My explanation of Judea Pearl's causal calculus: "If correlation doesn't imply causation, then what does?" http://www.michaelnielsen.org/ddi/if-correlation-doesnt-imply-causation-then-what-does/ …https://twitter.com/hardmaru/status/997987139726921728 …
-
An explanatory theory. Not 'data'.
Implicitly, there's some of this in the causal calculus - explanatory theories must be where the underlying graphs come from. But it'd be nice to put these two points of view properly together.
-
-
This is my biggest concern so far, reading his book. It looks like he first graphs out what the proposed causes are, i.e. he starts with an implicit explanation of causation but only draws it as a simple graph. This is just part of the background knowledge of the problem.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Data then backs up the theory
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
This Tweet is unavailable.
-
-
I did a recent blog post comparing Deutschian epistemology and Pearl's Causal Calculus. I fear I may have overplayed the overlap between the two in my exuberance. But the bottom line is that Pearl simply shows that once you have an explanation it's usefulhttps://www.mindfiretechnology.com/blog/archive/the-importance-of-asking-why/ …
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I don't see how this approach will lead to AGI. Because it still requires a human to first come up with an explanatory theory to get the process started. i.e. all the creativity is still in the human, not the AI. Or so it seems to me so far.
- 2 more replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.