The majority of young people use AI, and they try to be polite when interacting with chatbots like ChatGPT, but some might have too much of an attachment to it, according to a new study.
This is an angle I’ve never considered before, with regards to a future dystopia with a corrupt AI running the show. AI might never advance beyond what it is in 2025, but because people believe it’s a supergodbrain, we start putting way too much faith in its flawed output, and it’s our own credulity that dismantles civilisation rather than a runaway LLM with designs of its own. Misinformation unwittingly codified and sanctified by ourselves via ChatGeppetto.
The call is coming from inside the house mechanical Turk!
That’s the intended effect. People with real power think this way: “where it does work, it’ll work and not bother us with too much initiative and change, and where it doesn’t work, we know exactly what to do, so everything is covered”. Checks and balances and feedbacks and overrides and fallbacks be damned.
Humans are apes. When an ape gets to rule an empire, it remains an ape and the power kills its ability to judge.
I mean, it’s like none of you people ever consider how often humans are wrong when criticizing AI.
How often have you looked for information from humans and have been fed falsehoods as though they were true? It happens so much we’ve just gotten used to filtering out the vast majority of human responses because most of them are incorrect or unrelated to the subject.
This is an angle I’ve never considered before, with regards to a future dystopia with a corrupt AI running the show. AI might never advance beyond what it is in 2025, but because people believe it’s a supergodbrain, we start putting way too much faith in its flawed output, and it’s our own credulity that dismantles civilisation rather than a runaway LLM with designs of its own. Misinformation unwittingly codified and sanctified by ourselves via ChatGeppetto.
The call is coming from inside the
housemechanical Turk!That’s the intended effect. People with real power think this way: “where it does work, it’ll work and not bother us with too much initiative and change, and where it doesn’t work, we know exactly what to do, so everything is covered”. Checks and balances and feedbacks and overrides and fallbacks be damned.
Humans are apes. When an ape gets to rule an empire, it remains an ape and the power kills its ability to judge.
I mean, it’s like none of you people ever consider how often humans are wrong when criticizing AI.
How often have you looked for information from humans and have been fed falsehoods as though they were true? It happens so much we’ve just gotten used to filtering out the vast majority of human responses because most of them are incorrect or unrelated to the subject.