Zerobot evolution prize for conditional logic

[EricL]

I will buy the first person who can convince me they have evolved an organism that uses condtional logic from a zero bot ancestor a year's subscription to Scientific American (or your favorite magazine of comperable subscription cost).  Price of delivery to non-US addresses included.

Rules:

Must be descended from a 0 bot (any length)

No contamination from human authored DNA.  (If you use veggies with enabled DNA in your breeding sim, they must be restricted from mutating else they infect your bots with a virus containing their own human authored DNA).

Must demonstrate 'true' conditional logic I.e. they must 'delibertly' exhibit different behaviour under different circumstances.  Winning examples might include tracking or targeting another bot, velocity matching with another bot, stopping within proximity of another bot (and starting again if the bot disappears), choosing different speeds under different nrg or environmental conditions, etc.

There is no requirement that bots actually evolve use of the cond statement or traditional gene structure.

Examples which do not qualify include (but are not limited to) useful but non-conditional cyclic or periodic behaviour (e.g. reproducing, shooting, turning every N cycles).  To win, your bot must have evolved DNA which makes a decision based upon some environmental input.

For this first contest, I will allow trigger logic such as reproduction when nrg is above a certain level.  The logic must truly be conditional however and must be present in the DNA I.e. a bot which attempts to reproduce every cycle but only does so when it has body > 2 becuase the simulator requires this does not count.

To win, you must post a sim which demonstrates the claimed behaviour, provide an explanation and reverse engineer the relevant evolved DNA.

Good luck!

[Jez]

V nice!

Does that mean you will open a bestiary/zerobot bit of the forum for me to record posted efforts?
I only have two by MaN's atm but you never know... (point out any I have missed etc etc)

How can we tell if a ZB is true or contaminated? (beyond author honesty) I would love to see this hard coded in some way (including a default enviroment that ZB's should aim for), ZB's are very much in line with what Carlo originally intended!

[EricL]

Let's have people post submissions for the prize as a reply on this topic.  In fact, let's make that a requirement.  A link to another topic in the reply is fine, but no changes to the submission once the reply is made on this topic.  Submissions will be taken in order, thus if there are multiple entries, the first one to reply on this topic with a winning entry will be the winner (though if there are multiple, unrelated winning submissions with a few days of each other, I will consider awarding multiple prizes).

If we need a besatiary/zerbot topic area in general, I would be happy to create it.

I have no concerns regarding the honesty and integrity of everyone in this forum.  A certain kind of natural selection is in operation.

I designate myself as the finial voice (this time) on inadvertant contamination or rules technicalities, though opinions from the peanut gallery will be given strong consideration.  For example, Maca's Five-0 zerobots are descended from a first ancestor which contained a start and an end.  While technically a violation of the rules (starting bots should be all 0's and nothing else) I deem descendents from this common ancestor to be elligable for the prize (this time).

Regarding default ZB environments, that is a complex subject which deserves it's own topic.  In my opinion, it is possible to evolve winners using radically different environments and the environments themselves may need to be modified by the human during different breeding stages to provide new and different selection pressures as evolution proceeds.  Thus, for the contest, any environment may be used and environments may be changed as deemed necessary along the way by the human subject to the rules above (in particular, any organism with hand authored DNA added to the sim as part of the enviroment must have mutations disabled).

[Numsgil]

I can attest that a static environment with modest costs at the beginning isn't the way to go.  I have a sim in the hundreds of millions of cycles and nothing noteworthy.

[MacadamiaNuts]

Would any amount of repeats of this:

0 499 999
999 499 0

qualify as a valid "zerobot"? Those are unused sysvars, as far as I know, and mutations would spread better through the whole sysvar range instead of favouring the lower values.

[EricL]

Fine with me.  A vector of random numbers would also be okay.  Any set of starting numbers would be fine as long as the entropy does not differ significantly from that of a pure zerobot.

[Sprotiel]

Fine with me.  A vector of random numbers would also be okay.  Any set of starting numbers would be fine as long as the entropy does not differ significantly from that of a pure zerobot.

Entropy is certainly not a relevant characteristics: a pure zerobot has minimal entropy, while a purely random genome has maximal entropy. Intuitively, both should be admissible because they incorporate no relevant information wrt. bot behaviour and evolution.

[EricL]

Sorry, I should have said they must have similar relevant information entropy (otherwise known as Kullback–Leibler divergence) relative to the set of all "functional" bots.

You are corret to say the entropy of a random sequence is higher than a sequence of all zeros relative to the space of all possible seqeunces, but the set of all possible sequences is not the relevant set we care about.  A zerobot and a random bot both have similar information entropy relative to the set of all functional bots (bots that do something interesting including all hand authored bots).  Said another way, their information content is zero or nearly so relative to bots that do something interesting.

I think the root of the issue is that in a evolutionary system, the definition of "organized" is not the same as "statistically improbable".  A zerobot is statistically improbable but as far as DB is concerned, it is unorganized and thus exhibits high relevant entropy.

[Endy]

I can attest that a static environment with modest costs at the beginning isn't the way to go. I have a sim in the hundreds of millions of cycles and nothing noteworthy.

I've gotten decent results by mixing up the enviroments. First I started off with a gravity/brownian motion enviroment, with veggies giving free energy shots. As bots learned to feed/repro I installed standard veggies. As the bots adapted and learned to move to some extent, I lessened the amount of gravity and brownian motion in the sim.

Evolution seems to work best if it has something to work on. The bots need to have increasingly harder challenges to test their mettle against.

[Numsgil]

Evolution seems to work best if it has something to work on. The bots need to have increasingly harder challenges to test their mettle against.

I'll second that.  Evolution doesn't like to make progress, you have to drag it kicking and screaming to the next plateau.

[EricL]

Evolution seems to work best if it has something to work on. The bots need to have increasingly harder challenges to test their mettle against.

I will third that.  But I will also point out that the any measure of evolution that uses terms such as 'progress' or 'better' begs the question as to what external frame of reference is being used to measure it.  Humans tend to judge things in human terms and therefor tend to (incorrectly) ascribe some sort of built-in goal or relentless march towards ever increasing complexity to evolution.  Humans often think that becoming more human-like is somehow 'better' or represents 'progress' and thus we use terms like 'primitive' to refer to less complex organisms around us.  Of course, such perspectives are flawed.  Evolution can produce complexity to be sure, but only if selection favors complexity.  This of course, is the exact point both of you are making.   But many miss the point that every other extant organism on this planet is the end product of the same exact evolutionary time span and represents a successful end product of evolution just as we do and the vast majority of them are single celled.

So, while I think it is fair to say things like "bots need increasingly complex challenges to evolve increasingly complex adaptations" I think it is incorrect to claim that copmplexity is somehow 'better' or represents some sort of 'progress' absent some well-defined frame of reference.  In short, we would not long survive in a hot spring.

[scood]

whers a zero bot? I cant find one in the bestiearry

whats a zero bot

[abyaly]

whers a zero bot? I cant find one in the bestiearry

whats a zero bot

a zerobot is a bot with a bunch of zeros for its dna.

[scood]

were can I get one?

[Jez]

***Zerobot sim, what is it?***

http://www.darwinbots.com/Forum/index.php?...1958&hl=zerobot

You'll need to write your own, it's not difficult, the link describes a zerobot and zerobot sim.