<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body>
<p>glen wrote:<br>
</p>
<blockquote type="cite"
cite="mid:2d2c2b2e-0d97-4955-a2f0-7dc901115321@gmail.com">Yeah,
there is a distinction between structures that obtain from
anastomosis (that are now in/near some steady state) and
anastomosing/anastomizing, the process by which the structures are
formed. Maybe one could say that Jon's observation is about
post-anastomizing anastomotic code as opposed to anastomizing
"code". But code is data and data is code. Anastomizing code would
have to be anastomizing some other structure (e.g. copyleft stuff
puncturing intellectual property norms). But what anastomizes code
is the code creator/extruder, including prcesses like humans,
[semi]automated things like LLMs, genetic programming algorithms,
or especially hackers attempting to exploit the weak points.
<br>
<br>
Unfortunately, the word "anastomotic" can refer to either
structure that obtains or the process by which it's obtained.
Stupid English.
<br>
</blockquote>
<p>I once studied Latin, Greek, and Esperanto (alongside the myriad
goofball CS languages of the time) imagining that somehow one or
another would be more better for unambiguous speech. On one hand,
it made it easier for me to connotize (verbize that noun!) more
English words with various etymologies. On the other it also made
me more sensitive to the nuances which have me using
compounded/superposed/aggregate words and mal-understanding words
other have a common use for which I then hairsplit into
oblivion. <br>
</p>
<p>Or maybe my skills are more like those which drove Yogi Berra's
malapropisms... who knows?<br>
</p>
<p>Every day I am more and more aligned ("think I believe that what
I heard") with "what I think you said" about not believing in
communication (or somesuch)? <br>
</p>
<p>In CS/information theory <i>marshalling</i> and <i>serializing</i>
and <i>de-serializing</i> are all great "engineering" ideas which
have utility but perhaps they only create the illusion (maybe also
strongly typed languages and closure and ... ?) that what we are
saying/parsing is what is meant and it is unambiguous and all
context is carried. These tools might help with more casual
verification, but I'm not sure they are the magic bullets they are
often sold as?<br>
</p>
<p>This is one of the "beautiful" takeaways of Michael Levin's work
(IMO), that no matter how necessary and even elegant shite like
DNA might be, it is not *even* close to sufficient? <br>
</p>
<p>- Steve<br>
</p>
<br>
</body>
</html>