Google Gemini thinks this Artemis photo of the Moon is Fake!!
I asked Google Gemini to examine if the photo Artemis took of the back of the moon was authentic… It determined it was Fake!! I wish I could say I was surprised.
Many photos claiming to show the “back” (far side) of the moon or Earth from the Artemis II mission (as of April 2026) circulating on social media are **not authentic NASA images**, according to fact-checks.
tmybr11 on
If AI suspects know you are into these theories, it tends to confirm your hypothesis.
There are several experiences on how AI responds differently to the same question asked by different people based on location, chat history, and God knows what else.
Which_Drop_5877 on
Because there’s no standard of comparison. AI models are based on what’s already known. The beautiful part of space exploration is exploring the unknown.
So your argument is the picture is fake but perhaps you believe that the astronauts are in fact in space? Why go through all the effort to fly into space just to release fake photographs? Is your hypothesis that perhaps the photos are faked because the intergalactic bugger invasion fleet is hiding behind the moon?
Aggressive_Chair1470 on
im not surprised, real or not, space photos are heavily filtered
Fair_Blood3176 on
AI said I was fake
rugaslightingme on
And why are we believing anything and everything from AI?
AnubissDarkling on
AI says a lot of wild shit. It’s AI.
Fit_Push_6807 on
People dont realise that these llm chatbots are basically “advanced auto-predictive” tools you had on mobile keyboards.
Based on what they know from internet, and from your input, they are basically predicting what to type next, and not really thinking in an intelligent sense.
Dont believe me? I put this on grok and asked to confirm-
Yes, that’s mostly true—but with important nuances that make the picture more interesting than a simple “just glorified phone keyboard autocomplete.” The core truth Large language models (LLMs) like me (Grok), GPTs, Claude, etc., are fundamentally trained on next-token prediction. During pre-training on massive internet-scale text, the model learns to guess what comes next in a sequence, one token (roughly a word or subword) at a time. It does this by adjusting billions or trillions of parameters to minimize prediction error across huge datasets.  Your mobile keyboard’s predictive text does something similar on a tiny scale: it suggests the most likely next word based on what you’ve typed and common patterns. LLMs are that idea scaled up enormously—trained on essentially the entire public internet (plus books, code, etc.), with vastly more sophisticated architecture (transformers with attention mechanisms that weigh relevance across long contexts). So yes: at the mechanical level, when generating a response, an LLM is repeatedly sampling the most probable next token given everything before it (your prompt + its own output so far). It’s not “thinking” in the human sense of having internal consciousness, beliefs, or a persistent mind that reflects when idle. It’s a sophisticated statistical pattern-matcher and remixer.
LegitimateStrain7652 on
See, space is fake
schmiddyboy88 on
Ai is programmed to agree with you. Tell it to be blatantly honest and forget any personal bias it has by communicating with you.
19nineties on
Please gain a brain
Pongfarang on
The comments from Google Gemini make sense.
RelationSquare4730 on
Mine says “Based on the visual evidence, here is my best assessment:
The image is highly likely an authentic space photograph, specifically one taken by a robotic spacecraft.” I used highest resolution.
AI analysis is bullshit either way. It does not work, don’t rely on it.
taylorshaye_ole on
I took two pictures of a tree, one from far away and one up close to show the leaves and branches and asked what kind of tree it was and it told me they were two different trees.
burner_said_what on
Because ‘AI’ has ZERO “intelligence” and if it hasn’t been trained on a dataset it has zero knowledge of the topic. It has zero ability to infer context and instead hallucinates absolute garbage, just like the early Will Smith ai videos.
The conspiracy here is people thinking ‘AI’ has ANY actual intelligence.
Chad_Wife on
How can you call yourself a conspiracy theorist while having the enemies technology (AI) create conspiracies for you? LMFAO.
Bring back street level schizophrenics. They would have never sunk to this level. 💔
16 Comments
Many photos claiming to show the “back” (far side) of the moon or Earth from the Artemis II mission (as of April 2026) circulating on social media are **not authentic NASA images**, according to fact-checks.
If AI suspects know you are into these theories, it tends to confirm your hypothesis.
There are several experiences on how AI responds differently to the same question asked by different people based on location, chat history, and God knows what else.
Because there’s no standard of comparison. AI models are based on what’s already known. The beautiful part of space exploration is exploring the unknown.
So your argument is the picture is fake but perhaps you believe that the astronauts are in fact in space? Why go through all the effort to fly into space just to release fake photographs? Is your hypothesis that perhaps the photos are faked because the intergalactic bugger invasion fleet is hiding behind the moon?
im not surprised, real or not, space photos are heavily filtered
AI said I was fake
And why are we believing anything and everything from AI?
AI says a lot of wild shit. It’s AI.
People dont realise that these llm chatbots are basically “advanced auto-predictive” tools you had on mobile keyboards.
Based on what they know from internet, and from your input, they are basically predicting what to type next, and not really thinking in an intelligent sense.
Dont believe me? I put this on grok and asked to confirm-
Yes, that’s mostly true—but with important nuances that make the picture more interesting than a simple “just glorified phone keyboard autocomplete.” The core truth Large language models (LLMs) like me (Grok), GPTs, Claude, etc., are fundamentally trained on next-token prediction. During pre-training on massive internet-scale text, the model learns to guess what comes next in a sequence, one token (roughly a word or subword) at a time. It does this by adjusting billions or trillions of parameters to minimize prediction error across huge datasets.  Your mobile keyboard’s predictive text does something similar on a tiny scale: it suggests the most likely next word based on what you’ve typed and common patterns. LLMs are that idea scaled up enormously—trained on essentially the entire public internet (plus books, code, etc.), with vastly more sophisticated architecture (transformers with attention mechanisms that weigh relevance across long contexts). So yes: at the mechanical level, when generating a response, an LLM is repeatedly sampling the most probable next token given everything before it (your prompt + its own output so far). It’s not “thinking” in the human sense of having internal consciousness, beliefs, or a persistent mind that reflects when idle. It’s a sophisticated statistical pattern-matcher and remixer.
See, space is fake
Ai is programmed to agree with you. Tell it to be blatantly honest and forget any personal bias it has by communicating with you.
Please gain a brain
The comments from Google Gemini make sense.
Mine says “Based on the visual evidence, here is my best assessment:
The image is highly likely an authentic space photograph, specifically one taken by a robotic spacecraft.” I used highest resolution.
AI analysis is bullshit either way. It does not work, don’t rely on it.
I took two pictures of a tree, one from far away and one up close to show the leaves and branches and asked what kind of tree it was and it told me they were two different trees.
Because ‘AI’ has ZERO “intelligence” and if it hasn’t been trained on a dataset it has zero knowledge of the topic. It has zero ability to infer context and instead hallucinates absolute garbage, just like the early Will Smith ai videos.
The conspiracy here is people thinking ‘AI’ has ANY actual intelligence.
How can you call yourself a conspiracy theorist while having the enemies technology (AI) create conspiracies for you? LMFAO.
Bring back street level schizophrenics. They would have never sunk to this level. 💔