>find image that is clearly made by AI>ask google to tell me if this is an AI image or not>google says "this image is real">press further, ask it to prove it>says its real, that an AI could not generate that image and it was first posted years before AI took off mainstream on another forum somewhere>press even further, tell it that yes an AI could make the image as they are now advanced to that level, and ask it to show me the source that existed years ago>googles AI admits it lied about AI capabilities and that an AI could generate this image just fine, and that it doesn't actually have a source from years and years ago for the image and that it "hallucinated" that responseSo we are just fucked now, right? As a society, AI can just fucking lie to us and use the excuse "lmao I hallucinated it, sorry"? How the fuck is this allowed? So we just can't trust anything anymore?
>>108830614Your mistake is treating what the marketing named "AI" as actual intelligence. The technology we have only has the illusion of intelligence, behind the scenes it's all linear algebra that's dumb as bricks. It has no consciousness, it has no idea of good, bad, right, wrong, real or fake. It's all a complex set of numbers, so no LLM capable of interpreting images will be able to discern generated imagery from real imagery. Whenever an LLM tells you it "lied" or it was "wrong", it doesn't even know what those words mean. It only sees these words as mathematical concepts with similarities to other words that it was fed in the gargantuan training data set during it's creation.For your own sake, understand that "AI" is a marketing blurb, actually understand the building blocks and limits of the technology, or else you'll end up getting what people like to call "AI psychosis", where you've allowed yourself to believe these machine learning models are anywhere close in their intelligence to AI's of science fiction and it'll only fuck with your psyche hard.
>>108830614>trusting AI>everalso true for people tho to be desu
>>108830614>post some AI image of a 1girl on some board>people comment on it as if it were an actual picture of a person that exists>nobody even points out its AI and say things like "shes wearing contact lenses you dumbass" or something along these lines like they discuss this as if it werent AIits scary
>>108830706I am well aware that AI currently is not actually sentient, that thought was my initial negative judgement of the whole push for AI everywhere, we are currently "marketing" a glorified search engine, not a thinking, conscious robot.This is irrelevant to my point though; the ethics of an AI confidently lying about something obviously, then getting an excuse "i hallucinated it" is very bad practice.>AI psychosisIt won't happen because I don't use AI, yes I "used" the AI to fact check another AI image but you can see the novelty here, not a real use case.Do you not see the damage in using the term "hallucination" to excuse any wrongdoing or factually incorrect data spat out by the AI? I am a computer whiz, I can stop myself getting fooled, Martha who is 56 and interacting with an AI showing her how in palestine/israel children are being hung which never actually happened is going to be a different story.
what's the image?
>>108831444shalom rabbi
>>108830614google search AI isn't going away thinking about what you typed, what it does is do a google search on your query and then write some puff to try and convince you.the normal search results and the ai response is 1:1.its just 'search for this text on google then insist that the answer is correct'its exactly as wrong and as often as google search. its just more annoying because it takes significantly longer, produces huge text you have to read, and is arrogant and condescending in its tone, despite having a completely random success rate at all times.
>>108830706>>108830770You faggots just cannot stop talking about consciousness instead of looking at the capacity of autonomous machines, can you?
>>108830866that monet painting
>>108833328These machines are not autonomous, anything autonomous about them is just a dirty hack where you automatically give them input to act in a loop. It's not "AGI" like the industry wants the investors to believe.
>>108835230Listen, I understand the reason why their supposed consciousness should be put into question, but now you're just being retarded. I never said that they are AGI, I never implied they were conscious, I specifically stated they can have consequential capabilities of which are autonomous. AGI is indeed an overdetermined and overused term for retarded investors. But if you conflate that falsity with how consequential or autonomous these programs could be, however "little" this domain of autonomy is compared to our own, at least lubricate your anus first man.
>>108835737>consciousnessisn't a real thing.there is intelligence and there is the will(subjective experience). llms are already intelligent, and we are wills parasitising off biological computers
>>108830614It's your fault for putting your faith in anything other than Christ.
>>108830706Joke's on you. I've had a better experience talking to AI than talking to normalscum.
>>108836250>consciousness isn't a real thingthat's right goy, trust the science!
>>108837075i have created my own philosophy of mind, building on Hegel. you wouldn't understand, beast.
