Post has attachment
Virus like envelope with Turing machines a better animation than [1], taken from [2]. For TM in chemlambda see [3].

[1] Turing machines grafted on the mock ebola matrix constructor
https://plus.google.com/+MariusBuliga/posts/8E7v7jLAcXh

[2] https://medium.com/@chorasimilarity/how-i-became-a-face-model-for-ted-8ef40e250062

[3] Turing machines chemlambda style
http://chorasimilarity.github.io/chemlambda-gui/dynamic/turingchem.html
Animated Photo

Post has attachment
Complex math can be made so simple with the right point of view. Figure made for explanations of needs [1]. Wait and see, meanwhile enjoy the elegance of symmetry.

[1] https://github.com/chorasimilarity/needs
Photo

Post has attachment
... where I tell the whole story about the script for a #TED talk, very weird but hopefully funny. Then there's the script and the slides.

https://chorasimilarity.wordpress.com/2017/07/18/chemlambda-for-the-people/

Post has attachment
UPDATE: Amazing! I appeared on the list of speakers on CNN, but either:
- the talk was really cancelled for the good after this post you read, and the CNN page updated, or
- this was the last weird thing in a string of really weird events.

The proof is (as long as it'll last, I saved a cached version of the page for proof, for those willing to have one) here:

i2.cdn.cnn.com/cnnnext/dam/assets/170719163454-ted-global-2017-marius-buliga-small-169.jpg

TED talk cancelled. My Chemlambda for the people [1] TEDGlobal 2017 talk was cancelled at the last moment. The first version of the script was different, about Open Science [2] but did not pass. They were interested into chemlambda.

Weird is that I have not asked to give a TED talk, I've been invited to, probably because of the wide audience on the Net.

After the initial proposal, I said to myself "what the heck", I'm not the touchy-feely kind, but maybe I could use this for spreading the word on Open Science validation approach.

No, they wanted chemlambda. OK, if the mention of names of functional programming languages is "insider words" (how could I dare to mention lambda calculus, so I have to use these substitutes, I thought), if GitHub and Figshare are too much, then chemlambda will be. For the people, I can't be dark, right? I can start with some Neuromancer quotes, at least... Then I couldn't stop to write about "last season Netflix flavored meatballs for your doggie" in the line of "the internet can be your pet". Found support in recent results from the Harvard Medical School (the team of Church, what a weird coincidence, right?), where they deposited a 5 frames movie into the genome of living bacteria and then they recovered the movie from there [3].

Interesting challenge, I thought.

Well, what killed it was, it appears, the mentioning of the factorial function. Why, if chemlambda can be used to turn any computer program into a molecule and execute it by random chemical reactions, then I may show this by taking a simple (but not too simple) concrete program, like the factorial (I resisted to take the Ackermann function).

That was too much. Designer medicine, OK, they got it. Turning a real chemical process into a computer program, debug it and then turn it back into a solution (medicine), they could swallow this. But the factorial? What's the use of that? It is an illustration of the "can do anything", it works in both ways, you can turn a program (of your choice) into a chemical version and let Nature run it.

So you'll be spared from a chemlambda TED talk. And me too, they wanted me to come anyway to the show (maybe to put to use some vaccines I don't need? nah, probably consolation) but I don't need this.

I am left with the puzzle that any step which followed the initial invitation was faulty somehow:
- I've been baptized BULIGIA for some weeks, even in the official letter, wtf? it's a 6 letters word, consonant and vowels alternating
- been proposed a too short time for chemlambda (but maybe enough for validation/Open Science)
- with the exception of the comments from the scientific editor, who is a nice guy with a clear mind, nobody used the materials I sent, not even in the...
- rehearsal, where they asked me to use screencapture (were they afraid of my html slides? instead of Powerpoint they expect...), then it turned out that they had no sound (think furious sea waves and wind blowing instead), the animations appeared to them like random renderings with the frequency of 1/s,
- I was not aware of this, they are in the business of rehearsals, not me, but they let me go along mute and blind, was that a set up?

Who cares anyway, it was already very surprising that I received an invitation from an established show in the system.

[1] https://chorasimilarity.wordpress.com/2017/07/18/chemlambda-for-the-people/

[2] http://telegra.ph/Open-Science-is-RWX-science-07-18

Post has attachment
Factorial in edit distance space. Path through edit distance space for a factorial molecule, lisfact_2_mod.mol. There are blue bonds between states at edit distance less than or equal to 22. In red the evolution. What we see: there are many small clumps of states which are far in edit distance one from another (meaning that if we'd want to embed this space into an euclidean one than we'd need a big dimensional one, approx. equal with the number of clumps). These clumps are visited one by one during the evolution. These clumps are almost linear, excepting at the beginning and at the end.
Photo

Post has attachment
Edit distance space for a quine. This is the path in the edit distance space for a random evolution of the 28_quine.mol. It has been made as the Ackermann example of path in the edit distance space [1]. Each node represents a state of the quine during evolution and there are blue bonds for states closer than 12 in the edit distance. The red path is the evolution in time, from one state to the next one. The space is more confined, is connected, there is a central big blob of states which represent say the "healthy" quine states. There are some smaller blobs which are visited by the evolution, but the quine comes back to the central blob. That is a manifestation of the metabolism.

[1] Ackermann walks through edit distance space
https://plus.google.com/u/0/+MariusBuliga/posts/9Fej244AyJz
Animated Photo

Post has attachment
Ackermann walks through edit distance space.To each numbered (1...196) node corresponds a step in the random reduction of Ack(2,2). Then the molecules for the intermediary steps are encoded as "genes" and the edit distance is computed [1]. In the image two nodes are connected by a blue edge if they are at most 22 apart in the edit distance. Finally the red path shows the succession of the steps. Live version [2].


Fun facts: 1) the initial molecule and the result are less than 22 edits apart. 2) the reduction path, as seen in this edit distance space, has lots of big jumps, as seen in the heatmap of the edit distance [1].



[1] Jumps in edit distance
https://plus.google.com/+MariusBuliga/posts/HZCrpSPKde9

[2] http://chorasimilarity.github.io/chemlambda-gui/dynamic/random_walk.html

Post has attachment
Map of the library based on edit distance <= 20, uses [1]. It shows clusters of closely related molecules and trees connecting them. There are free floating clusters as well. Does not look like the result of randomness, nor of blind evolution. The mind behind is revealed :) Some form of the library map will be used for walking through the curated collection.


[1] Heatmap of the edit distance between 427 chemlambda molecules
https://plus.google.com/u/0/+MariusBuliga/posts/R9CfqUeQVkk
Photo

Post has attachment
Duplication as an outcome of conflict This is not a lambda term, tape or quine, it is a molecule which is made to duplicate by a very simple technique, involving the outcome of conflicts between A-FOE and A-L moves. cube_one.mol used, random reduction algorithm.
Animated Photo

Post has attachment
Jumps in edit distance appear during the evolution of a molecule, here my workhorse Ackermann function. In this random algorithm run the molecule takes 196 different configurations till the computation ends. Seen with the "genes" encoding, the evolution organizes into 3 big clusters (hence a 3X3 raster aspect of the distance matrix). I suppose that's because when seen via "genes", the small random rewrites which are done here and there translate into some big transpositions, sometimes. Interesting, beside the fact that the edit distance has its natural problems. UPDATE: see [1] for a representation of the edit distance space and the walk through it during this reduction.

[1] Ackermann walks through edit distance space
https://plus.google.com/+MariusBuliga/posts/9Fej244AyJz
Photo
Wait while more posts are being loaded