In the study of machine learning, they started to see that you can simplify quite complex decisions down to a string of ones and zeros. The computer is not sentient. It has just learnt through multiple repetitions of trial and error, with adjustments. Feedback loops added to complicated processes. It does not understand its own behaviour. Given the ones and zeroes, you can’t extract the knowledge they contain. You can only apply it and see what happens. We are sentient, but even we do not understand ourselves completely. And we have to accept that we can not understand ourselves fully. A lot of our behaviour carries deep knowledge that we can try to interrogate but we might not be able to come to the bottom of. We can constantly be on the path of understanding ourselves through self-awareness, self-search, and self-reflection. That includes recognition of limits, and creating an environment that can cope with lack of understanding and regularly being wrong. That can cope with lack of control.
No comments:
Post a Comment