I expect Anthropic to put out a statement on whether it was used in the selection of that school full of children in Iran as a target.
vomitHatSteve on
Incorrect! It is – in fact – humans making the decisions.
If I roll dice to generate a random latitude and longitude and then drop a bomb on that location, it was me who chose to bomb there, not the dice.
Full stop. The humans who sign off on the „recommendations“ made by complex RNGs are the ones who chose the locations to bomb.
shmightworks on
duh duh dum duh duh
ThrowawayAl2018 on
The people who press the „launch“ button should be held accountable. See Nuremberg trials – illegal orders passed down is still illegal.
Wasn’t there a recent report that AI resorts to war in 95% cases of a simulation!?
c7hu1hu on
„If you can’t tell if you’re targeting a day care or an SA-21 Growler launch site, calm down! You’re not crazy — you just need to look at the telltale differences between them.“
celtic1888 on
This will end up being the excuse for atrocities like the Girls School Bombing
We didn’t do it ! It was that AI guy….
celtic1888 on
If you kill a bystander by accident during a bank robbery you still get charged with murder….
tabrizzi on
Like it or not, that’s the future of warfare.
Tribe303 on
I’ve assumed it was AI trained on an old dataset, because that girls school that was bombed WAS part of the nearby naval base.. Until 2016 when walls were added to separate them, and are visible on the satellite pictures from 2016 onwards.
OGBeege on
Maybe? Ya think?
BusyHands_ on
Would be a waste of munitions to randomly drop bombs when you consider the overall cost to make one.
Definitely needs Intel and humans making the final decision. AI probably selects targets based on parameters but a human is calling the shot
MoneyManx10 on
This will be their excuse for dropping a nuke.
El_Gran_Che on
Obviously not humans, next step is autonomous weaponized AI
BarmyBob on
If this is the future if war, certain real estate areas known to have previously been targeted now become places that may be targeted in the future, driving down value.
tim3k on
New captchas will be just some satellite images with „select all squres that are NOT a school“
BadSkeelz on
„Claude, what do I bomb to get the world to stop talking about how I like little girls?“
„Well there’s this school…“
JLRfan on
I got permabanned from r/artificialintelligence for posting about this
JLRfan on
I think it’s possible it was a Claude error that led to the US bombing of the Iranian elementary school.
NovelDraft5175 on
Where ever Israel can get the most ethnic cleansing for their buck.
CopiousCool on
Don’t fall for this BS, the first thing you should think of when adding any automation is who has the responsibility to check it works … otherwise the person implementing it is responsible
They’re trying to pass this off as if it’s a new phenomenon when in fact this question was raised when automated weapons were created and tested in Korea and more recently in Palestnie
>*“Samsung SGR-A1: South Korea developed and tested the SGR-A1, an autonomous sentry gun, in collaboration with Korea University and Samsung Techwin (now Hanwha Aerospace). First tested in the Korean Demilitarized Zone (DMZ) in 2010, the system was designed to replace human guards along the border with North Korea“*
OVER 15 YEARS AGO
If they still haven’t been bothered to check the procedure its humans at fault, plain and simple
gearstars on
I thought Captain America wanted to stop Project Insight? Why does this administration hate Captain America?
hughdint1 on
Fox drunk used ChatGPT to do his job
Leave A Reply
Du musst angemeldet sein, um einen Kommentar abzugeben.
22 Kommentare
I expect Anthropic to put out a statement on whether it was used in the selection of that school full of children in Iran as a target.
Incorrect! It is – in fact – humans making the decisions.
If I roll dice to generate a random latitude and longitude and then drop a bomb on that location, it was me who chose to bomb there, not the dice.
Full stop. The humans who sign off on the „recommendations“ made by complex RNGs are the ones who chose the locations to bomb.
duh duh dum duh duh
The people who press the „launch“ button should be held accountable. See Nuremberg trials – illegal orders passed down is still illegal.
Wasn’t there a recent report that AI resorts to war in 95% cases of a simulation!?
„If you can’t tell if you’re targeting a day care or an SA-21 Growler launch site, calm down! You’re not crazy — you just need to look at the telltale differences between them.“
This will end up being the excuse for atrocities like the Girls School Bombing
We didn’t do it ! It was that AI guy….
If you kill a bystander by accident during a bank robbery you still get charged with murder….
Like it or not, that’s the future of warfare.
I’ve assumed it was AI trained on an old dataset, because that girls school that was bombed WAS part of the nearby naval base.. Until 2016 when walls were added to separate them, and are visible on the satellite pictures from 2016 onwards.
Maybe? Ya think?
Would be a waste of munitions to randomly drop bombs when you consider the overall cost to make one.
Definitely needs Intel and humans making the final decision. AI probably selects targets based on parameters but a human is calling the shot
This will be their excuse for dropping a nuke.
Obviously not humans, next step is autonomous weaponized AI
If this is the future if war, certain real estate areas known to have previously been targeted now become places that may be targeted in the future, driving down value.
New captchas will be just some satellite images with „select all squres that are NOT a school“
„Claude, what do I bomb to get the world to stop talking about how I like little girls?“
„Well there’s this school…“
I got permabanned from r/artificialintelligence for posting about this
I think it’s possible it was a Claude error that led to the US bombing of the Iranian elementary school.
Where ever Israel can get the most ethnic cleansing for their buck.
Don’t fall for this BS, the first thing you should think of when adding any automation is who has the responsibility to check it works … otherwise the person implementing it is responsible
They’re trying to pass this off as if it’s a new phenomenon when in fact this question was raised when automated weapons were created and tested in Korea and more recently in Palestnie
[https://www.lawfaremedia.org/article/foreign-policy-essay-south-korean-sentry%E2%80%94-killer-robot-prevent-war](https://www.lawfaremedia.org/article/foreign-policy-essay-south-korean-sentry%E2%80%94-killer-robot-prevent-war)
>*“Samsung SGR-A1: South Korea developed and tested the SGR-A1, an autonomous sentry gun, in collaboration with Korea University and Samsung Techwin (now Hanwha Aerospace). First tested in the Korean Demilitarized Zone (DMZ) in 2010, the system was designed to replace human guards along the border with North Korea“*
OVER 15 YEARS AGO
If they still haven’t been bothered to check the procedure its humans at fault, plain and simple
I thought Captain America wanted to stop Project Insight? Why does this administration hate Captain America?
Fox drunk used ChatGPT to do his job