0
$\begingroup$

This refers to a system described in a book by Nick Lawrence titled "Correlithm Object Technology". The author coined the term "correlithm" as a combination of "correlation" and "algorithm".

The Correlithm Objects ("COs") described in the book are, in the author's view, "primary data tokens" in biological neural systems and are central to "all high-level data representation, storage, and manipulation". In addition to this the author describes the COs as "important mathematical objects of the statistics of bounded, high-dimensional spaces".

Considering these descriptions it seems like these COs could be useful for AI. What I was curious about is whether they are actually used anywhere in the industry, and if so, in what kind of situations?

$\endgroup$
3
  • 1
    $\begingroup$ Per a talk by Dr. Lawrence, they are heavily used by the U.S. Department of Defense. The specific application of math is supposedly non-obvious and highly useful. $\endgroup$ – Grault Oct 16 '18 at 20:25
  • $\begingroup$ @Grault, would you have a reference (URL or info on where it was given) to that talk by any chance? Thanks $\endgroup$ – user100487 Dec 24 '18 at 15:34
  • $\begingroup$ Unfortunately no. It was an internal talk for Bank of America at DevCon 2018. $\endgroup$ – Grault Jan 14 '19 at 0:04
2
$\begingroup$

Looks like lots of marketing around a specific case of what everyone was already doing to sound bigger than it is. So, are people implementing mathematical frameworks in software to extract and use the structure within data? Absolutely. Are people using the word 'Correlithm' when doing it? Not as far as I've seen.

$\endgroup$
1
  • $\begingroup$ I'll add that I've never heard of "correlithm" before. Also, no one I know would use that word. Correlation is already a tremendously abused word. $\endgroup$ – Jim Apr 1 '18 at 18:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.