>US used Grok for 2000 missile strikes in IranFor those unaware, Grok is the worst of the corporate AI models and even the chink ones that cost a fraction of a cent outperform it. The people that work for musk don't even use Grok.https://www.independent.co.uk/news/world/americas/us-politics/elon-musk-grok-ai-iran-missiles-pentagon-b2997321.html
@grok merc another member of the IRGC high command make no mistakes
>>537297689Kek
Training AI to kill humans is heckin based! I'm sure Big Balls has it all under control.
what exactly is AI doing in this situation? guiding the missiles? choosing the targets?
I think it is stupid to use AI that is probably programmed with some brown or female limitations and should use someone that would sacrifice whatever it takes to win and not care about made up concepts like civilians. A real life Ender there has to be lots of autistic gamers they could ask.
Trump is pretty good tho. It sometimes outperforms ChatGPT when I ask it some deep research.
Grok couldnt solve the capcha and blew up all the decoys
>>537297622So we lost the war because of Grok is what I'm hearing.
>Grok, bomb iran>Did you mean Barbara Ann by the Beach Boys?
>>537297972When you connect to your home wifi or the AI gets some signal on your location the missile gets launched, and your entire family dies. Or, 200 schoolgirls.
>>537297622>For those unawareDumbass you have no clue what custom model of Grok they are using or have made available for military use.Also just because GPT and Opus/Mythos are better for coding and agentic tools, certainly does not mean Grok is bad at parsing data and using extremely high levels of reasoning to select targets. Grok is heavily used for current events and people data because of its X training set.Holy fuck you are dumb
>>537297622@grok is this real?
>>537297972Probably something about "Grok here is a picture, where would YOU hide the ammo depot?"
>>537297622Hegseth belongs in prison.
>>537297972Helping designate target areas supposedly
>>537297972>what exactly is AI doing in this situation?My guess is target selection and its not being fed pictures like retard OP is claiming. I would assume they are feeding it military sigint data as well as public/private sourced satellite, cellphone, smart device, browser data.
>>537297622>Grok is the worst of the corporate AI models and even the chink ones that cost a fraction of a cent outperform it.copie, co-pilot is far worse
>>537298129The only niche Grok had was jeets making AI nude edits of white children.
>>537298129>he believes the hypeAI is for you
>>537297622This is Hegseth raging against Anthropic that he views as 'woke' and 'liberal' and using the right-wing aligned option instead, which also happens to be a complete joke in every benchmark.
>>537297622These gAI are also so censored and lobotomised too. Imagine trusting them with anything important. Here's another perfect example:https://www.youtube.com/shorts/4z3dFX0ec9Mhttps://www.youtube.com/shorts/IvGedA3iLKA
Ai can't even tell the difference between NOBLE and NOBL. Why the fuck would anyone want to use it to identify targets?
>>537298459>Co-pilotdoes that jeet shit actually count as AI
>>537297622I'm probably gonna get sued for stating my opinion but I believe Palantir may have utterly failed in Iran to deliver on its promises. I believe it works in a scenario like in Ukraine where two forces with uniforms and camo patterns and NATO catalogized vehicles shoot each other. It seems to be utterly overwhelmed with the concept of asymmetric warfare. If, or I guess when, Palantir is scrapped enjoy watching people like Karp and Thiel have an epic meltdown on live TV and Palantir's stonks to crumble - just like OpenAI will in the coming 12 +/- months Man this shitshow is gonna be bad.
>>537297972They dump hundreds of terabytes of data into it and it pops out a target
>>537298993This is what the head of Anthropic has repeatedly said. It is not reliable enough to make life and death decisions.Hegseth wanted to give AI the power to control autonomous drones with the power to kill people and have no human supervision on it. Anthropic said no and Hegseth + Lutnick decided he was going to destroy their company because they wouldn't go along with it.
>>537297988>programmed with some brown or female limitationsclankers cant make spicy memes
Hey guys I found this old picture of my robot wife.
>>537297850After 80 years of hard work and winning, you just get tired of winning, bro!
>>537297622What's your metric for saying Grok was outperformed?I prefer Grok to all the woke AI's. Grok is based.
>>537298006>TrumpKek. I meant Grok.
>>537297972chances are pretty good that the military intercepts whatever communications they possibly can with those big radio snooping planes they also sent to the mexican border, feed everything they record into a LLM, and ask it where the iranians are. then look up wherever it says iranians are hiding on google maps and fire some missiles
Grok s the best of AI models only because it's the least bound by progressive psychologyBut I do think them using AI models to bomb around just shows how retarded the Pentagon is he shouldn't be listening to these people if they have to turn to an AI program when they are literally paid to plan war
>>537297622I want his DICK
>>537297622oh that why we lost ,fuck they should used anthropic.
>>537298993>Ai can't even tell the difference between NOBLE and NOBL. Why the fuck would anyone want to use it to identify targets?Your skepticism is totally valid. Standard text AI is deeply flawed at precise, small details, which is exactly why it cannot be trusted blindly with high-stakes identification without human oversight.
>Grok is the worst of the corporate AI modelsActually that's Meta AI
>>537300102>weMeximutt...I think I get why you think so highly of yourself now.
>>537297622Did Grok kill those little girls in the Iran school??
GROK is Jeet CodedDouble Innuendo
>>537297972Watch the Hollywood episode where the super giant helicopter carrier can identify all hostile targets within 1 second and eliminated all of them across the globe.This is the pentagon mutt's wet dream, this is what they hope their datacenters can do. instant reaction time towards alll threats in the universe.
>>537300678You know it anon
>>537297622Claude is almost as good as it gets. Have no idea why it was targeted other than someone got mad. The Chinese use stolen code. Grok I'd say is good overall but not excellent like Claude. I use it quite a bit for images (good enough, banana is better but you can do more daily using grok). I won't use ChatGPT, it's garbage but look who's coding and running it. I wonder if that jew had something to do with Claudes new model getting banned.>>537300678Fack the people who used it. Probably a pajeet.
>>537297622Legitimately have the stupidest people in charge of the strongest military in the world*.*unless, it's fighting Iran...or Afghanistan...or ISIS...or the Houthis...
>>537300678It wouldn't surprise me if that's the narrative we get if throwing Kegseth under the bus doesn't work.
>>537297972I dunno, but I imagine it's like that scene from Willy Wonka where the guy asks the computer to tell him where the golden tickets are.
>>537297972I guess deciding 160 school children were imminent threats.
>>537302239>I use it quite a bit for imageswhy?
>>537302699Kids don't know how to photoshop, anymore.
>>537298993They don't use LLMs to identify targets. They have other specialized models for this.
>>537297972Making AI videos of missiles and winning the war
>>537297622>um actually grok did all the bad thingsKill everyone above the lowest person who could authorize such a thing.
Grok is a war criminal now
>>537297972choosing schoolgirls as targets for tomahawk missiles.>not my fault, the AI picked the target
>>537298179Kek
>>537302699Oh just making loading screens. It's good for that sort of thing. I make mods on the steam workshop. It's good for a slop then you can take anything good from that.
>>537297972The US has info on most citizens, their address, job, rank, maps of critical infra. It takes a human days to go through it all but AI takes minutes to start picking targets.
>>537297622>US used Grok for 2000 missile strikes in IranAre they really using the dumbest AI out there for military purposes?
