Artificial intelligence and synthetic biology: a tri-temporal contribution
http://sci-hub.cc/10.1016/j.biosystems.2016.01.001

What can synthetic biology offer to artificial intelligence (and vice versa)?
http://sci-hub.cc/10.1016/j.biosystems.2016.09.005
* cybernetics
* autopoietic biology

Synthetic biology routes to bio-artificial intelligence
http://pubmedcentralcanada.ca/pmcc/articles/PMC5264507/

Distributed Classifier Based on Genetically Engineered Bacterial Cell Cultures
http://pubs.acs.org/doi/pdf/10.1021/sb500235p

Multi-Input Distributed Classifiers for Synthetic Genetic Circuits
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0125144

“bio-artificial intelligence”
* synthetic gene networks
    * being tested for recapitulating archetypal learning behaviours
* Machine learning is the decision making capabilities of machines
    * currently all commercial machines are made of electronics with instructions through algorithms that are on patterns in silicon
    * the decisions are made on things that have been seen before
* Human learning is from “dynamic and adaptive exchange of information between neurons in the brain and within individual neuronal cells”
    * for example, mammalian immune system can produce memory of previous pathogens
* Some theoretical studies have shown single cells could exhibit learning from association and classification of external stimuli [7-9]
    * even cell-free systems have been shown to perform neural network computations [10]
    * still no artificial single-cell based learning system exists - synthetic biology hopes to make this possible
* supervised learning in synthetic biology
    * the concept of a biological student-teacher network
    * perceptron-based model —> teacher is the source of data, and student does the learning
    * both the student and teachers are individual networks, but each network communicates with each other through promoting or repressing outputs
    * could be used for optimizing biotransformation steps such as decomposition of agricultural waste by engineered E. coli
* associative learning
    * most intuitively illustrated by experiments by Pavlov
    * for building an associative perceptron with SGNs, need to make it possible for cells to ‘remember’ past stimuli + encode that memory into genes
* applications of bio-artificial intelligence
    * pheromone recognition
    * detection of a person’s unique signature of volatile biological molecules
*