Personal post triggered by the coincidence of the year’s end and a round number of posts here: 500.
I started this year with high hopes about the project described in the article GLC actors, artificial chemical connectomes, topological issues and knots. Louis Kauffman wrote in the introduction some unbelievably nice words about graphic lambda calculus:
Even more generally, the movement between graphs and algebra is part of the larger picture of the relationship of logical and mathematical formalisms with networks and systems that was begun by Claude Shannon in his ground-breaking discovery of the relationship of Boolean algebra and switching networks.
We believe that our graphical formulation of lambda calculus is on a par with these discoveries of Shannon. We hope that the broad impact of this proposal will be a world-wide change in the actual practice of distributed computing. Implemented successfully, this proposal has a potential impact on a par with the internet itself.
But in June we got news that the project “Secure Distributed Computing with Graphic Lambda Calculus” will not be funded by NSF, see my comments in this post. We got good reactions, from the reviewers, on the theoretical side (i.e. what is described in the GLC actors article), but fair critics on the IT part of the project. Another mistake was the branding of the project as oriented towards security.
I should have follow my initial plan, namely to start from the writing of simple tools, programs, which should also have some fun side, ideally, but which at least would allow to start the exploration of the formalism in much more detail than the what pen and papers allows. Instead of that, as concerns this project, there has been a waste of time from Jan to Jun, waiting for one puny source of funding before doing any programming work.
I parallel, a better trend appeared, one which I have not dreamed about two years ago: artificial life. During the summer of 2013 I thought it is worthy to try to get rid of a weakness of the graphic lambda calculus, namely that is has one global graph rewrite, called GLOBAL FAN-OUT. That’s why I wrote the Chemical concrete machine article, describing a purely local graph rewrite formalism which later was renamed chemlambda. That was great, I even made an icon for it:
which is made by two lambda symbols, one being up side down, to suggest that writing linear conventions are obsolete. The lambdas are arranged into a DNA like double spiral, to suggest connections with life. (Of course that means I entered in the alife field, but everything about that was so fresh for me. Later I learned about Alchemy and other approaches, mixing either lambda calculus with alife, or rewriting systems with alife, but not both, and surely not the way is done in chemlambda, as abundantly documented in this blog)
Several (personal as well) novelties related to the article. One was that the article appeared on figshare too, not only on arXiv. This relates with another subject which I follow, namely what means of dissemination of research to use, seen that publishing academic articles is no longer enough, unless you want your work to be used only as a kind of fertilizer for future data mining projects.
More about this later.
The initial purpose of chemlambda was to be aimed at biologists, chemists, somewhere there, but apparently, with almost certainly, if a guy does computing then he will not do chemistry, or if he does chemistry then he has no idea about lambda calculus. I’m mean, there are exceptions, like to be part of an already existing team which do both, which I’m not. Anyway, so initially I hoped for interactons with biochemists (I still do, looking for that rare bird who would dare to lower himself to talk with a geometer from a non dominant country, motivated purely by the research content, and who would have the trivial idea to use his chemistry notions for something which is not in the scope of already existing projects).
With Louis Kauffman, we set out to write a kind of continuation of the GLC actors article, this time concentrated on chemlambda, as seen from our personal interests. The idea was to participate at the biggest event in the field, the ALIFE 14 conference. Which we did: Chemlambda, universality and self-multiplication.
Putting together the failure of the NSF project and the turn towards alife, is only natural that I set to write more explanations about the formalism, like this series of 7 expository posts on chemlambda described with the help of g-patterns:
- part I
- part II (definition of g-patterns)
- part III (definition of moves)
- part IV (g-patterns for lambda terms 1st part)
- part V (g-patterns for lambda terms 2nd part)
- part VI (about self-multiplication of g-patterns)
- part VII (an example of reduction) .
This was a bridge towards starting to write programs, that’s for later.
In parallel with other stuff, which I’ll explain in the second half.