4 ms·

> If you can explain the mechanism behind the manifestation then you can understand why it isn't real. I watched, with my family, the Algorithms documentary av

by moosey 10y ago

> If you can explain the mechanism behind the manifestation then you can understand why it isn't real. I watched, with my family, the Algorithms documentary available on Netfix. While not perfect (but very, very good), it certainly did a great job of explaining in simple terms some algorithms and how they relate to the world. One of the most interesting parts, in my opinion, was the admittance during the section where they were discussing the XBox Kinect. It was all done through some machine learning algorithms, and impressively, they confessed to not being completely certain how the code actually worked. You could argue that this means that they cant explain the mechanism behind the manifestation. Now, of course, this is a massively simplified version against the near-future Westworld narrative. However, it does call into question whether we would understand the mechanism of something so advanced. Also, applied to humans, if we end up understanding the mechanisms behind the brain, then does that make us less real? They point out in episode three I think that one of the 'Hosts' are experiencing physical discomfort while going through diagnosis. If such a thing were coded, and it was made possible to respond to it (I would argue that the best possibility for making this possible would be a learning algo, similar to the Kinect analogy), then how would that be different from the very physical and chemical responses that a human has to experiencing suffering? It's hard for me to make a decision in either direction. It is a complicated and nuanced problem. It reminds me of the short story "They're Made Out of Meat" [1], and could just be the result of human-centric (or in this case, biologically-centric) thinking. 1: http://www.terrybisson.com/page6/page6.html http://www.terrybisson.com/page6/page6.html

solatic 10y ago
There's an axiom of software engineering that any sufficiently large system (or, I suppose, algorithm) is incomprehensible by a single person. That's to say, a single person may be a high-level architect and understand how the high-level modules work together, but he won't understand how the inner workings of those modules work, especially if there are multiple layers to those modules, both bureaucratically and technically. And somebody who busies themselves understanding and working on the lowest levels, won't have the time to work on the higher architecture, nor would it interest him, as long as boundaries are well established and respected. It's not that far of a stretch at all to claim that strong AI would be similar. At some point, everything worked on by today's AI researchers becomes some refined lower-level module, and tomorrow's researchers will use them as building blocks for the architecture for strong AI. Those lower-level modules will have coefficients derived from massive training sets, and tomorrow's researchers will simply understand that those coefficients work, without any real understanding of how those coefficients were derived. And that's OK.

by moosey 10y ago

solatic 10y ago

There's an axiom of software engineering that any sufficiently large system (or, I suppose, algorithm) is incomprehensible by a single person. That's to say, a single person may be a high-level architect and understand how the high-level modules work together, but he won't understand how the inner workings of those modules work, especially if there are multiple layers to those modules, both bureaucratically and technically. And somebody who busies themselves understanding and working on the lowest levels, won't have the time to work on the higher architecture, nor would it interest him, as long as boundaries are well established and respected. It's not that far of a stretch at all to claim that strong AI would be similar. At some point, everything worked on by today's AI researchers becomes some refined lower-level module, and tomorrow's researchers will use them as building blocks for the architecture for strong AI. Those lower-level modules will have coefficients derived from massive training sets, and tomorrow's researchers will simply understand that those coefficients work, without any real understanding of how those coefficients were derived. And that's OK.