hi I’m grant ⍾ and welcome to my blog! Learning to Adapt ⚶ is a blog about learning– the biological kind! (sorry, machines).
Who am I
I’m grant ⍾. I studied bioengineering and worked in biotech for a few years, helping build microscopes and look at molecules. A funny confluence of factors gripped me: I was working with inverse design software (Zemax), trying to understand concepts in ML, studying animal physiologyfor the MCAT.. i know as an engineered system, and reigniting a passion for computational biology. I started thinking about the kinds of systems that optimization produces, how to better design a system so as to be optimized, and the sorts of optimization algorithms that exist.
What I’m studying
Complex systems are systems with many components where each part behaves according to simple, predictable rules yet the whole is unpredictable, hard to describe, and gives rise to novel features. Alternatively, they are systems that you could find yourself talking about at lengths, even when you try to keep it short. We find ourselves in a world enriched with complexity and have managed to build some complex systems ourselves.
Biology has built a number of complex systems and has done so for billions of years. These systems are some of the most persistent things we know, and I believe that by studying how they work, how they’re built, and how they change, we can better understand how to build persistent systems of our own. I’ve long had a fascination with computational biology and watching intricate structures emerge from simple rules. Evolution is a tinkerer, not a designer, but I have a hunch that it’s smarter than we give it credit for. This blog is dedicated to examining the irreplicable systems that it has produced, and for exploring the higher order phenomena that give it a bit of a speed boost.
Learning is the process by which behavior changes in response to the acquisition of knowledge about the world, encoded in memory (Kandel). From Old English leornian: to learn ← Proto-West Germanic *liʀnōn ← Proto-Germanic *lizaną ← PIEProto-Indo-European, the reconstructed parent to numerous languages across Europe, the Middle East, and south Asia. *leys- (to trace, to track). (Wiktionary)
An adaptation is “a trait that increases the ability of an individual to survive or reproduce compared with individuals without the trait” (Freeman). From Latin adaptō: to fit, adjust, modify, via ad- (towards) + aptō (to fit, to make ready) ← PIE *h₂ep- (water; to join, attach, fasten, fit). (Wiktionary)
What I’m reading:
Here’s a list of books that I’m reading from at the moment.
Principles of Neural Science; Kandel, Schwartz: the brain bible
Developmental Biology; Barresi, Gilbert: an overview of the toolkit that our cells use to build our bodies from the ground up
Physical Biology of the Cell; Phillips, Kondev, Theriot, Garcia: a rich, digestible resource for conceptualizing biophysics. Accessible for physicists and biologists
Information Theory, Inference, and Learning Algorithms; MacKay: theoretical basis for the mathematical and algorithmic side of learning
Statistical Mechanics, Entropy, Order Parameters, and Complexity; Sethna: for understanding what statements we can and can’t make about systems with N»1
Molecular Driving Forces: Statistical Thermodynamics in Biology, Chemistry, Physics, and Nanoscience; Dill, Bromberg: statistical mechanical formulations of various useful physicochemical systems relevant to biology. Really helps with polymers
Evolutionary Neuroscience; Kaas: my primary reference on evolution of the vertebrate nervous system
Vertebrate Life; Pough, Janis, Heiser: great for understanding the vertebrate lifestyle and how our adaptations enable and facilitate it
Biology of the Invertebrates; Pechenik: for comparison against vertebrates
Introduction to Systems Biology; Alon: models for understanding how molecules handle information
Wikipedia: the whole thing. support our queen
How I’m doing it
I’m going to be showcasing a couple tools. Python mostly.
Disclaimer
This is a blog (not a scientific journal) and I post posts (not papers)! I can be wrong, I haven’t read all literature yet, and I love to speculate. I also like to use teleologic language as shorthande.g., 'the cells learned to communicate' instead of 'the cells who possessed traits allowing for signal transduction due to random genetic variation were more likely to pass on their alleles' and generally elide the bulky language of evolutionary biology.. sorry. Please feel free to contact me at grant@learning-to-adapt.io with any corrections, concerns, or feedback!