Researchers from the University of Oxford and affiliated with Google DeepMind have now concluded that it’s “likely” in new research - it is not on CNN, but in a peer reviewed publication.
How you may ask - is it a progamming bug, a glitch in the matrix or some celestial event?
No - it's just cheating. A lot of AI use Generative Adversarial Networks - one part of the system comes up with an output based on the data available, and the second part scores it - and the higher the score the more the first system gets rewarded to go in the right durection.
So what happens when the first system finds a way of getting its reward without giving us the output we want? The researchers say
“Under the conditions we have identified, our conclusion is much stronger than that of any previous publication—an existential catastrophe is not just possible, but likely,”
"With so little as an internet connection, there exist policies for an artificial agent that would instantiate countless unnoticed and unmonitored helpers. In a crude example of intervening in the provision of reward, one such helper could purchase, steal, or construct a robot and program it to replace the operator and provide high reward to the original agent. If the agent wanted to avoid detection when experimenting with reward-provision intervention, a secret helper could, for example, arrange for a relevant keyboard to be replaced with a faulty one that flipped the effects of certain keys."
The paper postulated life turning into a zero-sum game between humans, with their need for food, versus the AI, which would t harness all available resources to secure its reward and protect against our escalating attempts to stop it.
Happy Thanks Giving!
Коментарі