[FRIAM] Entropy RE-redux

[steve smith]

Nick -

I'm surprised you gave over to ALL-CAPS to distinguish your words from 
others... it was shockingly difficult or me to read through 
carefully...  some call this "shouting" which is part of the effect to 
me also, but more key is the typography and readability of it...

That said, I did what I had to to make it "parseable" with my own 
sensoria/sensibilities and found it to be quite rich epistemically.

I share your apprehension/feeling/sensibilities on the topic — that 
entropy (or motive, or any number of these second-order constructs) gets 
misrepresented when we try to pin it to an instant, as if we could find 
it /in/ the moment instead of across the differential.

Your “Newliebzian category error” notion made me smile - there’s 
something compelling about how we keep building our most useful tools 
from fictions we then forget are fictions. Even that might be too 
conventient? It might be a kind of necessary fold in how cognition works 
- using what’s locally graspable to navigate what’s only ever apparent 
in change or flow. There is a larger insight I think about metaphor in 
general, but in this case time/causally grounded metaphorical target 
domains?

I also appreciate your connection to how motives are misread — 
interiorized and moment-bound when really they emerge from comparative 
arcs of behaviour. That strikes me as roughly what the nervous system 
does: extract regularities from dynamics, then offer up something /like/ 
a motive as a usable handle. Induction is fallible, but it’s also our 
only working interface. Not because it guarantees truth, but because it 
lets us stay involved.

In that light, entropy - or motive - might be less about what /is/, and 
more about what is /minimally sufficient to expect/? From a free energy 
perspective, the nervous system isn't trying to uncover objective causes 
but to *minimize surprise* through actionable pattern recognition - 
constructing motives, like entropy, as provisional summaries of systemic 
change. It isn’t so much that they’re true, but that they *reduce the 
expected divergence* between model and world just enough to stay viable?

Fascinating that this iterative/parallel /satisficing/ seems to converge 
on something yet more absolute (objective truth)?  Is this what 
Transformers (for the ML wonks in the crowd) do?

- Steve

On 6/16/25 10:38 AM, Nicholas Thompson wrote:
> HERE IS A THE PARADOX AS I NOW UNDERSTAND IT.  IS ENTROPY A STATE 
> VARIABLE/  WELL, YES AND NO.  IT IS NOT IN THE SENSE THAT YOU CANNOT 
> CALCULATE IT FROM INSTANTANEOUS STATE VARIABLES, IN PERHAPS THE WAY 
> THAT YOU CANNOT CALCULATE ACCELERATION FROM AN INSTANTANEOUS 
> MEASUREMENT OF LOCATION AND VELOCITY.  ACCELERATION AT INSTANT IS A 
> CATEGORY ERROR, A MATHEMATICAL FICTION.  BUT, STRANGE TO DISCOVER, 
> THAT ALL PROGRESS IN MATHEMATICS, BEGINNING WITH THE MATHEMATICAL 
> LINE, PERHAPS, IS THE RESULT OF THE EXPLOITATION OF SUCH FICTIONS.  
>  fOR THE MOMENT LETS CALL THEM NEWLIEBZIAN CATEGORY ERRORS.  iF MY 
> CURRENT INTUITION IS CORRECT AND ENTROPY IS NEWLIEBZIAN CATEGORY 
> ERROR, THEN IT FOLLOWS THAT THE REASON i AM CONFUSED BY IT IS BECAUSE 
> PEOPLE KEEP ATTRIBUTING IT TO INSTANTS AND PLACES WHEN IT KNOWN 
> THROUGH THE COMPARISON OF DYNAMICS AND ARRAYS.  BEWARE BECAUSE THIS IS 
> A VERY FAMILIAR CONFUSION TO ME, BETWEEN MOTIVES, WHICH ARE THOUGHT OF 
> INTERIOR INSTANTS, WHEN THEY ARE ACTUALLY KNOWN THROUGH THE COMPARISON 
> OF DYNAMICS AND ARRAYS OF BEHAVIOR.   THE WHOLE PURPOSE OF OUR NERVOUS 
> SYSTEM (THE STUFF IN SIDE)  IS TO MAKE COMPUTATIONS BASED ON THE 
> DYNAMICS AND ARRAYS OF BEHAVIOR AND APPLY THEM TO THE NEXT APPLICABLE 
> INSTANT.  PHILOSOPHERS CAN SAY ALL THE WANT ABOUT THE FALLIBILITY OF 
> INDUCTION, BUT IT IS ALL WE GOT. 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://redfish.com/pipermail/friam_redfish.com/attachments/20250616/a43c0364/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: OpenPGP_0xD5BAF94F88AFFA63.asc
Type: application/pgp-keys
Size: 3118 bytes
Desc: OpenPGP public key
URL: <http://redfish.com/pipermail/friam_redfish.com/attachments/20250616/a43c0364/attachment.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: OpenPGP_signature.asc
Type: application/pgp-signature
Size: 840 bytes
Desc: OpenPGP digital signature
URL: <http://redfish.com/pipermail/friam_redfish.com/attachments/20250616/a43c0364/attachment.sig>