10 Replies to “UPDATE to a needle in a haystack calculation

  1. 1
    bornagain77 says:

    semi OT:

    Bernard d’Abrera on Butterfly Mimicry and the Faith of the Evolutionist – October 2011
    Excerpt: For it to happen in a single species once through chance, is mathematically highly improbable. But when it occurs so often, in so many species, and we are expected to apply mathematical probability yet again, then either mathematics is a useless tool, or we are being criminally blind.,,, Evolutionism (with its two eldest daughters, phylogenetics and cladistics) is the only systematic synthesis in the history of the universe (science) that proposes an Effect without a Final Cause. It is a great fraud, and cannot be taken seriously because it outrageously attempts to defend the philosophically indefensible.
    <a href="http://www.evolutionnews.org/2.....51571.html

  2. 2

    “And, if one wishes to argue that somehow, a much simpler configuration formed a self-replicating molecule, and then it elaborated itself and moved from RNA molecules . . .”

    Of course there is also the little problem that there is no such thing as a self-replicating molecule.

    “it took “2,737,850 million billion billion billion monkey-years” to reach 24 matching characters”

    One of the interesting things about this is that they weren’t trying to reach a particular sequence or reproduce a particular play. Any sequence within all the works of Shakespeare would do the trick. This highlights just how difficult it is to come up with functional specified sequences.

    Now, I’m not sure that coming up with language sequences is a perfect example for nucleotide or amino acid sequences. It seems to me that it might be harder to come up with language sequences. Nevertheless, the principle illustrated by the exercise is important.

    Question: When they talk about “monkey years” above, I presume they are talking about a single monkey? So the 24 letters were reached by a billion billion monkeys in 2.7 million million billion years. Am I understanding that right? How fast were the monkeys typing and how does that compare to the very conservative approach of using the Planck time?

  3. 3
    kairosfocus says:


    They did some sort of reasonable typing speed estimate for that and converted computer sim time into monkey time.

    10^50 possibilities is searchable within resources, but 10^150 simply is not.

    GEM of TKI

  4. 4
    kairosfocus says:

    A serious point of concern!

  5. 5
    rhampton7 says:

    All of which demonstrates the remote probability provided that an exhaustive search is the only method accessible to ‘nature’. That’s a big assumption being challenged on multiple fronts, one of which is Novelty Search:

    …While we often assume that complexity growth in natural evolution is mostly a consequence of selection pressure from adaptive competition (i.e. the pressure for an organism to be better than its peers), biologists have shown that sometimes selection pressure can in fact inhibit innovation in evolution. Perhaps complexity in nature is not the result of optimizing fitness, but instead a byproduct of evolution’s drive to discover novel ways of life.

    A recent extension of novelty search called minimal criteria novelty search (MCNS) is a potential further step in this direction: Natural evolution has discovered many ways to meet the constraints of survival and reproduction. MCNS is a general formulation of such constrained novelty discovery that abstracts natural evolution as a search for many ways to do the same thing (and this same thing need not always be survival and reproduction like it is in nature).

    List of publications

  6. 6
    kairosfocus says:


    First, I thank you for sharing your thoughts; which help us all see how people on your side of the design theory issue think about the evidence and analysis we face.

    However, the following, I must directly note, is directly contrary to what the OP actually says, to the point of regrettably being a — probably inadvertent — strawman distortion:

    All of which demonstrates the remote probability provided that an exhaustive search is the only method accessible to ‘nature’. That’s a big assumption . . .

    1 –> On the contrary, the very POINT of the above, is that just for 73 ASCII characters worth of info, the PTQS resources of the solar system, could only sample 1 in 10^48 of possibilities. Just the opposite of “exhaustive search.”

    2 –> For double that, 143 the PTQS resources of the observed cosmos could sample only 1 in 10^150 of possibilities. (And in fact, much less, it takes 10^30 PTQS’s to do the fastest chemical interactions.)

    3 –> So, the maximal BLIND — chance + necessity sample, can only take the sort of 1-straw sample as described from the field of possibilities.

    4 –> THAT IS WHY, ON SAMPLING THEORY, THE SAMPLE WILL WITH MAXIMAL LIKELIHOOD BE TYPICAL OF THE DISTRIBUTION AS A WHOLE, NOT ATYPICAL. That is, we will with maximal likelihood, get straw, and nothing else, even if huge ABSOLUTE amounts of atypical resources lurk.

    5 –> The only empirically verified — and routinely seen — workaround to that, is intelligence, as can be seen from posts in this thread and blog, as a quick and familiar illustration.

    6 –> In short, the only empirically and analytically known — on billions of examples and zero credible counter examples — source of FSCO/I is intelligence. So, as with the laws of thermodynamics, we are entitled to infer on that pattern until and unless a credible counterexample is presented.

    7 –> So also, the matter is not the strawmannish “assumption” that you would project to dismiss. It is an empirically and analytically well-warranted inference to best scientific explanation of a quite common phenomenon. That is in the present.

    8 –> And, for the remote, unobservable past,the same application of the principle of uniformity that was pioneered by the likes of Newton [across space to other, remote, star systems], and by the pioneer origins scientists including Lyell and Darwin, is used. On empirically tested and reliable signs, and their best explanation, we reconstruct the past we cannot directly observe or the remote reaches of space we cannot directly observe. So, your projection of an “assumption” is a selectively hyperskeptical dismissal.

    9 –> Going further, sampling theory also tells us what would be required to make such a rare outcome more plausible, would be a quasi-infinite number of cases. That is, the real alternative to the above is a quasi-infinite multiverse.

    10 –> But, as I have discussed here (cf here also in the ID foundations series at UD), such a philosophically speculative — there is precisely zero observational evidence for such — multiverse hypothesis only serves to promote the complex functional organisation issue up one level.

    11 –> For, it turns out, our observed cosmos sits at a local cosmological operating point that is credibly locally fine-tuned for facilitating C-chemistry, cell based life. So, the underlying sub cosmos would have to be functionally specific and complex in its organisation to search this zone of possibilities with exceeding fineness.

    12 –> Just, to get to a sub-cosmos in which abundant H, He, C, O and N are possible, and with the sort of galaxies that provide suitable habitable zones for solar systems hosting long term stable terrestrial, earth-like planets. In short, ours is already an exceptionally privileged planet BEFORE you get to the problem of creating a ladder of chemistry that leads to life based on plausible environments and reaction chains.

    [ . . . ]

  7. 7
    kairosfocus says:

    13 –> Now, you go on to cite as a “challenge” the following highly speculative suggestion:

    Perhaps complexity in nature is not the result of optimizing fitness, but instead a byproduct of evolution’s drive to discover novel ways of life.

    A recent extension of novelty search called minimal criteria novelty search (MCNS) is a potential further step in this direction: Natural evolution has discovered many ways to meet the constraints of survival and reproduction. MCNS is a general formulation of such constrained novelty discovery that abstracts natural evolution as a search for many ways to do the same thing (and this same thing need not always be survival and reproduction like it is in nature).

    14 –> See how many REAL, speculative assumptions are built into the “challenge,” and how far we are from serious empirical or analytical warrant? Notice, how it is a priori ASSUMED that macro evo has happened, per chance variations and natural selection, and now that is to be generalised? See the a priori Lewontinian materialism at work, resulting in a raft of begged big questions?

    15 –> The first unanswered problem is that such a constrained search across a tree pattern, has to have a root. Without reproduction capacity, life as we know it cannot possibly evolve by chance variation and differential reproductive success of sub populations.

    16 –> But, already, that implicates a vNSR, which is enormously and irreducibly complex and functionally specific, requiring codes, algorithms, data structures and organised executing machines. (And if one suggests instead a ladder of spontaneous self replicators, Shapiro and Orgel as cited in the OP have cut that off, absent strong direct empirical demonstration, which is conspicuously lacking.)

    17 –> And when it comes to the origin of novel body plans, as also cited in the OP, you must first show from the fossil record — 1/4 mn plus species, millions of fossils in museums, billions in known fields — that there is an actual branching tree structure, contrary to the summary of Gould et al [cf. OP again].

    18 –> In short, you again knocked over a strawman here, and begged the question on whether there is actually as a matter of fact a tree-like structure to the architecture of life forms. Why do you imply, yes, and how do you show Gould — who helped found an alternative theory to explain why speciation, the gateway to such novel body plans, is persistently missing in the fossil beds — in error?

    19 –> In addition, on observing genomes and comparing the reasonable requisites for novel tissue types, organs etc, the dozens of novel body plans from say the Cambrian era forwards, will each require 10 – 100+ mn bits of additional protein and regulatory genetic information. How are you going to account for such on forces of chance and mechanical necessity, given the blind search challenge that is already insuperable at 500 – 1,000 bits?

    20 –> Perhaps, the cite is proposing a novel set of laws of physics and chemistry, leading to an automatic process of complexification and diversification in biology. So — apart from begging the question as to whether body-plan level macro evo occurred [there is a serious alternative on the table with empirical and analytical support, for the origin of FSCO/I] and the similar question as to the capability of chance variations and culling out of sub-pops on failing in the struggle for existence — how do you ground and empirically warrant such novel laws of physics, chemistry and biology? [As a starter, perhaps you could explain the homochirality of life on such laws.]

    21 –> but, even if you were able to propose and empirically warrant such laws, these would be laws of the cosmos, that program in life. In short, you would have proposed and empirically warranted yet another degree of cosmological fine-tuning and programing of our sub-cosmos.

    22 –> In short, you would have provided evidence for cosmological design and front-loading.

    GEM of TKI

  8. 8
    kairosfocus says:

    PS: Springing a list of publication-links as though that provides evidence amounting to warrant is the fallacy known as elephant-hurling, or in academic guise the literature bluff. Not good.

    I note from the linked, that Stanley and Lehman [presumed responsible parties] suggest:

    Rather than converging on a single optimal solution, nature discovers a vast variety of different ways to meet the challenges of life. As an abstraction of novelty-discovering in nature, novelty search directly rewards behaving differently instead of rewarding progress to some ultimate goal, which is the traditional approach to search. That is, in a search for novelty there is no direct pressure to be better.

    To be more precise, evolutionary search is usually driven by measuring how close the current candidate solution is to the objective. That measure then determines whether the candidate is rewarded (i.e. whether it will have offspring) or discarded. In contrast, novelty search never measures progress at all. Rather, it simply rewards those individuals that are different.

    Instead of aiming for the objective, novelty search looks for novelty; surprisingly, sometimes not looking for the goal in this way leads to finding the goal more quickly and consistently. While it may sound strange, in some problems ignoring the goal outperforms looking for it. The reason for this phenomenon is that sometimes the intermediate steps to the goal do not resemble the goal itself.

    What this actually does first of all is to underscore the key blunder of typical genetic algorithms, i.e they are intelligently structured hill-climbing searches of a well-behaved island of function.

    The issue that trends may lead to pitfalls is commonly ignored in discussions, including those we have had at UD over recent months. This, we have had to point out in laborious details, and the force of the point was often not acknowledged.

    So, the clip helps move things forward in that way.

    However, there is a key breakdown here, the search method is again an intelligently designed and structured one that critically depends on the design for its success. And, there is a subtle allusion to of course a reward function.

    This is a case of indirect hill-climbing, and a subtler level of targetting, in short.

    Next, there is a major gap here: the assumption that mere variation and novelty are to be rewarded, on the further assumption that “evolution” is a datum of fact to be drawn upon.

    But, the problem is not variation, it is successful variation, rewarded through differential reproductive success, crossing the space of possible configurations of atoms and molecules to achieve novel islands of function. Which brings us right back to the centrality of the need to find islands of complex, specific function in large config spaces.

    Surpise, the researchers give us an answer: they exemplify it — INTELLIGENT DESIGN. (Those coded algorithms did not suddenly appear by monkeys hitting keyboards at random . . . )

  9. 9

    Not only is the multiverse idea unhelpful generally, as far as the origin of life goes, it is utterly irrelevant. The multiverse generating machine can spit out universes all day, but it doesn’t answer the origin of life. Specifically, *given* a universe like ours, finely-tuned to accommodate life, what are the odds of life arising? There was apparently a time when there was not life in the universe. Now there is. Thus, at some point in time, within our universe, life arose. That life did not arise due to some multiverse machine outside of our universe. Rather, it arose within our universe, within the timeframe of our universe, and within the parameters of the laws of chemistry and physics that exist in our universe. Thus, it is simply irrelevant what else is going on in other alleged universes.

    Even granting that we got lucky and ended up with a finely-tuned universe through some extra-cosmic fluke, given our universe, what is the best explanation for the origin of life?

    [Note, the multiverse idea might be relevant to the question of whether the universe itself was designed or the result of chance. There are, as you mention, lots of caveats and problems with the multiverse idea, but at least it would be relevant to that question.]

  10. 10
    kairosfocus says:


    The speculative idea — recall, there is no observational basis for it — is that by multiplying opportunities without effective limit [quasi-infinite] then any logical-physical possibility will be actual.

    Even with such a possibility, the problem of FSCO/I arises for the search to find the operating point for a locally fine-tuned world, is such that the best explanation is still design. Unless, you can put forward observational evidence for the quasi-infinite multiverse. Which has not been forthcoming.

    We have seen designers set things up at operating points, but for multiverses, there is zero empirical evidence. This speculation is philosophy not science, never mind the lab coats being worn by the speculators.

    And, if the subject is phil, then ALL the relevant evidence and argument come to bear, not just that put forth by a priori evolutionary materialists wearing lab coats.

    A very different balance emerges.

    GEM of TKI

Leave a Reply