To all of you AI-doubters,OpenAI just published a paper providing proofs to ten frontier level mathematics problems across different fields. https://cdn.openai.com/pdf/ten-proofs-oai.pdfNo, I do not personally understand them. Nobody on this board will. These are the kinds of problems that one has to dedicate their entire life to a single narrow field to get to the point where they can take a crack at it. Each one has maybe a dozen people in the world who are qualified to work on them.It is safe to say that AI is officially superhuman at mathematics.This is what I've been wanting from AI all along, and I'm ecstatic that it's finally here. So many fields of study have grinded to a halt over the past century. When was the last time that physics was truly turned on its head like what was happening every year in the 1920's? Humans have just about gotten to the limit of what we can do, and all this stagnation will finally come to an end. This is worth the internet being filled with slop, this is worth endless data center construction, and this is worth enormous CapEx on the part of big tech.Progress is cool, and its totally justified how obsessed society is with this technology right now.
AGI WILL be used by those in control of it to further their petty human interests. typical human interest does not involve the betterment of less fortunate people.the funniest part is that those that AGI will serve are not even the researchers that did EVERYTHING to create it, but those that inputted money at the lottery machine we call R&D.I hate, hate, HATE our species.
>>540069854>AI solved a hyper-specialized problem that in no way helps the world and was made up by mathematicians to justify their salarywake me up when it makes it so everyone can be a neet
It's still predicting tokens retard, it simply has layers of reasoning to refine the prediction process into more likely results
>>540069854Now ask GPT how to open Hormuz strait.
>>540070084The water in your body could be helping a datacenter right now, buddy. Better think hard about how you're going to justify your existence.
If the technology was useful you wouldn't tired to post this. The market would just adopt it.AI models are being used by companies but it's mostly chinese models that are cheaper , nobody needs open AI at their current price
>>540070032>typical human interest does not involve the betterment of less fortunate peoplenor does mine!Who cares WHO it serves? It will discover math and science that humans could not, create things that humans could not, and that is all I want.
>>540069854Ill be impressed when AI actually shitstomps general relativity and we can stop pretending jewish misdirection about the multidimensional nature of the universe is canon
>>540070341>create things that humans could notAll if it likely to fail because reality is far more chaotic than a clean sterile and "perfect" over engineering blueprint or computer environment
>>540070129So? It's working. Any complex process seems simple if you get so disingenuously reductive about it.>>540070162The water problem is extremely exaggerated. One shower uses more water than any number of prompts that you individually could realistically hope to make in a day unless you were deliberately trying to overload it.>>540070274I'm posting about it because I think its cool, and I want to discuss it. But yes, open chinese models are also very cool and only ~3 months behind in progress. Definitely the better experience to use, if you have the means to run one.
>>540069854Has it finally proven P equals NP? If not then everything else is a waste of time
AI needs to stop replacing future engineers and PhD candidates. This is a genocide on the third world
>>540070590>disingenuously reductive>disingenuouslyLol, lmao
