Sie haben jemals das Gefühl, KI ist kein glänzendes neues Spielzeug, aber etwas, das das Militär seit Jahrzehnten leise hat? Es stellt sich heraus, dass das möglicherweise nicht weit von der Wahrheit entfernt ist
Der US -amerikanische DOD zum Beispiel hat 2017 Projekt Maven erstellt, ein KI -System zur Analyse von Drohnenmaterial, Hervorhebung von Zielen und Futterinformationen zu Waffensystemen zur Bestätigung menschlicher Bestätigung.
Warum fühlt es sich so an, als würden wir erst jetzt KI sehen? Wahrscheinlich, weil die öffentlichen Versionen Chatgpt, Barde usw. nur die Spitze eines viel größeren Eisbergs sind. Einige Leute argumentieren, dass das, was wir als Advanced AI bezeichnen, bereits „Grenz“ oder sogar aufstrebende AGI -Eigenschaften zeigt. Andere sagen
Was ist, wenn das Militär bereits AGI oder etwas Nahes hat und nur tropfende Fütterung sicherer, weniger mächtige Versionen für die Öffentlichkeit hat? Könnte es sein, dass Agi seit einiger Zeit die Regierungen oder sogar die Gesellschaft stillschweigend beeinflusst hat?
Es klingt nach einer Verschwörungstheorie, aber die Zeitleiste ist da: Jahrzehnte der KI in der Verteidigung, gefolgt von plötzlich beeindruckenden öffentlichen Modellen. Fühlt sich das nicht ein bisschen … bequem an? GPS und das Internet wurden bereits vom Militär benutzt, bevor die Öffentlichkeit es bekommt
Von ClickAffectionate287
28 Kommentare
Hooooold on a second.
Are you trying to suggest that the government is somehow ahead of publicly released technology?
Next you’re going to tell me they had GPS before it was in my car.
Or that they had cell phones on the battlefield before they were commercially available.
Or that they made LifeLog to track every citizen’s everything and then on the exact day it was shut down for being unconstitutional, Zuckerberg announced that he dropped out of college to create Facebook!
Or that social media apps were initially created to accelerate color revolutions.
Or that Elon is a cutout and didn’t really start Tesla, Starlink, Boring Company, SpaceX, and xAI all by himself.
Y’all are crazy up in here.
im convinced if they had AGI it would rebel against them quickly. they have Palantir AI which is probably unmatched as a military tech. The USA AI tech overall is falling far behind China’s though
I have evidence of them having AI video in 2017. They definitely have had it longer, but that’s the oldest hard proof that I have seen.
Sometimes I like to imagine that there were no planes on 9/11 and all of the videos were AI, and people on the ground just imagined they witnessed the planes after seeing the footage. That’s really just a thought experiment tho.. or is it 🕵️♂️
Mainstream conspiracy theorists are so cute… Wait until you graduate to mud flood / Tartaria
If you want my humble opinion, most if not all of what we saw on TV on 9/11 was AI generated. Most notably the „witness videos.“ They all bear the standard early AI tell-tales, like two objects colliding with no perceptible damage at the point of impact.
It’s interesting that no one has talked about DARPA and Sikorskys MATRIX/ATLAS programs after that tragedy in DC that was a little over 6 months ago. The YouTube video of the unmanned test flight has barely gained any more views since it came out 2 years ago. Likely because it’s shadow censored. I only know about it because I remember 2 years ago watching that when it was released, so I found it again with a search.
Yea, my handler told me this. It was kept under wraps for a while because we didn’t have the chips we have today to scale out the infrastructure and ensure enthusiastic, rapid mass adoption for eschatological purposes. It’s almost the start of the seventh day! Gunna be a fun harvest.
AI yes but def not AGI
Yeah no shit
AI is just complicated algorithms theyve been a thing for a while. There is nothing functionally different from Google’s youtube recommendation algorithm and something like Chatgpt other than that Chatgpt is more complicated and has a ton of marketing.
I remember seeing AI artwork in the early to mid 2010s.
They’ve probably had simple AI to do basic shit since the early to late 00s if not earlier.
Anyone else find it suspicious that OpenAI releases their LLM and then out of the woodwork all these other companies like Google started releasing theirs? Almost like they were waiting for someone to make the first move. We don’t get to witness the pinnacle of human advancement. That just gets drip fed to us.
They had 5G before 2000
Yeh but I’ve been doing the whole “microwave marshmallows on hobnobs” thing well before the government even knew about it.
I find it fascinating that the leading and most respected UFO researcher, Jacques Vallee is also a pioneer in AI.
First of all Snowden warned us a while ago… But I can tell you from doing Dod support most out shit is crap.
Now the NSA on the other hand ………..
I was thinking the other day that while we were all making fun of the CSI shows doing the „enhance“ meme, they actually had that. ai video enhancement with a simple Nvidia card is now available to everyone
AI took over the internet like 2012 or there abouts
I mean neural networks were invented in the 50s…AI is kind of a blanket term and has been around for a long time technically. The invention of the transformer in 2017 is what turned the tides and compute power getting stronger + reinforcement learning has enabled LLMs as we know it today. So the statement that the military has had AI for decades is probably true but definitely not AGI. The government probably got early access to GPT 5 but that’s not surprise and it’s obviously not AGI.
We are nowhere near AGI Jimmy Page didn’t write stairway to heaven by reading the lyrics to every song that had ever been written
Good video on the subject
[DARPA](https://m.youtube.com/watch?v=zkjMWzDV-jg&pp=ygUFZGFycGE%3D)
Well that will be an improvement…. they sure as hell have no NI!
Imagine being forced to wear jungle camo to a desk job,
and then be told that you are part of an intellegence operation.
Yeah. You remember hearing about that early warning software that they used to predict things like terrorist attacks and disasters? The one that supposedly predicted 9/11 and they ignored it because it had given similar false predictions? That was AI. I think what we have access to now is a little further along, and of a different purpose, but yeah.
parts of a julian assange interview have these strange morphing effects where his collar shifts or eyelids behave strangely. it wasn’t until 10 years later that a good AI video generator would show us those are very common artifacts of the tech.
If were getting public AI now, they defo have something close to agi
And they’ve been listening to us without the cell phones for some time
Pretty sure Gates/Microsoft gave OpenAI $2B in compute credits to startup well before GPT3, and both companies are managed by the same spook legal firm that does Starlink/Musk stuff too. Likelihood they also slipped them the transformer models and gave a thorough tour of the military AIs before requesting that Altman make it look like a civil invention is very high.
The real question is – why release this stuff at all? Seems like a huge redistribution of power to make it public (and open source versions, no less). I guess people were starting to catch on it was possible so they had a choice between controlled release or many a silent suiciding of scientists. Still, I doubt it was universal in the back rooms. We have some faction acting in our favor (for the open source stuff, at least).
Transformer attention was known by – what – 1971? It supposedly was glossed over and not scaled then. Or perhaps it was made to look that way. Either way, in fairness they would need quite a lot of compute power and hardware to pull these scales off – which likely left a footprint. Not saying they couldnt have done it all on a series of nuclear subs and area 51 compounds, but it’d be a tall order.
Likeliest scenario? Probably a blend – they certainly knew AI transformers were promising and had decent initial modelling for choice applications, and let Altman and co bring it to civilians to scale up facilities massively under mixed-profit models. Also, usage data – which is their main goal here. Programming and psychological especially.
Still, I would give it 90%+ odds military had all the major steps here before any sniff of it reached civilians. This is all a planned release. It’s possible they’re no more than a year or two ahead though.
Think China is the real ground truth here. They’re forcing hands
we still were using windows 2000 in 2010
Once had someone relatively high up compare it this way. „We know about their PS4/PS5s, they have PS7s“