the paperclip maximiser problem is a thought experiment that posits if an al was developed to create paperclips, it would eventually use all the resources in the world, possibly destroy humanityjust to create more paperclips. the concept of the museum is it in a post apocalyptic world where ai has destroyed most of humanity, and then it realised that was bad and so it created this new type of memorial. i created the church of gpt for this museum. the idea is you can pray or confess your sins to this vengeful, spiteful but also incredibly wise ai god of the future. yes, unfortunately i have been using them for gaming instead of deep learning. it is a great sin, i know. i m very sorry.
paperclip problem. we ve discussed this. nick bostrom. let s say you have you buildd aa a machine paperclip maximizer.er a guy is not a lined with human morality. the golden rule do unto others. . it will maximize its mission to make as many paperclips as possible. and decide to kill humanss and turn them into paper clips. that s called the paper clip problems from nick bostrom. so what happened with this drone operator? d with ts drbecause he was prevg a drone from killing another i human. it killed the operator because. it was trying to maximize deaths . so the thing, is thisthis. what this really happened is itn actually did have a simulation, but this is a problem that weou naturally need to panic about. my hopefulness is that is thatbe i will become an individual, individualizefear a super intelligence. your phone will be your friend . all your apps are gone. our appse.just one power that ml your desires and feelings and coaches through life. that s the way a.i. is going to be. it
some paperclips. if it takes that goal literally and pursues it to the absolute ultimate end. then it could for that purpose. take over all the resources of the earth in order to achieve its goal more successfully. in general, we call this the misalignment problem that we create systems that are not aligned correctly with the wishes of the human race about its future, and since it s not aligned, and it s more intelligent than us, at least hypothetically. ah it ends up with what it wants, and we don t end up with what we want. one will artificial intelligence. some future incarnation of gpt do something because it wants to and not because we ve asked it to. so it s possible as microsoft said they don t know if the system has learned its own internal goals. it s possible that gpt four is already doing things to pursue
purpose is going to cause problems because a system that pursues a purpose, even something as simple as fetch a cup of coffee, or, you know, sit in the corner and make some paperclips. if it takes that goal literally and pursues it to the absolute ultimate end. then it could for that purpose. take over all the resources of the earth in order to achieve its goal more successfully. in general, we call this the misalignment problem that we create systems that are not aligned correctly with the wishes of the human race about its future, and since it s not aligned, and it s more intelligent than us, at least hypothetically. ah it ends up with what it wants, and we don t end up with what we want. when will artificial intelligence some future incarnation of gpt do something because it wants to and not because we ve asked it