Information Theory is the New Central Discipline

Replied to Information Theory is the new central discipline. by Nassim Nicholas Taleb (facebook.com)

INFORMATION THEORY is the new central discipline. This graph was from 20y ago in the seminal book Cover and Thomas, as the field was starting to be defined. Now Information Theory has been expanded to swallow even more fields.

Born in, of all disciplines, Electrical Engineering, the field has progressively infiltrating probability theory, computer science, statistical physics, data science, gambling theory, ruin problems, complexity, even how one deals with knowledge, epistemology. It defines noise/signal, order/disorder, etc. It studies cellular automata. You can use it in theology (FREE WILL & algorithmic complexity). As I said, it is the MOTHER discipline.

I am certain much of Medicine will naturally grow to be a subset of it, both operationally, and in studying how the human body works: the latter is an information machine. Same with linguistics. Same with political “science”, same with… everything.

I am saying this because I figured out what the long 5th volume of the INCERTO will be. Cannot say now with any precision but it has to do with a variant of entropy as the core natural generator of Antifragility.

[Revised to explain that it is not *replacing* other disciplines, just infiltrating them as the point was initially misunderstood…]

Nassim Nicholas Taleb via Facebook

[My comments posted to the original Facebook post follow below.]

I’m coming to this post a bit late as I’m playing a bit of catch up, but agree with it wholeheartedly.

In particular, applications to molecular biology and medicine are really beginning to come to a heavy boil in just the past five years. This particular year is the progenitor of what appears to be the biggest renaissance for the application of information theory to the area of biology since Hubert Yockey, Henry Quastler, and Robert L. Platzman’s “Symposium on Information Theory in Biology at Gatlinburg, Tennessee” in 1956.

Upcoming/recent conferences/workshops on information theory in biology include:

At the beginning of September, Christoph Adami posted an awesome and very sound paper on arXiv entitled “Information-theoretic considerations concerning the origin of life”  which truly portends to turn the science of the origin of life on its head.

I’ll note in passing, for those interested, that Claude Shannon’s infamous master’s thesis at MIT (in which he applied Boolean Algebra to electric circuits allowing the digital revolution to occur) and his subsequent “The Theory of Mathematical Communication” were so revolutionary, nearly everyone forgets his MIT Ph.D. Thesis “An Algebra for Theoretical Genetics” which presaged the areas of cybernetics and the current applications of information theory to microbiology and are probably as seminal as Sir R.A Fisher’s applications of statistics to science in general and biology in particular.

For those commenting on the post who were interested in a layman’s introduction to information theory, I recommend John Robinson Pierce’s An Introduction to Information Theory: Symbols, Signals and Noise (Dover has a very inexpensive edition.) After this, one should take a look at Claude Shannon’s original paper. (The MIT Press printing includes some excellent overview by Warren Weaver along with the paper itself.) The mathematics in the paper really aren’t too technical, and most of it should be comprehensible by most advanced high school students.

For those that don’t understand the concept of entropy, I HIGHLY recommend Arieh Ben-Naim’s book Entropy Demystified The Second Law Reduced to Plain Common Sense with Seven Simulated Games. He really does tear the concept down into its most basic form in a way I haven’t seen others come remotely close to and which even my mother can comprehend (with no mathematics at all).  (I recommend this presentation to even those with Ph.D.’s in physics because it is so truly fundamental.)

For the more advanced mathematicians, physicists, and engineers Arieh Ben-Naim does a truly spectacular job of extending ET Jaynes’ work on information theory and statistical mechanics and comes up with a more coherent mathematical theory to conjoin the entropy of physics/statistical mechanics with that of Shannon’s information theory in A Farewell to Entropy: Statistical Thermodynamics Based on Information.

For the advanced readers/researchers interested in more at the intersection of information theory and biology, I’ll also mention that I maintain a list of references, books, and journal articles in a Mendeley group entitled “ITBio: Information Theory, Microbiology, Evolution, and Complexity.”

Venn Diagram of how information theory relates to other fields.
Figure 1.1 [page 2] from
Thomas M. Cover and Joy Thomas’s textbook Elements of Information Theory, Second Edition
(John Wiley & Sons, Inc., 2006) [First Edition, 1991]
Syndicated copies to:

Twitter Changes Rules on Users. No Auto-Follow. | Kyle Lacy

Replied to Twitter Changes Rules on Users. No Auto-Follow. by Kyle Lacy (kylelacy.com)
(hat tip to ZDNEt and Chris McEvoy for the lead) From ZDNet: “With no notice, Twitter yesterday “pulled the rug out from under its developers” one developer says, by discouraging auto-following and imposing 1,000 person-per-day following limits.” Now… this is not news to me because of the “pulling the rug out from under its developers” thing or the 1,000 person-per-day following limit… The news to me as a Twitter user… I don’t really remember getting a message or alert that the new limits were going to be enacted. I use Twitter on a daily basis. It seems fairly odd that I would not know about the change.

I’m personally glad they’d be implementing something like this and wish they had done it about a month ago. Eventually without any controls the site would have become a waste land. In the spirit of using it as the tool it has become, they needed to implement changes like this as the site scaled up to more and more people. It’s very similar to the changes they instituted in the fall of 2008 when they created a cap of being able to follow more than 2000 people when your own number of followers wasn’t commensurate with that number. As a game theorist, I’m sure that people will somehow find some other way to artificially game the system.

As a separate note, who really wants to waste the time building thousands and thousands of followers when none of them are really going to ever pay attention to you? Yes, it’s great to have a high number, but really what is your ultimate reach? How many people are you engaging?

Syndicated copies to: