options, and a number of that generation was L1 Arabic or Farsi speaking.

By the 2000s-2010s, the younger generation was 2nd or 3rd generation and truly "Israeli" so a lot of the cultural nuances in the Arab World fell by the wayside.

[0] - https://www.nytimes.com/2023/11/30/world/middleeast/israel-h...

[1] - https://www.bbc.com/news/world-middle-east-67082047

I used to work for a Bay Area startup whose claim to fame was that it was all 8200 alumni productizing 8200-derived technology.

Yep!

It was Shlomo Kramer (Checkpoint, Imperva, Cato Networks) and Nir Zuk (Palo Alto Networks) that started the 8200-to-entrepreneurship pipeline in the 1990s-2000s.

Shlomo was basically the primary reason Israel does so well in cybersecurity entrepreneurship.

There was some overlap with the IIT alumni network as well due to Rajiv Batra (cofounder with Nir at PANW) that caused the entire Enterprise SaaS space to be lead by Israeli-American and Indian-American founders.

Might not be him, but there's a facebook profile with that name that happens to like an Israeli-built "publish book to Amazon tool", the Israeli Ministry of Defense, Palo Alto Networks, etc.

I see no lapse here.

Some people publish under a pseudonym to disguise their true identity, others to separate their public persona from their private one. This is pretty clearly the later case.

He published the book under his initials in combination with a hint to his job. That he did not take a serious effort to hide his email is just consequential.

This is as exciting as had they reported they'd found out the real name of JJ Lehto.

EDIT: Thinking about it, it is simultaneously a nom de guerre and a nom de plume.

finally, I was waiting for this comment.

Opsec is hard.

Also. No matter how bullish you are on the potential of AI, it's horrifying and categorically unethical to use them to decide who to kill. It creates the possibility of atrocity with no accountability.

Although the fact that this guy was anonymous to start with definitely indicates that accountability isn't the goal here.

The Guardian reported more on the IDF’s use of AI a couple days back, and it’s grim:

https://www.theguardian.com/world/2024/apr/03/israel-gaza-ai...

Although in that article, most of the horrors are coming from the human operators - there’s what comes off as a pretty strong disinterest in finding reasons to distrust the machine.

There must be a line somewhere between "pretty strong disinterest in finding reasons to distrust the machine" and "using the machine as plausible deniability for doing what you hoped to do anyway"; is there a relatively simple description of this line, or is it a fractal boundary?

Yeah, I phrased that as generously as possible because I didn’t really want to kick that particular hornets’ nest, but the article does a fine job making that point.

Very charitable of you! (do we need an emoticon for charity, like we have for sarcasm? symmetry would suggest yes.)

Upon further reflection, it occurs to me that these situations are exactly why the Geneva Conventions[0] require combatants to have (a) uniforms, and (b) a command structure:

(a) prevents someone low from disingenuously stopping the buck by claiming "tweren't no orders; I'm acting on my own"

(b) prevents someone high from arguing as Uber does: "you see, we're not a traditional army that gives orders to our fighters — we just provide a service that matches attackers and targets online, and really, anything bad that allegedly happened was only between those two parties"[1]

[0] if we're not doing these anymore, just go ahead and tell me "the future is now, old man", and I'll take my mutterings to the pétanque piste.

[1] > "Once the rockets are up, who cares where they come down? That's not my department," says Wernher von Braun —TAL

Tinder for combat: swipe to kill.

Morality-washing was always going to be one of the first uses of AI, in hiring and in policing and in war (or ethnic cleansing).

> Morality-washing was always going to be one of the first uses of AI

They even made a cruel joke by naming the automated killing system "Lavender". Lavender flowers were used in clothes washing since ancient times. The name of the flower cames from the Latin word for washing. Portuguese inherited it, for example, with the word for lavender, "lavanda", sounding similar to the verb to wash, "lavar".

Using one of the most biased and anti-israel as a source what could go wrong? Edit: A different and less biased title would be: the use of technology and AI to minimize civilian casualties in guerrilla warfare. No? How much civilians did US kill during Vietnam war? There was no AI then and the Vietnamese didnt murder and rape civilian Americans did they?Edit 2: downvote me until it gets removed and you enjoy your echo chamber of beliefs.

>>Using one of the most biased and anti-israel as a source what could go wrong?

Which one exactly? The guardian article quotes literally hundreds of sources - which ones are you bothered by?

>>the use of technology and AI to minimize civilian casualties in guerrilla warfare.

It achieves literally no such thing, it's dishonest to suggest otherwise.

>>How much civilians did US kill during Vietnam war?

Peak whataboutism. As if everyone on the internet is American or believes that we should be comparing everyone else to America.

> the use of technology and AI to minimize civilian casualties in guerrilla warfare

The article makes it quite clear this is not what’s happening and not why the systems are in use. They’re being used to generate a target list faster than could be done by traditional methods, and are being used despite agreeing with traditional methods only 90% of the time. They’re being used to accelerate the war, not make it more accurate.

> use of technology and AI to minimize civilian casualties

In the year 2022 Russia invaded ukraine, flattened two (?) cities, showed no concern for civilian life, committed war crimes.

And yet Israel has killed more children in 6 months than Russia did in 3 years.

Russia killed 4 western aid workers in 2022, Israel already more than that in 1 week, we are suppose to believe these are accidents?

I think those differences are primarily due to Gaza being urban guerilla warfare with Hamas not wearing uniforms and Ukraine being conventional warfare with frontlines.

Look at the battles of Grozny for a more apt comparison of Russian urban warfare.

> Edit: A different and less biased title would be: the use of technology and AI to minimize civilian casualties in guerrilla warfare.

In what universe is that a less biased title? It is just biased in a different direction.

The original(?) +972 article was also on HN: https://news.ycombinator.com/item?id=39918245

It was an excellent and detailed piece on the various systems, and their use by Israel in the Gaza conflict.

>> In an unprecedented move, according to two of the sources, the army also decided during the first weeks of the war that, for every junior Hamas operative that Lavender marked, it was permissible to kill up to 15 or 20 civilians; in the past, the military did not authorize any “collateral damage” during assassinations of low-ranking militants.

Needs no comment and basically sums up the war.

The names of the systems are “The Gospel” and “Lavender”! How is it that a Jewish army is using Christian references for their AI killing systems? I don’t think that is unintentional.

is "Lavender" a Christian reference?

Lavender's history has biblical roots. It is referred to as Spikenard in the Bible. Mary used it on the infant Jesus and anointed Jesus after the crucifixion when she was preparing him for burial. So it has a "life and death" connotation which makes a lot of sense when you think of the system as a roulette for those chosen to die.

Exactly. There is no way that whomever named these systems didn’t do it intentionally. The question is why? what was the motivation?

It's a form of ideological hand-washing, they don't want to link the killing to their own religion but it's ok to link the other well known religion. Coincidentally hand washing refers to the story in St. Matthew’s gospel, when Pontius Pilate ‘washed his hands before the multitude, saying, I am innocent of the blood of this just person’.

Sources? What Christ was offered in his birth was myrrh (by one of the Magi) and his body was wrapped in myrrh after his death (https://en.wikipedia.org/wiki/Myrrh#In_the_New_Testament). Myrrh and Lavender seem to be very different. Is there some popular association of the two (like yams with sweet potatoes)?

A quick google search for “lavender christianity” shows many many people have associated them together, possibly with some historical or primary source inconsistencies, however it’s very obvious connected. In addition to search engines a conversation with any LLM has them immediately explain the connection.

Now you might be correct there is some mis-associating from a purely academic standpoint and that actually makes it even more interesting the IDF used it as a name, considering the other name is “The Gospel”. Isn’t the message to Christians - “you’re idiots and we kill our enemies using AI systems named to make fun of you”?

I was raised Greek Orthodox and that's not a message I receive at all. I commented as above because for me there is no association between Christianity and lavender. Asking others who were also raised Greek Orthodox, lavender has no special meaning for any of us at all. To clarify, that's not from an academic standpoint, but from the point of view of tradition.

Perhaps it's a Catholic or Protestant thing, but, if so, I don't know where it comes from. I used to love reading the New Testament and I really don't remember any mention of lavender at all. It could be an Old Testament thing though, so, to be fair, more Jewish than Christian. In Greek Orthodox practice the aromatic of choice is frankincense (every church smells of it because the priests burn it in long-chained incense burners that they swing wildly left and right during liturgy), and, separately, myrrh (although that's associated with death and funerals, I guess). Lavender we mainly use to keep moths off clothes, to be honest. I got small muslin-wrapped packets of it in my closet.

I also am not sure why Israelis would want to send a message to Christians about an AI system used to kill mainly Muslims. I mean, in practice, the Israelis enemies are Arabs who tend to be Muslim, so why would they send a message to Christians?

In any case, can you please share some of the information you found in your quick googling? I'm curious to see what people say. I had a quick look online also and couldn't really find anything specific. The wikipedia page on lavender also lists no specific relation with Christianity, as opposed e.g. for myrrh and frankincense.

There is definitely a large groups of Catholics and Protestants who are associating Lavender with Virgin Mary, Jesus’s birth, and others who believe it goes back to the garden of Eden and Adam and Eve.

One hypothesis is a confusing Spikenard, here is what GPT4 says about that:

> Yes, there can be some confusion between spikenard and lavender, partly due to historical naming conventions and translations of ancient texts. The confusion often stems from the use of the term "nard" or "spikenard" in historical texts, including the Bible, and how it has been interpreted or translated over the centuries. > Spikenard (Nardostachys jatamansi) and lavender (Lavandula angustifolia) are indeed different plants, both botanically and in terms of their historical and cultural significance. However, the confusion might arise because both were used in ancient times for their aromatic properties and were considered valuable for perfumes, medicinal purposes, and religious rituals.The term "nard" comes from the Sanskrit word "narada," while "spikenard" refers specifically to Nardostachys jatamansi. The translation and interpretation of ancient texts, like the Bible, have sometimes led to a blending of these plant identities, especially when the specific botanical knowledge of the original texts was not fully understood or when names were translated in ways that did not precisely match the botanical realities.Additionally, the historical trade of these substances, along with others, through ancient markets might have contributed to the blending of their identities. Merchants and consumers across different cultures might not have always distinguished clearly between the two, especially given the value placed on aromatic herbs and oils in ancient times for both practical and symbolic uses.Despite these confusions, modern botanical and historical scholarship distinguishes clearly between spikenard and lavender, recognizing their unique characteristics and the distinct roles they have played in history and tradition.

As far as the message from IDF, my understanding is the true believers of Orthodox Judaism are actually both anti-muslim and anti-christian. The hidden message I take is “you Christians are fools, easily tricked, we have tricked you also, we laugh at you behind your back, and desecrate your beliefs as our weapons of war”.

Do you have a source that corroborates GPT4? It is well-known that it can happily generate unsubstantiated information.

>> As far as the message from IDF, my understanding is the true believers of Orthodox Judaism are actually both anti-muslim and anti-christian. The hidden message I take is “you Christians are fools, easily tricked, we have tricked you also, we laugh at you behind your back, and desecrate your beliefs as our weapons of war”.

Isn't that just a little bit too much to infer from a single word? If they wanted to send a message to Christians, why not choose something obvious like "Cross" or "Golgotha" or "Herod"?

“The Gospel” is pretty damn obvious.

I don’t think it’s too much to infer, for me it’s extremely obvious, especially considering the religious history.

People don’t realize that Orthodox Jews literally spit at Christian worshipers in Israel.

We have to admit what is going on here. The IDF built multiple AI killing machines with US and Western money (Christian money) and then named them intentionally offensively and with direct hostility to Christianity. There is absolutely no possibility that the people who selected the name “The Gospel” and “Lavender” were not doing so in a religious context.

From a religious standpoint Islam and Christianity are incredibly close. Orthodox Jews only view Christians as useful idiots in their agenda not as religious allies, quite the opposite.

Regarding Lavender - copilot with references below. However I think you are likely correct that the accuracy of the connection is potentially flawed, however it’s the beliefs that matter, and a large number of Christians believe there is a direct connection.

Lavender has *profound connections* to divine symbolism within *Christianity*. Here are some references for you to review:

1. *Biblical Meaning of Lavender*: - Lavender's spiritual essence is depicted in various contexts within biblical scripture. - It is considered a powerful tool for *spiritual purification* in Christianity. - The scent of lavender uplifts the spirit, fosters inner peace, and aids in establishing a connection with the Divine¹.

2. *Folklore and Symbolism*: - When a woman washed Jesus' feet, the lotion she applied contained lavender. - Mary, Jesus' mother, hung his swaddling clothes on a lavender bush, transferring his scent to the plant. - As a result, lavender came to represent *cleanliness* and *purity* in Christian symbolism³.

3. *Spiritual Benefits*: - Lavender is associated with calmness, purification, and tranquility. - It is closely linked with the *crown chakra*, which connects to spiritual realms. - Smelling lavender during meditation and yoga can purify and cleanse the air, aiding in spiritual practices⁵.

Feel free to explore these references further to deepen your understanding of the connection between lavender and Christianity!

Source: Conversation with Bing, 4/5/2024 (1) Biblical Meaning of Lavender - Divine Symbolism. https://biblewithus.com/biblical-meaning-of-lavender/. (2) Lavender Folklore: The Tales Behind This Calming Purple Plant. https://www.icysedgwick.com/lavender-folklore/. (3) The Spiritual Benefits and Meaning of Lavender - Original Botanica. https://originalbotanica.com/blog/spiritual-benefits-lavende.... (4) The Spiritual Meaning of Lavender - TheReadingTub. https://thereadingtub.com/lavender-meaning-spiritual/. (5) History’s Love of Lavender: From Mummies to Bathhouses and Beyond!. https://www.ancient-origins.net/history-ancient-traditions/l....

Christianity: • Jesus ascended into Heaven • Jesus is the Son of God, the Messiah, & the final prophet • Jesus’ soul & body are in heaven • Jesus will return to destroy Satan and establish peace on earth • Jesus came from a virgin birth • Jesus performed miracles

Islam: • Jesus ascended into Heaven • Jesus is the word of God, and his messenger and a deliverer of the scripture • Jesus’ soul & body are in heaven • Jesus will return with Imam Mahdi to destroy a False Messiah and establish peace on earth • Jesus came from a virgin birth • Jesus performed miracles

Judaism: • Jesus did not ascend • Jesus was the most damaging false Messiah • Jesus is in boiling excrement in Hell. • Jesus will not return • Jesus had a normal birth • His miracles were of the devil

Those don't seem like references that have anything to do with Christian tradition and practice, certainly not as I understand it from my background. For example, meditation is not a Christian concept.

I would kindly suggest that if you want to understand the meaning of Christian symbolism you should rely on the New Testament and possibly the Old Testament also. You don't need internet searches and chatbots for that.

Or, of course, you could chat with a Christian.

Specifically for this:

>> However I think you are likely correct that the accuracy of the connection is potentially flawed, however it’s the beliefs that matter, and a large number of Christians believe there is a direct connection.

That is what needs a source and I can't find one in what you have provided. Who says that "a large number of Christians believe..." etc? In my experience, nobody I know who is Christian believes that.

> That is what needs a source and I can't find one in what you have provided. Who says that "a large number of Christians believe..." etc? In my experience, nobody I know who is Christian believes that.

I’m in the US and every Catholic I’ve asked associates Lavender with faith. Several of the references I gave are directly Christian themed sources.

We know Christianity is very diverse and has numerous interpretations. I think it’s possible in more Orthodox leaning cultures that Lavender is substituted.

Regarding Old and New Testament, I certainly agree those are the primary and authoritative sources and as I referenced earlier the translations are in dispute and likely source of this confusion. Different traditions appear to have differing translations of the underlying plants being mentioned.

Well that's a good point. I guess I'll ask some Catholics and Protestants about it. It's true that there may have been mistranslations of the Gospels over the years. Come to think of it, that would explain a whole lot.

that may be so, but lavender being a Christian reference seems a bit of a stretch to me. it may well be referred to in the bible, but so are thousands of other things. to me it sounds more like they're giving them light and positive names that bely their destruction in a dark and comical manner

"The Gospel" has wider connotations

https://en.wikipedia.org/wiki/The_gospel

probably from the Greek

https://en.wiktionary.org/wiki/εὐαγγέλιον

That’s a stretch considering any Jewish person I’ve ever known would immediately associate “The Gospel” with Christianity and what officially their religion considers as a blasphemy and heresy. There is absolutely no possible way that whomever named the system wasn’t consciously linking it and invoking the religious connotation.

Correct. The Jewish people think of the Gospel as "that which the masses blindly accept as true but we know internally to be false" which is apt given what they have intended to use it for.

Which perhaps they even expounded on by using “Lavender” which potentially is something Christians may have mis-associated with Jesus’s birth (see parallel comments on Lavender). The message to Christian’s is clear - “thanks for all the funding, you’re a bunch of fools, and we are killing our enemies in your name”

I believe the laws of war as commonly understood authorise civilian casualties insofar as they are "proportionate". 1:20 is technically a proportion, but...

> the laws of war as commonly understood

There’s a whole lot of wiggle room in that phrase.

1:2000000 is also a proportion.

Btw, the proportionality requirement in the law of armed conflict relates civilian casualties not with military casualties but rather the military advantage gained. The military advantages of killing a single low-level Hamas combatant are rather limited, and thus the room for civilian casualties should be rather limited as well.

Proportionate in war is not about going tit-for-tat.

> The principle of proportionality prohibits attacks against military objectives which are “expected to cause incidental loss of civilian life, injury to civilians, damage to civilian objects, or a combination thereof, which would be excessive in relation to the concrete and direct military advantage anticipated”.

The way it's worded is to prevent destroying civilian targets for no military gain.

Thats a very... charitable reading.

20:1 is a proportionality I'd find horrifying. 1:20 is unspeakable.

There is no justification in the current situation in Gaza to justify anything close to the numbers quoted.

More of a slaughter than a war at this point.

"at this point" being the last 6 months, I'd assume?

Yes.

There's one quote in that article that continues to make my blood boil:

> "You don’t want to waste expensive bombs on unimportant people"

This, regarding the bombing houses full of civilians, because an infantryman might have been inside. This, in a statement by somebody who is critical of the AI effort. By being more concerned at the loss of munitions; the true basis of the calculation is revealed. The IDF -- even those members who are actively critical of it -- views its enemies not as humans, but as vermin to be exterminated. This is genocide, pure and simple.

None of what you said is the definition of a genocide.

”Machines never make the decision, they’re merely collecting data to assist human operatives”

They’ll keep saying this forever, and there’s a shroud of plausible deniability to it! Not to mention it can soothe the conscience of the operatives who have to live with their decisions - they don’t even have to look through the family photos - the AI has already filtered out the suspicious parts. In reality, it’s a good ole recommendation engine, just like Netflix or TikTok. And when you’re tired, have too much work stacked up, you’ll be more likely to let autoplay take over – or approve the strike of someone’s home.

To those that think this is harsh: please read the article. It’s that bad.

It's not a matter of being tired and having to catch up. The operators are explicitly instructed to treat the AI results as orders without questioning the results.

In other words, operators are threatened with punishment if they take the time to inspect the results more closely before following orders. It's not even an option!

> In order to speed up target elimination, *soldiers have been ordered to treat Lavender-generated targets as an order, rather than something to be independently checked*, the investigation found. Soldiers in charge of manually checking targets spent only seconds on each person, sources said, merely to make sure the target was a man. Children are also considered legitimate targets by Lavender.

Right! The extreme point of laziness/stress is just pressing approve. At which point the machine is making the de-facto decision.

In either case, the role of the recommendation engine is immensely impactful, as we already know from consumer products. But here the software engineers are directly involved in life-and-death decisions at scale. I really hope they know that.

They know that because it was designed to increase civilian casualties by unprecedented civilian:"combatant" ratios and confidence levels. These were design requirements, not faults.

The notable element of this news is not that the decision-making was automated. It's that they set the civilian collateral death ratio to ~100 (and other details such as considering children to be valid combatant targets), regardless of whether the process that arrives at that was automated or not.

The Guardian article on this had quotes from the "operators" involved. They don't question or examine any of the output just hit approve.

The only apparent operator act that seems to be something they are involved in is turning the dial up or down to get more or less "hits". Depressingly this appears to be related to superiors shouting at them as they don't have enough targets.

It's absolutely sickening.

When you're a tiny, easily overtaken (geographically) country with genocidal neighbours and your race and religion is a global minority, of course you're going to be bullish on taking out threats to it, and your people, and less caring about collateral damage. Your nation is under threat from almost all sides

Being less concerned about collateral damage and other similar concerns in that condition will perpetually increase that threat.

Not when you have fire and technical superiority.

Fully autonomous drones, from other countries, have been killing people since at least 2020. It's not just the IDF using AI.

This was the earliest official report sourced by the UN, pegging Turkey: https://www.fox5ny.com/news/fully-autonomous-drone-was-used-...

The US has killed plenty of non comabatants using human guided drones for decades...

In question are fully autonomous, artificial intelligence-based, unmanned drones

Can't make them too artificially intelligent or they might draw conclusions: https://news.ycombinator.com/item?id=38362711

The link is, indirectly to Peter Watts' story "Malak", which, yeah, is exactly what I think of whenever people talk about this.

>It creates the possibility of atrocity with no accountability.

As gross and unfortunate as it is, this is a selling feature to some governments.

Hey at least Palantir says their death machine is ethical because it shows you a flowchart of a bunch of widgets. Do other weapons have widgets?!? I didn't think so!

Do you think this holds up in court? It gives me Nuremberg defense vibes ("I was just following orders"), except humans are following an AI's orders.

That's why you work to remove humans from the loop altogether.

Man, Fuck Ted Faro.

That's terrifying, I hate where this is all headed. I still hope that the people that ordered it setup and those that had it developed would be held accountable (I know, pipe dream).

...I'm having trouble discerning whether you understand the abject stupidity of that suggestion, or if you're just being sarcastic. Poe's law strikes again.

To be honest, I could see this happening. What's going on right now is absurd enough as it is. I pity The Onion.

More sardonic or grimly ironic than sarcastic, but yes.

That probably isn't enough - representatives from IG Farben, Krupp and Flick were all tried at Nuremberg for war crimes because their companies helped war crimes and crimes against peace; therefore the people behind the AI targeting civilians are in theory liable for it murdering civilians.

> It gives me Nuremberg defense

You only need a Nuremberg Defense if you lose, and your country is taken over.

If you are an American soldier, only the US will be able to hold you accountable.

Israel has nukes and the backing of America. I think it is unlikely that they would ever lose so badly as their soldiers face a Nuremberg like trial.

Rules for thee but not for me (or any of my best "friends").

If they lose that badly, I can't imagine anyone around them would bother with trials.

Yea now you know why they're fighting so hard. Their genocidal neighbours want them dead regardless of any of this stuff.

> It creates the possibility of atrocity with no accountability.

No it doesn't. If you use any tool to kill people (ie a gun, bomb, computer) the person (or people) who used the tool are accountable.

What it does is create more levels of indirection and detachment.

Indeed, they gave each hit 20 seconds - if that - to manually verify. That said, airstrikes, drones etc already add a layer of indirection and detachment. The thing that this "AI" tool allowed was to mark targets faster, or if you will, with less scrutiny, increasing the efficiency and speed of the killing machine.

We've been creating indirection ever since the first homo ancestor carved/fashioned a spear tip or arrow, or the first metallurgist poured a bronze sword. But.

There's matters of degrees. And this is just one or two steps away from full autonomous murder.

Reminds me of this section Frank Herbert's God Emperor of Dune:

... She would carry with her forever afterward the clear sights and sounds and smells. The seeking machines would be there, the smell of blood and entrails, the cowering humans in their burrows, aware only that they could not escape . . . while all the time the mechanical movement approached, nearer and nearer and nearer ...louder...louder!

    ...like the terrible machines of that apocalyptic vision, the predator could follow any creature who left tracks.*

Indirectly, the voters who put a warmongering leader into office are accountable…but when a war goes wrong, they never blame themselves.

Cynically, it’s not who is actually responsible and by how much, it’s who is forced to take the blame.

when a war goes wrong, they never blame themselves

Arguably the German nation/culture has done this. At least in large part. Not sure any cultural repentance would ever be enough though.

I also feel like there is in Canada at least a growing understanding of the horrors of colonization and attempted genocide and assimilation against First Nations here. Whether that ultimately leads to reconciliation and a better path forward is hard to say.

Well, there’s war, and then there are wars which are driven by popular cultural and social ideologies. Desert Storm and WW1 are pretty different than WW2 that way.

People who use the best tool (i.e. other people) to kill are rarely held accountable.

> it's horrifying and categorically unethical to use them to decide who to kill.

Absolute statements are hard. I mean, I agree with the sentiment, but I can't get behind such a strong statement.

If you have a 10 person group deciding who to kill, and reduce their workload with AI so that the 10 person group can take more time double checking-- that's "using AI to decide who to kill" but probably improves outcomes.

The problem is, in practice, we know that people are just going to punt responsibility to the machine and shrug.

The fundamental reality is that compressing decision timelines creates more effective military outcomes.

Everything else flows from that.

At a high-level, that decision compression will be an arms race between competitors for military advantage, where the argument to relax human-scale decisions will be "If we don't do it, we'll lose to someone who will."

You saw the same thing in the nuclear arms race, as launch-to-impact time compression (as technology advanced) eroded available human decision time.

> The fundamental reality is that compressing decision timelines creates more effective military outcomes.

But popping off half-cocked often results in bad outcomes, so there's a tradeoff between decision quality and decision time.

In theory, having AI in the loop could improve both. In practice, I suspect you're right.

Unfortunately, the negative impact of killing civilians currently seems to be... minimal.

And maybe always has been, internationally-speaking?

It would be nice to reform UN and supra-national rules of warfare in light of modern developments.

Even a "this is an acceptable ratio of civilian collateral deaths, above which constitutes a war crime" would be helpful for curtailing nation's baser urges.

> And maybe always has been, internationally-speaking?

If I had to guess, it was even less. Now at least there can be media backlash.

> Even a "this is an acceptable ratio of civilian collateral deaths, above which constitutes a war crime" would be helpful for curtailing nation's baser urges.

Having some kind of metric would be nice. However, there are problems with it being too simple (your enemy has a say in how exposed their civilians are; and we'd hate to leave countries with less precision in their militaries without the ability to defend themselves).

> reduce their workload

If killing the right person is too much work then you shouldn’t be killing

We will find the "evidence" to extract our pound of flesh. It was Curveball then, it happens to be "AI" now.

US decided not to bomb Kyoto in WW2 because Henry Kissinger had visited it and was enthralled by its beauty. Somehow such decisions are ok because made by a human?

https://www.bbc.com/news/world-asia-33755182

Some historians say US Secretary for War Henry Stimson had a personal reason for sparing Kyoto. (edit : wrong Henry)

Military technology has been helping people decide who to kill for ages. For example, long distance cameras and night vision devices use extensive image processing. They're designed to improve visibility but can undoubtedly lead to false positives in the heat of the battle.

> Also. No matter how bullish you are on the potential of AI, it's horrifying and categorically unethical to use them to decide who to kill. It creates the possibility of atrocity with no accountability.

How much technology is it ethical to use in the decision-making loop for kinetic decisions? I don't think there's an easy answer here that actually engages meaningfully with the question.

He didn't advocate for fully anonymous AI systems, but ones where humans will ultimately make the decision. Obviously these can be implemented in a variety of ways, from basically data mining software and some neat dashboards to help a bunch of humans decide on a target in a responsible way, to something where all the boring bits are hidden, a list of names is oh-so-helpfully drawn up and a rubber-stamp from an under-trained early-20s "intelligence officer" is all that's needed to kill someone, and their family, in their home.

AI is a tool. It can be used to kill more people, and it can be used to kill less people.

Imagine if the superweapon US developed during WWII was AI. They'd use it to deliver devastating destruction to military factories in Nagasaki and Hiroshima, saving 100k+ civilian lives.

>It creates the possibility of atrocity with no accountability.

It most certainly does not. And comments and thinking like this enable that. There is a human who is responsible for deciding to use the AI, that human is responsible. If they can't handle the responsibility, then don't use AI.

its true. Whenever you're in a sensitive position and must practice opsec hygene, you need to be consistent and get it right every time.

Your adversary only needs to get lucky once.

> No matter how bullish you are on the potential of AI, it's horrifying and categorically unethical to use them to decide who to kill.

Is it any more horrifying or unethical to decide without AI?

> It creates the possibility of atrocity with no accountability.

Who was held accountable for hiroshima? Or do you know how much atrocities were committed to create israel? You act like we had accountability before AI. We've never had accountability for atrocities. Just victor's 'justice'.

AI can be better than humans when it comes to atrocities. It can be as bad as humans, But it can never be worse. Humans set the standard.

> Is it any more horrifying or unethical to decide without AI?

It is harder to get 1000 people to participate in an unethical military campaign than it is to get 10 people. So if you can make those 10 people 100x as efficient...

> It is harder to get 1000 people to participate in an unethical military campaign than it is to get 10 people.

No. It's easier to get 1000000 to participate than 1000. But that's not the point. We are talking about accountability. I'd say it's easier to hold 10 people accountable than a 1000.

> Although the fact that this guy was anonymous to start with definitely indicates that accountability isn't the goal here.

Well, it isn't. This is a strategic choice, and one pretty much every military took

That's basically what AI as a concept is. Smoke as screens to hide responsibility and costs.

No it doesn't. That was always a possibility. Atrocity with no accountability long predates AI, and the ethics of "using them to decide who to kill" are isomorphic with the ethics of using a dice roll to decide who to kill.

In this case, the people you could choose to hold accountable is everybody who planned, built, or activated the machine made of a Wacky Waving Inflatable Tube Man mounting a chaingun, designed to spray bullets indiscriminately in all directions. The fact that no human trigger finger is actually on the trigger does not AT ALL defuse ethical considerations, does not exculpate anybody. We kill with machines in every military, but it's human decisions all the way down no matter how long ago a human last physically touched the munition.

The lack of accountability starts to occur when a journalist embraces headlines like "AI Made Decisions About Who To Target", rather than using scarequotes and agency - "IDF-run AI system 'Made Decisions' about who to target". The IDF are the actors in this construction, the AI system is just a tool. The lack of accountability continues in the political sphere where no war crime investigation takes place, no red line is instituted, no Geopolitical Consequences arise. It finalizes in the polity when no liberal / humanist vigilante takes revenge into their own hands, no 4chah/kiwifarms investigation doxes their family, no airline pilot diverts to the Hague, we just sort of embrace treating atrocity-committers humanely because we're lazy cowards.

It doesn't really matter if his identity gets leaked, when you are a director, you always become more visible no matter what you do. You cannot hide if you are a director.

In military matters, high ranking officers don't protect their identity, because they represent the military.

Although it is obviously true that lower and medium ranking soldiers and personnel ALWAYS need to protect their identity, because they should not be identified since they are the ones who carry orders and take the real risks.

Of course it would have been better for him to not have his identity disclosed, since it would expose him to assassins if he would travel, but honestly this is a non-story, this will just generate hate comments in anti-israel crowds.

Your last paragraph is a weird take. It's a story because he was ostensibly trying to hide his identity and failed pretty miserably.

> Published in 2021 using a pen name composed of his initials, Brigadier General YS

Also it doesn't seem like he was REALLY trying to hide his identity

Here's the book:

https://www.amazon.com/Human-Machine-Team-Artificial-Intelli...

Author is listed as "Brigadier General Y.S" - hardly difficult to guess Yossi Sariel?

The 'About the Author' doesnt really try to hold info back either

> Brigadier General Y.S is an expert analyst, commander of an elite intelligence unit and technology manager with 20 years of experience working on national security issues and challenges threatening the State of Israel. His areas of expertise include cyber and data science. In 2018, he was awarded the prestigious Israel Defense Prize for a new artificial intelligence based anti-terrorism project. He wrote his most recent book, The Human Machine Team, during his year of studying for a Master’s Degree at the National Defense University in Washington, DC. Y.S also has written other books in Hebrew on the topics of intelligence, strategy, and Hezbollah. He holds a BA in Middle East Studies and another BA in Psychology and Sociology. He grew up in Haifa and currently lives with his family in the beautiful Galilee region of Israel.

I believe the revelation is that this particular Brigadier General is leading specifically unit 8200. I'm sure that the line

> n expert analyst, commander of an elite intelligence unit and technology manager with 20 years of experience working on national security issues and challenges

Fits neatly into quite a few punched-up resumes and 'about' pages.

How many matches do you think there are for "In 2018, he was awarded the prestigious Israel Defense Prize for a new artificial intelligence based anti-terrorism project"?

okay, yeah. Not many. That is patently poor opsec.

Yeah how was this such a secret?

Perhaps whichever HR department in the IDF covers publishing books has a checklist that says "serving members cannot publish books about intelligence operations under their own name", but no one who actually enforces an anonymization policy.

Yeah this is either an opsec blunder or a smokescreen.

Sariel is an anagram for Israel.

Very nice! Sariel is one of the 5 archangels in Judaism. I think (in descending order) they are Michael, Gabriel, Raphael, and Sariel. I don't remember the 5th.

Also, how many brigadier generals could Israel have? It seems like a smokescreen, but... idk idk

Personally, I don't care about his identity, but this article also discusses the contents of the book he published, which is much more interesting.

Brilliant 4D chess move by Yossi Sariel: he's cleverly feigning a mistake, aiming to exit his role without formally resigning. This "mistake" provide his superiors with a pretext for his departure, all while sidestepping the events of October 7.

Given his position, is revealing his identity really that great of a move for an uneventful retirement? The reduction in personal security (given that many people might want revenge on him or extract information) seems like a bad trade.

He already explicitly wanted to publish the book with his actual initials. It would have been quite easy for adversaries with just general public information to get from the initials to the person, right?

So perhaps revealing the full name doesn't make much of a difference, which is why it could just as well have been "accidental" (quotes intented).

But the book was released 3 years ago? He made this mistake on purpose 3 years ago to have some way of exiting his role without having to resign?

But not until very recently did a publication suddenly "discover" this

Ah, but how do you know it was published 3 years ago? Did you buy a copy then, or are you just relying on Amazon telling you so?

i think he was being sarcastic

This is why it's 4d chess. Your linear perception of time is why you're not the Israeli intelligence chief.

Especially since the book was publicly meant to be released using his initials YS in the first place. It's not like he really wanted to keep this a secret.

It seems to me that any blunder can be described as "brilliant 4D chess move" by those insisting that they did not mess up and it is us who do not see the game.

Based on the events that have been taking place from October 7, neither party is capable for any form of strategizing. Let alone "brilliant 4D chess moves".

A characteristic of highly effective leaders is the skill to blunder, laugh authoritatively, and claim it was all part of a bigger plan.

No thats bad leaders with good staying power.

> This "mistake" provide his superiors with a pretext for his departure, all while sidestepping the events of October 7.

I am not convinced that the Mossad, a competent intelligence agency from what I have read, did not actually know what was going to happen. The October 7th attack gave a perfect political justification for wiping out as much of Gaza as possible.

Read the article from Maariv that is linked in the parent article. Quite eye-opening. The plans were known for two years, but the hubris blinded them. Another recent, very painful article is this one - https://www.haaretz.co.il/magazine/2024-04-04/ty-article-mag...

Politely letting you know that it’s spelled quite, not quiet , so that you can learn from this. The English language can be really dumb

Yeah. Many people in the Israeli government have been quite explicit about wanting to wipe out all Palestinians. They're not trying to hide it.

Many people in the Israeli government have been quite explicit about Hamas’s actions being useful to them as a pretext for doing so, too.

So, while I am not arguing that is is the case, it is not at all implausible (especially if the underjudged the scale or likely impact of the attack) that they might allow an unusual attack that was likely to be pretextually useful to occur to provide an excuse to bring the hammer down, even before considering the specific, personal pressures on Netanyahu and how he might view an excuse for a justified war as a means to deflect them.

This theory doesn't make sense when you factor in the political consequence for Netanyahu.

A country's leader is always blamed for a lapse in decent, and Netanyahu knows that.

Maybe he is just left there as a puppet

By...?

I don't know but does it look like he's hugely powerful now?

Yes and no, but either way, he was hugely powerful before.

Pointing out that he doesn't look as powerful now after dominating Israeli politics for decades is supporting my point.

lapse in defense*

Mossad's competence is a complicated topic.

Exhibit A: Yom Kippur war. They had convinced themselves that Egypt would wait on armament deliveries, so their obvious preparations for war, troop movements, and massive exercises next to the border were not a problem. Syria's obvious troop movements and cancelling of leaves weren't a problem either because obviously Syria wouldn't attack without Egypt, and Egypt would obviously wait for more Soviet shipments of armaments. Spoiler: they both attacked, and although defeated later on, demonstrated that Israel wasn't as invincible as everyone had thought, and it resulted in Sinai being returned to Egypt and normalisation of relations between Egypt and Israel (thus long-term a victory for Egypt, even if they had their asses handed to them, militarily).

Exhibit B: Lillehammer affair, where Mossad mistook a random Moroccan waiter for a high level Palestinian leader, and killed him.

Exhibit C: Assassination attempt against high level Hamas leader, Khaled Mashal which was a failure and seriously jeopardised the relations with Jordan.

For the October 7 attacks, we simply don't know if it was a political decision to let the attacks happen, or if hubris convinced Israel/IDF/Mossad that the intelligence can't be true.

Why would he want to look like an idiot when he could just resign and become a civilian?

He could even go full Snowden if he wanted to.

The real 4D chess, is in writing a thought leadership manual for computerized mass murder occurring right now, and being able to extricate yourself from any responsibility by just resigning.

If you looks at the deaths per bomb dropped it shows you the amount of restraint that Israel is exhibiting and that it is not at all a mass murder.

It's just war.

How many countries have been at war with their neighbour for 75 years post-WW2? How many wars have a body count that's tilted 20:1 in favour of the stronger side?

"It's just war" might win you glibness points online. Just be careful not to use it when referencing say, attacks on Warsaw ghettoes in the 1940s. After all, anybody there could have been part of the resistance, could they not?

"Deaths per bomb dropped" was probably somewhere in the Nuremberg defense lawyer's arsenal, but left out because the Allies could just point to the mass starvation that was directly related to bombing. Plus ca change....

Be careful not to pull a muscle with all that reaching.

It's nice to know this kind of stuff is hard for the experts too

> Sariel’s critics, the report said, believe Unit 8200’s prioritisation of “addictive and exciting” technology over more old-fashioned intelligence methods had led to the disaster.

Do they also use Kubernetes and the latest frontend frameworks?

Kubernetes is old and crusty technology at this point. It would be more like firing the senior developers to hire junior ones with a ChatGPT subscription and expecting big improvements in development speed.

Nope, they moved to serverless. They compromise random computers in basements and send small functions to process in a distributed manner. Since the data is so small, it's not possible to construct what they are doing.

SETI@HOME in Soviet Russia fashion. You process their data, even without knowing it.

The only thing you can do is to format your system to get rid of the backdoor (hopefully).

I didn’t know reformatting disabled intelME!

I hoped the person used a firewall and/or removed ME forcefully for the second round.

Deep in the bowels of Mt. Sinai, a clandestine IDF bunker guards a terrible secret: a giant useEffect() loop that controls all the murderbots, individually, by their hardcoded names. They didn't have the budget to research memoization yet, so it runs a bit slow, but it was fine on powerful enough hardware.

The ballistics were all calculated in WebGL, leading to a global GPU shortage the last few years. And sometimes they'd get the viewport size wrong, which is why there was collateral damage :(

Having learned their lesson, Israel's next AI project will stick with the tried and true and be written entirely in CSS instead.

Opsec is hard. 99999 times done everything right. 1 time making a mistake and you're done.

Is it really OpSec if you're putting your initials on a self-published Amazon book?

> An electronic version of the book included an anonymous email address that can easily be traced to Sariel’s name and Google account. Contacted by the Guardian, an IDF spokesperson said the email address was not Sariel’s personal one, but “dedicated specifically for issues to do with the book itself”.

So it doesn't sound like his "real" email was leaked somehow through the book publishing workflow but more like there was a contact email listed (Maybe even in the book itself) in the book and it was not sufficiently private. Could be as easy as leaking some letters or a profile picture through the password reset workflow.

The implications of this is that anyone is easily able to find the name behind a google account?

They used to show the profile picture, not sure what's the current status but there's still a few characters of the email address visible usually.

> The security blunder is likely to place further pressure on Sariel, who is said to “live and breathe” intelligence but whose tenure running the IDF’s elite cyber intelligence division has become mired in controversy.

It seems weird to me that the news outlet that ousted Sariel is foreshadowing that the disclosure could put pressure on him. If they say "we try to oust Sariel," and also "ousting Sariel will put pressure on him", then through the transitive property I interpret it as "we are trying to put pressure on him." And, that seems like politics instead of journalism.

Email account aside, he

- Published a book using his real initials

- Was one of four recipients of an IDF prize in 2018

- Did a Master’s Degree at the National Defense University in Washington when writing the book

This is either a blunder like no other or there is no Yossi Sariel. His last name being an anagram for Israel and the fact there's zero traces online for someone who's been in the IDF his entire life, this seems like the most likely explanation.

EDIT "Sariel" is actually the name of an angel from a Judaic tradition https://en.wikipedia.org/wiki/Sariel

>"One section of the book heralds the concept of an AI-powered “targets machine”"

https://www.youtube.com/watch?v=nSQ5EsbT4cE

https://en.wikipedia.org/wiki/Brazil_(1985_film)

Adjacent for some time to those participating in the secret squirrel pantomime, I find this all so terribly tiresome.

Drawing a veil under the pretense that revealing what's behind would harm the country, the reality is that those participating are engaged in rather mundane drudgery and would simply prefer to not be made available to uninformed speculation by the peanut gallery.

Nice try, book promoters.

"Published in 2021 using a pen name composed of his initials, Brigadier General YS" - Really? Misdirection?

I don't run any secret military units (that I'm willing to admit to), but I do leave occasional product reviews on Amazon under a pen name.

I discovered that I was doxxed by creating a Goodreads account from the same email address, now that Amazon owns Goodreads.

In the 1950s or 1960s, the American journalist Steward Alsop printed the name of the head of the British MI-5 (or maybe MI-6). He told someone that the man (Menzies) was somewhat placated because Alsop had referred to him as "legendary".

US LEO like to say The bad guys only need to slip up once to get found out.

This seems to be true.

If the chief spy of a country doesn't understand the implications of publishing books from his personal account, then you cannot trust him whatsoever... The stupid are controlling the whole thing.

Ego. What's the point of success and mastery if you can't convert it to what you really care about which is status, even if it's lived vicariously through a pen name.

It's like organized crime people always having to buy cars with their money then the tax authority comes knocking. Beyond a certain point money is useless if you can't convert it to status and get those feel good chemicals flowing.

If nothing else you can use money to help people. Money is never useless my friend. Money can be a burden though, like if you have it but can’t spend like someone else who has the same amount but legally I can agree to that.

Ive played at really high levels.

I'd say 90-98% are really smart.

That is the interesting part, a few people arent. They are there for some other reason. I know someone who was inspiring but a dolt, I know someone who had their parents money, I know someone who bandwagon a winning candidate.

Who knows is this person is there on merit or not, but its pretty interesting how smart people get when you start moving up. Its the exceptions that are weird.

I know there are smart people there, the point is that the stupid are really in control. It is the same for most companies: you'll find a lot of smart people (that's why they're successful), but you'll find also a few stupid people who reach the top by other means and make the company a piece of hell for everyone.

Ok but why did his identity need to be secret in the first place? https://en.m.wikipedia.org/wiki/Director_of_the_National_Sec... Like hey look here’s the director of the NSA! Who cares?

The NSA is probably a bad example, given its entire existence was denied for two decades by the US.

There's some logic to hiding the identity... don't know who/who's family to hack/threaten? But, you can just target the 2nd, or 3rd in command it'll be just as effective really.... so I'm not sure what the point is, besides some sort of security theater coolness.

The US has oceans on both sides, and the bureaucracy involved in getting airfare reimbursed for a bunch of agents is too much for most governments.

It's less the airfare than the fact that they would stand out. Fort Meade is a defence installation, people don't just "pass by". Most people in the area would have military connections and be wary of even-slightly-foreign-accented individuals suddenly popping up near critical areas or where heavily-protected personnel actually live.

That's not really apples to apples though. If we start bombing Indian reservations and trying to exterminate Native Americans, I'm pretty sure they would quite deliberately become a threat to American equivalents of this guy.

(...or perhaps I'm misreading you? not sure)

There have been reasons for folks to be motivated to attack personally the leads of the various American spy agencies over the last century, there really hasn’t been a significant amount of time where we weren’t in some kind of military conflict, very many of them with questionable to explicit human rights violations.

Sounds about right, but currently the American spy chief doesn't have domestic enemies. That guy does.

The head of Israeli intelligence is operating in a different threat landscape (both figurative and literal) than the head of the NSA.

Human: There are hundreds of Hamas hiding in Al Shifa hospital.

AI: Go in and kill them

-- Is that why the picture of "destruction near Al Shifa Hospital" is in there, to lead us to believe it was the heartless software that led to the raid?

(https://edition.cnn.com/middleeast/live-news/israel-hamas-wa..., as if you can believe anything coming out of there)

So he published a book using his own private email. It doesn’t seem like he was really trying to keep it a real secret.

No - he included a different email address in the book to gather feedback from readers. This email address was then linked to his personal Google accounts (how, is not described - but I bet Krebs On Security would know how).

Where is the "Mossad Is going to burn your house down" copypasta now? Just for laughs.

Given how unreliable LLM is in general usage with its tendency to "hallucinate" -- a grossly anthropomorphic term, when it's just an alogrithm with no sense of what's "actual" or "fictional" -- to use this in government and law is irresponsible and criminaly negligent.

To use it in war is abhorrent.

Does it say anywhere that they were using LLM?

The book was published three years ago. The timing of this revelation is convenient.

In the subject article there is a link to Maariv article (In hebrew) https://www.maariv.co.il/journalists/Article-1078519.

Basically the guy drunk "Startup Nation" kool-aid and thought he was running a startup. 8200 built this data lake with billions of data points and started to rely on it, decreasing the importance of intelligence analysts. So they were getting signals but couldn't act upon them, something that senior analysts would have helped. That's why they ignored the senior analyst "V" alarms and called her delusional. And on that damned early morning when they started to get "low significance" signals from the system and the chief of staff, Mossad and the intelligence head decided not to do anything significant. At the same time, an analyst could have connected the dots.

This guy will be out and the rest as well.

> However, it has been criticised over its failure to foresee and prevent Hamas’s deadly 7 October assault last year on southern Israel, in which...

Does anyone believe that they didn't know about it?

I do, I think a lot of times massive fuckups get the "it was secretely a 4d chess move" thing but it's pretty believable (to me at least) that Israeli capabilities have degraded significantly over the decades as the IDF shifted from a war fighting army to prison guards

It's possible, of course, it just doesn't sound likely.

"Making an entire book about my AI death machine is the best idea I ever had!" is the intelligence equivalent of The Simpsons "Videotaping this crime spree was the best idea we ever had!"

edit: Since this thread is getting turfed so hard that my other, serious comment is already gone. I would ask you to read this thread about said AI death machine and make your own thoughts about what is currently happening.

https://news.ycombinator.com/item?id=39918245

“Is you taking notes on a criminal conspiracy?"- Bibi, probably.

Another 27-year-old brigadier general?

The major challenge in fighting terrorists is distinguishing them from the general population. Killing innocent people is against international law and harms the country doing it. While killing terrorists is acceptable and even desirable to many people.

If a targeting system is more intelligent, it should be better at reducing the loss of innocent life while still delivering the same kill rate of actual terrorists.

Israel’s system, based on reporting, does the opposite.

So we can see that their use of an AI system to do this targeting reflects an insight not into the operational benefits of AI, but the cultural benefits.

By spending tons of money and time and calling it “artificial intelligence,” the targeting team has developed a far more flexible pretext for violence. “This incredibly advanced AI system said he’s a target” is now enough to go drop a bomb… even if the incredible AI system is simply set to have its filters wide open.

This is what YS means by “A team consisting of machines and investigators can blast the bottleneck wide open.” The humans build the machines to be extremely credulous, and then tell everyone else that the machine ID’d each target as a legitimate target.

> Killing innocent people is against international law and harms the country doing it. While killing terrorists is acceptable and even desirable to many people.

Enemy combatant has a very specific meaning and that _can_ include people aiding and abetting the enemy. Thus, "actual terrorist" is a fairly nebulous term along with "innocent people". War is hell for a reason. Except for in obvious cases like Mai Lai it is never clear whether to call a person "innocent" for allowing an enemy combatant to seek shelter in their home - for example. But you have to make a snap decision: is the woman carrying what looks like an oddly shaped baby actually carrying bomb? Groceries? Why is she walking so close to the convoy? Did we issue a warning? Did she heed it? A bad decision will cost you many more lives than taking one. War is hell.

> the targeting team has developed a far more flexible pretext for violence

Doubt. The kill order chain is most likely AI flags target -> fed to human operator -> operator confirms -> kill order. Of course, your counter argument is that the human is given pretext ("this is a target") and thus assesses it with the assumption but in theory it's actually a better situation. 80% accuracy is still better than previous generations warfare. You dont want to end up on the thermals of an apache carrying a tube you need to replace your broken sewage system.

The news is polluted with pro-Palestinian nonsense. It's very clear absolutely no one on social media has any idea what hamas does, or what the people of palestine have done to aid and abet the terrorists. This is not to justify just dropping bombs on people without cause but I would be pretty satisfied with a best-effort approach to the loss of civilian life given how deeply embedded hamas is.

And then the Israelis put that target they only have partial confidence about into a system called "Where's Daddy?" which will track their target until he goes home to his family at which point a kill order is sent out to drones and warplanes to massacre his entire family. [0, 1]

Is that the "pro-Palestinian nonsense" you were referring to?

[0] https://www.theguardian.com/world/2024/apr/03/israel-gaza-ai...

[1] https://www.972mag.com/lavender-ai-israeli-army-gaza/

See the part where I said aiding and abetting a known target makes you a target. "What about the children" is a very poor argument. Would you rather a fireteam kick down the door and have to shoot the entire family when they all scurry to a room and grab arms because daddy told them to? Because that's what happens - all the time. There's no "put down your weapons" like you see in movies. You have about 50ms to make a decision to neutralize a target.

I know people who are very screwed up after having to shoot children carrying weapons because their parents told them to. I know people who've had to make the kill/no-kill decision on a woman and her child approaching a convoy. It's a fucked up situation. That's why we try to avoid war.

You can't just deploy operators when a guy goes home to his family. It's expensive, dangerous, and difficult. If you leave him alone he could be ordering people from his home to kill your people. Unconventional war is by definition unconventional. If they fought like a regular military their families would most likely not be targeted. I have no data but I can tell you absolutely no one wants to kill women and children. Except for hamas, that is.

At this point, "aiding and abetting" seems to be "existing while inside Gaza." See the deliberate execution of the world central kitchen workers in a three-phase strike to ensure maximum casualties as an example.

I mean, that's not true. But you won't agree with me as it's en-vogue to be pro-hamas. You should actually be ashamed of yourself how little you understand about how these terrorists operate and how boldly you claim shit like "genocide" and "execution".

I am against systems designed to target and slaughter civilians. I am against using starvation as a weapon of war. I am against rules of engagement so loose that it allows the military to kill a convoy of people coordinating directly with the Israeli military on a aid mission who are using GPS devices to transmit their location to the Israeli military using three separate strikes over 2.5 km. I'm against dropping 2000-pound bombs on dense civilian areas. I am against rules of engagement so loose it allows the Israeli military to kill three of the hostages it was purportedly trying to save while they're waving white flags. I am against using snipers against people hiding in churches and hospitals. I am against an approved collateral civilian kill rate of 15-20x.

I consider these positions to be pro-humanity.

If in your view, holding these positions makes me pro-hamas, I encourage you to examine if you are using an unnecessarily binary framing to analyze the death and destruction going on.

You can't just scowl seriously and declare "war is hell", and pretend that excuses everything. War may be hell, but it's only hell because of the individual decisions of actual people up and down each belligerent's chain of command.

Yes, but Hamas instigated this by attacking civilians and the PLO—and only the Palestinians and its silent backers—threw away a dual-nation solution during the Clinton era.

So when we include decisions, it’s important to be impartial here.

Blaming the failure of the 2000 Camp David talks solely on the PLO is a convenient lie that seems to only serve to justify the increasingly harsh treatment of Palestinians by Israel. The historical record shows that both parties were not willing enough to compromise in order to create the two state solution [1].

[1] https://en.m.wikipedia.org/wiki/2000_Camp_David_Summit

The convenient lie is papering over the blame where it squarely belonged.

Like many of my generation—Gen X—I was closely following the on-goings because it seemed so close (Berlin wall, Russia seemed like an ally; more peace in our time); but in the end Arafat scuttled the deal because he did not want to compromise and give-up any part of Jerusalem.

“Arafat rejected Barak's offer and refused to make an immediate counter-offer.[100] He told President Clinton that, "the Arab leader who would surrender Jerusalem is not born yet.”

https://en.m.wikipedia.org/wiki/Yasser_Arafat [1]

The citation [100] here is Jimmy Carter's book, which not only does not support the wiki summary (Camp David was based on a proposal from Clinton--not Barak, and Barak had "twenty pages of reservations" about the proposal himself), Carter also concludes:

[1]

> There was no possibility that any Palestinian leader could accept such terms and survive, but official statements from Washington and Jerusalem were succesful in placing the entire onus for the failure on Yasir Arafat. Violence in the Holy Land continued.

In regards to the following talks at Taba, which perhaps the wiki is conflating with Camp David, Carter continues:

[2]

> A new round of talks was held at Taba in January 2001, during the last few days of the Clinton presidency, between President Arafat and the Israeli foreign minister, and it was later claimed that the Palestinians rejected a "generous offer" put forward by Prime Minister Barak with Israel keeping only 5 percent of the West Bank. The fact is that no such offers were ever made.

[1]/[2] Jimmy Carter - Peace Not Apartheid (it's on Libgen, go look for yourself like I just did)

> but it's only hell because of the individual decisions of actual people up and down each belligerent's chain of command.

If a 10 year old takes up an AK-47 and points it at you what are you going to do? If a woman carrying a child sets a suspicious package down what're you gonna do? This isn't a "decision by the chain of command" outside of ROE for the AO. Everything else is on you.

If your answer is "call the chain of command" your entire unit is now dead. Congratulations. The variance in the decisions is due to the fact you can't make decisions by debating the internal psychology of the moron who may be trying to kill you. Innocent people will die. That's war. I'm getting downvoted to death here because it seems the posh liberals sitting at their desks don't understand what an actual life or death decision feels like. To be honest I'm not shocked. The toughest decision most people make is what to have for lunch and even that is enough to give the yuppies anxiety.

> Killing innocent people is against international law

It’s not.

Specific instances of killing innocent people where a number of factors are met is against international law.

Innocent people inevitably die in war. That’s why the world so desperately wants to avoid it.

But international law acknowledges that even if a actor does everything right, innocent people will die.

Yes, there is “fog of war” and “collateral damage.” International law requires that combatants take steps to minimize that. And not intentionally target innocent life.

It is. By definition (Because they are innocent)

Even if you fail to avoid shooting innocent people and can find a scapegoat to save yourself from the consequences, homicide is definitely, unequivocally, illegal. And assassination (killing somebody that is helpless, like bombing its house or a sniper shooting a child) is even more illegal.

And any army that does not try at least to study a basic understanding of international war laws are a bunch of unprofessional criminals.

Where does this misunderstanding come from?

https://ihl-databases.icrc.org/en/customary-ihl/v2/rule14

“an attack which may be expected to cause incidental loss of civilian life, injury to civilians, damage to civilian objects, or a combination thereof, *which would be excessive in relation to the concrete and direct military advantage anticipated*”

Note that proportionality is ill-defined, leaving a lot of room for interpretation.

> Where does this misunderstanding come from?

Greed. War is very profitable.

There will be always people trying to twist the words, and justify the unjustifiable.

Right, but have you given thought to the greed and maliciousness of those that prop-up Hamas?

Also, war has never excluded the casualty of innocents except in the movies.

What’s not reasonable is when an attack goes out of its way to kill civilians.

It’s why it was a big deal and a disgrace when Trump pardoned US soldiers who knowingly and without cause killed civilians.

And why what Hamas instigated was especially appalling and inexcusable. And why Netenyahu has overplayed his hand, to cover for his failings.

> Greed. War is very profitable.

That's the reason for your misunderstanding?

Because it's your statement that is incorrect.

> it's your statement that is incorrect.

Wrong

There is not a single country where I have the "right", granted by law, to shoot dead your 8Yo children. Period.

Some people, like dictators or criminals, just can jump over the law without consequences. And laws have exceptions that modify the punishment so in a few --extreme-- cases I could do it without serious consequences to me.

But: "Is an 'all you can kill' party and there are good chances that I will not be punished"; is still profoundly different as "is legal to kill children (as long as they are in a minority)".

And It always will be different.

> There is not a single country where I have the "right", granted by law, to shoot dead your 8Yo children. Period.

Sure there is.

If I'm a policeman, and in order to save lives I need to use my weapon and it results in the deaths of an 8 year old, that's permissible by law.

And so do the laws the of war.

If, in order to destroy the war fighting ability of an aggressor, I bomb a civilian center where the aggressor has placed offensive weapons, and is is judged by probability and proportionality to be reasonable, then it's legal.

Israel gets to define anyone killed as a terrorist so the AI can never be wrong

Alternatively they can just label anyone resisting extermination as such so the entire population can be cleansed

> Israel gets to define anyone killed as a terrorist so the AI can never be wrong

I will point out that the US does the exact same thing. Both are wrong of course.

I was once the first full-time employee at a small company. I had to answer the phone for sales calls sometimes when the owner wasn't around.

I started there as a consultant who designed and coded the sales CRM application. I had full control of everything.

When I answered the phone for sales, and someone would ask for a discount, I would always say "I'm sorry, the computer just won't let me do that." The blame was passed to the computer, and I got out of that uncomfortable situation. The customer always understood. After all, "it's up to the computer," what can you do?

What we are reading about in this article is the kill list version of that.

A couple years ago, I wrote on this website that I am much less worried about AGI killbots, than I am about humans using dumb AI to kill people. Now, here we are.

> In one chapter of the book, he provides a template for how to construct an effective targets machine drawing on “big data” that a human brain could not process. “The machine needs enough data regarding the battlefield, the population, visual information, cellular data, social media connections, pictures, cellphone contacts,” he writes. “The more data and the more varied it is, the better.”

They should hold more strategy-oriented courses at the Israel Military Academy, that is if they have a military academy worthy of that name and if this intelligence guy is a graduate of one. Because it looks like he can look no further than the tactical level ("how do we kill leader X of enemy group Y at time moment Z?"), which is a suicidal thing to have as your top-military commanders in the middle to long term (because, no, killing enemy leaders X1...Xn won't win you any wars in the middle to long term).

Statistical models are used for at least 80 years to calculate the size of bombs when hitting targets and to factor in civilian casualties. Today's statistical models have much better PR but it is not much different.

Also war is bad, but sometimes it is a necessity when the alternative is worse. It is always quite easy to cast yourself as a neutral moral judge on the sidelines, but we all saw what happened to the USA in the early 2000s when faced with a lesser threat than Israel.

With the way the world is heading what we are hearing are simply echoes of the pre-ww2 appeasement attitudes which of course are a great way to prevent future deaths and war

Reply on news.ycombinator.com