[FRIAM] Abduction

[uǝlƃ ☣]

Forget my incompetence in ε-machines for a minute. 8^)  They say:

> Take a glass shattering upon impact with the floor. In one temporal direction, the future distribution of shards depends only on the glass's current position, velocity,and orientation. In the opposite direction, we may need to track relevant information regarding each glass shard to inferthe glass’s prior trajectory. Does this require more or less information?

What's the actual answer to that question?  It's not at all obvious that "the future distribution of shards depends only on the glass's current position, velocity, and orientation."  Don't you also need the shape of the glass (tumbler or tulip), the material properties of the glass (leaded?), etc?  Sure, somewhere deep down, there may or may not be some randomness.  And that randomness may be asymmetric.  But this motivating example feels like a trojan horse for some reason.

In particular, with the heralding coin, aren't they baking in the temporal asymmetry by replacing the FIRST 0 with a 2?  That "first" is called "ordinal" for a reason.  But perhaps I'm missing the human experience (ha!) needed to understand how the heralding coin is canonical and to which class of other examples it refers?




On 1/3/19 9:03 AM, Marcus Daniels wrote:
> Possibly of interest..  https://journals.aps.org/prx/abstract/10.1103/PhysRevX.8.031013

-- 
☣ uǝlƃ