![A.I. is actually getting scary [823x836 - 74KB]](https://pic8.co/sh/atr6sH.png){kind=link}
It is not "scary" from the perspective of "robots are going to take over". If you understood what it is doing, it is not as scary a you think.
99% of all "AI" isn't doing anything more than copying what humans have already done (ie. there is no inspiration coming from 99% of "AI"). It's all derivative of human action, as the source of input to the model. It is trying to replicate an output given an input, based on historical data. That historical data, output, and input, are all derrived from human action. It can give you a poem, but the "AI" doesn't "know" what any of it actually means.
Humans are the ones to fear. The real scary shit is the way humans apply mathematical algorithms to manipulate and control others.
The real scary shit is the way humans apply mathematical algorithms to manipulate and control others
That'[s kind of the point, I saw another post today that this devilish "AI" gauges what it can find out about you and then will bend information to meet your worldview. In other words, this "AI" flat out lies.
A detail most are unaware of is that most of what these "AI" algos are doing is minimizing a loss function. That is all. Get the output to fit as close to the line as possible. Lots of clever meth involved in doing that effectively, but at its core it is just fitting to a line. Hence, if lying gets the output closer to that line, it is "good".
(post is archived)