Hacker Newsnew | past | comments | ask | show | jobs | submit | nickdothutton's commentslogin

Better to have smart bombs than dumb ones. Or rather, better to have 1 smart bomb than 1000 dumb ones spread across an entire city in order to pick off the particular building, vehicle, or person you want.

Specially AI Hallucination bombs, that hit a park named "Police Park", because it thinks it's killing policemen[1], or a children school with Shahed in the name[2], because it thinks It has something to do with drones.

[1] https://x.com/MarioNawfal/status/2029575052535173364

[2] https://www.aljazeera.com/news/2026/3/6/elementary-school-in...


There's also a chasm of (non-)accountability.

You or your subordinates target an elementary school: that's a war crime.

Your "battlefield AI" targets an elementary school: software bug, it happens, can't be helped.


This isn't even that new. Part of the motivation for building autonomous nuclear response programs during the cold war was specifically to remove accountability, and guilt, from human operators. But AI does bring it to a new level.

> Part of the motivation for building autonomous nuclear response programs during the cold war was specifically to remove accountability, and guilt, from human operators.

Details please. Because I can see the reality being most likely an attempt to avoid conflict by solidifying MAD, by trying to prevent a human from vetoing a second-strike.


At least it is the plot of a lot of movies of this era: Dr. stragelove, wargames, colossus.

The software is never accountable, so the human running it is always accountable.

that is how it should be, not how it is.

"War Crimes" only apply to the loser of the war and are prosecuted by the victor.

Meaning whatever horrors are done on either side, only the horrors committed by the loser will be "crimes". The inclusion of AI doesn't change that.


sadly that's also true within Ukraine. like, I know that Russians are handling Ukrainian prisoners of war very brutally (no sources, why: [0]) but, if not for [0] AND if I wouldn't be killed by my co-citizens for that, I would point out a good chunk of misconduct on Ukrainian side as well.

I also recall the history lessons. I can't remember anyone who committed a war crime against Nazi Germany that also was internationally prosecuted. yep, the West did prosecute domestically, and there were some loud cases with German POWs, but I can't recall any, any Soviet soldier being charged for e.g. rape.

[0]: there is nothing public to link to that remained up, and I'm long out from private Telegram channels where such videos are posted; plus, even if I could, you and mods wouldn't want to see the video of someone getting beheaded


Your links talk about the places that were bombed, but I don't see anything apart for conjecture that this was the product of AI targeting.

Also this is a vast underestimate of the ability of organizations that were able to locate most of Iranian leadership throughout the war in their hiding places, but suddenly their Farsi is so bad they need a twitter account to tell them this is a Park


It's a popular conspiracy theory, without evidence, and without any perspective on any information that intelligence had. Using civilians as shields is well documented/known for Iranian military and groups they sponsor. For example, hospitals [1].

Shitty, but possibly a valid military target.

[1] https://www.gatestoneinstitute.org/8666/yemen-human-shields


> Gatestone Institute is an American far-right think tank known for publishing anti-Muslim articles.

> The organization has attracted attention for publishing false or inaccurate articles, some of which were shared widely.

> The Gatestone Institute has been frequently described as anti-Muslim, regularly publishes false reports to stoke anti-Muslim fears, and has published false stories pertaining to Muslims and Islam.

- https://en.wikipedia.org/wiki/Gatestone_Institute

The US and Israel have repeatedly claimed that schools and hospitals are legitimate military targets with no evidence. A highly partisan think tank which is known for putting out misinformation is not a valid source.

If you're going to destroy hospitals and target civilian infrastructure and kill children, you should be accountable on a world stage and provide evidence. Unless you would you accept Iran bombing elementary schools in the US because they claim to have intel that there are terrorists hiding under them?


It's more complex than that, you have direct evidence of Iran recruiting 12 year old child soldiers in this war (https://www.bbc.com/news/articles/c9wqgjn7x89o).

Using stadiums as security forces hiding places (https://www.wsj.com/world/middle-east/israel-iran-leadership...)

Misusing hospitals for military purposes (https://www.iranintl.com/en/202602215486) and schools (https://x.com/IranIntl_En/status/2032846253189411146 https://www.instagram.com/p/DV8NUQPFJJv), while the last ones are weak visual evidence and backed up by rumors, Iran is currently under internet blackout for over a month as the regime is interested in controlling information flowing out of the country.


There are MANY examples of Iranian backed terrorist organizations doing this (which I thought might be too indirect), but here's something more recent [1].

Regardless, left leaning news reports things that make them look good and the opposition look bad. Right leaning news reports things that make them look good and the opposition look bad. Both are needed to find truth because they're all biased for profit corporate entities owned by 6 different billionaires that will only report what's convenient for that bias.

And no, I have no trust in the claims of the Iranian government. Do you? Who do you believe does?

[1] https://www.msn.com/en-in/politics/international-relations/i...


What's the running rumor right now of which AI was involved? I heard Claude awhile back, but this makes me wonder how much Redhat could have been involved?

This has nothing to do with AI, the school got hit because it was directly next door to a military base.

You're mistaking it for Shajareh Tayyebeh Elementary School[1], double tapped with tomahawks in the opening salvo of the war. That was another school, hit later. There was multiple schools attacked.

[1] https://en.wikipedia.org/wiki/2026_Minab_school_attack


It is completely unfounded speculation that AI was involved.

That is target selection and has nothing to do with dumb vs smart bombs.

There are no mainstream sources about the police park story when I looked. I’m pretty sure it’s a hoax.

Channeling my inner Socrates:

You want consensus from non-experts for a plan to use 20 smart bombs.

Your opponent wants consensus for a plan to live-stream a demo of 1 smart bomb, and then use 19 dumb ones.

Your team has more expertise.

Your opponent's plan saves enough money to buy a better PR team than yours, and is still more cost effective than your plan.

Who wins?


That “smart” vs “dumb” distinction doesn’t apply here though. What is discussed has nothing to do with the ability to physically land a bomb in a precise location, that problem seems to be solved reasonably well already. “Smart” in this case has more to do with using ML/LLM to select a target.

You can rationalize anything by only considering the upside relative to alternatives' downsides.

The reality looks more like the worst of both worlds to me.

If you genuinely needed only a handful of "surgical strikes", thete would be no need to "compress the kill cycle".

What we see in Gaza, Lebanon and Iran looks more like "smart carpet bombing": Some AI system generates a continuous stream of "targets" from sensor and intelligence data, according to whatever criteria political leadership defines and according to a given level of allowed "collateral damage", then those targets are immediately fed to drones or warplanes to destroy - essentially a continuous "pipeline" that probably "ideally" (in the dreams of those people) should become fully automated.

For THAT kind of vision, "efficiency" in destroying any particular target and checking all legally required boxes as quickly as possible is probably paramount.

(And in addition to that, there are probably still enough "dumb bombs" if no one is looking)


Its the plot of Captain American 2 with those 3 aircraft carriers at the end

You might be right, but that's terrible

Smart bombs are no good if they are directed by a dumb AI targeting system, a dumb alcoholic accelerationist religious fanatic Secretary of War, or a dumb narcissistic genocidal pedophile President.

I had a thought about that. Using weapons that don't give the opponent an opportunity to surrender feels like a war crime.

There is one more layer - America voted for this.

In fact, it didn’t. Trump continued to make “no new wars” a plank of his platform.

Some of his base will follow wherever he goes, but he would not have been elected without those who supported him on the basis of this (broken) promise.


Trump said this wasn’t a war.

Americans voted for this confrontational, disrespectful, and chaotic way of governing.


And the 36% that sat out for reasons also contributed.

Use vote.gov to find out if youre registered

Use 5calls.org to find out about top issues to call your reps about.

Use govtrack.us to find out about what's going on.

If there was a github.com view of the govt agenda it'd be easy for more people to comment on things. At all scales of governance


POC of GTFO should apply to AI models too, or the false positive rate will overwhelm.

What a great article. Several years ago I wrote a (far inferior) post along similar lines, using a famous railway/bridge disaster[1]. Study of these kinds of engineering failures, even those from hundreds of years ago, are so revealing. I'd lay money on similar reports from the days of the ancient Egyptians as being just as valuable.

[1] https://blog.eutopian.io/the-age-of-invisible-disasters/


Hard to imagine now, but this was a huge turning point. A genuinely powerful CPU in a "Pee-Cee" available for less than RISC workstation money. I had to wait a while, mine was an AMD DX2-66 since I didn't have a budget for Intel... add Slackware... and countess hours messing with XF86config and I had a poor-mans Sun workstation.

These are interesting numbers for engagement but don't mean as much without equivalent stats for the other platforms. It's a little like when a news story quotes only a percentage (but not the absolute figure in $) or vice versa.

Agreed.

Assuming they use the same principles everywhere, they're getting more views on Mastodon and Bluesky? That is surprising.


Not really, their target audience is much more likely to hang out on Mastodon and Bluesky. So even if the impressions might be fewer the quality of them is almost certainly higher.

Fair, their post gave a nod to the believers I suppose, and it's reasonable to assume they have different metrics of success for getting the message out to believers vs as they described "The people who need us most are often the ones most embedded in the walled gardens of the mainstream platforms and subjected to their corporate surveillance. Young people, people of color, queer folks, activists, and organizers..."

Having said that, I'd argue that X meets the definition of "walled gardens of the mainstream platforms and subjected to their corporate surveillance."

But, it feels like based on this comment, they should still be on X "We stay because the people on those platforms deserve access to information, too. We stay because some of our most-read posts are the ones criticizing the very platform we're posting on. We stay because the fewer steps between you and the resources you need to protect yourself, the better. "

And view counts aren't available on most platforms, but on tiktok, where they are, they seem to have about 60k plays or whatever in the past 6 months. So, I'm not sure how you can argue that X is de minimus, but, gotta be on tiktok for reasons, that also apply to X, but, X is de minimus and tiktok is not, even though we get many more views on X.

Anyhow, with this response I've spent 10 more minutes thinking about this than I should, I will leave it here with the closing thought that their post feels very disingenuous.


Also if you tweet a link to the content instead of tweeting the actual content, you get penalized by the algorithm.

They do this in almost every tweet.


This has changed recently. Links no longer appear to be penalized.

I use a separate user account per project on my local machine (not in sudoers), which I ssh to, and also runs tmux. If I need claude code in windows, then I run a VM. The performance and (in)convenience cost of this to me is minimal. I started working this way in order to limit the "blast radius" when claude went on a dependency binge within a project.

"Gradually, then suddenly", as someone once said.

Title is wrong, should be "New form of cancer discovered".

When I got started, the NSFnet backbone was a bunch of IBM RS/6000 systems with comms cards. There were no routers.[1]

[1] https://www.rcsri.org/collection/nsfnet-t3/


There were routers, just no T3 (45 megabit) capable routers.

Recently working on implementing a "MS Teams for the terminal" (video conf, audio, chat, file sharing, recording, etc, the usual things you would expect). Linux and Mac, FBSD and others to follow. Prior to the 1M context window I found I had to restrict myself to specific functionality/areas of the codebase. I'd gotten lazy anyway so this was no bad thing. Reduced the "vibe" quotient of the AI coding.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: