Speaking of robust
…chimes with @GaryMarcus on AI
…data-only path very iffy
…many Tech Titans beguiledhttps://twitter.com/seanmcarroll/status/1001485745062739968 …
-
This is completely irrelevant and uninteresting. The single parameter is a real number of arbitrary precision: in other words, an encoding of the data. Occam’s razor is about bits of information, not the number or parameters
-
-
"constraints…enforced independently [of]…data"—>data alone often misleads…my 2 c causal-model-free "fit" often unrobust
-
Causal models and non-causal models both depend critically on the choice of representation. If a bad representation is employed, both will fail miserably. Because not all variables can be manipulated (via a Do()), causal models depend even more on having the right representation
End of conversation
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.