Discoveroids Don’t Fear Intelligent Robots

The best thing about this new post at the Discovery Institute’s creationist blog is its brevity: Intelligent Design and Artificial Intelligence — The Connection. It was written by Klinghoffer. Here are some excerpts, with bold font added by us for emphasis:

It seems obvious on a moment’s reflection: intelligent design and artificial intelligence have something in common, and that is intelligence.

BWAHAHAHAHAHA! They do share the word “intelligence.” But artificial intelligence actually exists, while intelligent design, the Discoveroids’ “theory” that the universe, life, and you were miraculously created by a designer — blessed be he! — is a vacuous ark-load of babble.

But Klinghoffer thinks there’s something important going on here. He says:

What’s the significance of that? In an illuminating conversation for ID the Future, Robert Crowther talks about the connection with Dr. Robert Marks of Baylor University, co-author of the recent book Introduction to Evolutionary Informatics.

Here’s a link to that book at Amazon. What did Marks say to Crowther in the Discoveroid podcast? We’re told:

Marks and his fellow researchers have shown that evolution isn’t computable, meaning it can’t be successfully modeled — “There exists no model successfully describing undirected Darwinian evolution,” as Marks puts it. And you know what? The qualities that make human intelligence special are similarly not computable.

There probably isn’t one grand mathematical model for the whole theory of evolution, but there’s quite a bit of modeling going on — see Mathematical and theoretical biology. As for computing the “qualities that make human intelligence special,” we don’t know what he’s talking about.

Having made his point — whatever it is — Klinghoffer ends with this:

That, as Professor Marks explains among other helpful observations, makes fantasies about AI robots taking over the world, developing consciousness, or displacing the human race incompatible with reality. Listen to the podcast here. [Link omitted.]

[*Sigh*] At times like this, we miss Casey more than ever.

Copyright © 2017. The Sensuous Curmudgeon. All rights reserved.

add to del.icio.usAdd to Blinkslistadd to furlDigg itadd to ma.gnoliaStumble It!add to simpyseed the vineTailRankpost to facebook

. AddThis Social Bookmark Button . Permalink for this article

16 responses to “Discoveroids Don’t Fear Intelligent Robots

  1. Michael Fugate

    The DI’s mantra:
    If intelligence can’t do X, then only intelligence can do X.

  2. Derek Freyberg

    And Marks is hardly neutral in all of this – he’s an old friend of the DiscoTute and Dembski; not to mention having Dembski as a co-author on the book. Also, see his entry in the Encyclopedia of American Loons.
    But the book has wonderful reviews on Amazon: 9 of them, of which 8 are 5-star with one outlier at only 3-star.

  3. Marks and his fellow researchers have shown that evolution isn’t computable, meaning it can’t be successfully modeled — “There exists no model successfully describing undirected Darwinian evolution,” as Marks puts it. And you know what? The qualities that make human intelligence special are similarly not computable.

    Even if that’s true at present, it doesn’t follow that it will always be true. And I’m far from convinced that it is true now, for either evolution or AI. The models may not be perfect, but, again, that doesn’t mean they can’t ever be. And in any case I’m very doubtful that Marks et al. have actually “shown” anything, no matter what Klinghoffer and Robert Crowther claim. Klinghoffer in particular has proven to be, ahem, a less-than-reliable source of information.

  4. Ceteris Paribus

    Klinghoffer concludes:

    AI robots taking over the world, developing consciousness, or displacing the human race [are] incompatible with reality.”

    Apparently Klinghoofer is unaware of that special robot which I asked Santa Clause to drop off at my house this year.

  5. Michael Fugate

    That is quite a “distinguished” list of reviewers on the back of “Introduction to Evolutionary Informatics” – I guess they couldn’t find anyone who wasn’t a Christian apologist affiliated with Reasons to Believe, Marks’ Evolutionary Informatics Lab or the Discovery Institute. I do love that on Marks’ site he has a Christian Apologetics section – really! Who knows where the Chaitin quote comes from – not from any searchable review.

    And of course Marks hasn’t shown that evolution isn’t computable. He claims that, but he hasn’t shown that. For discussion see “The Skeptical Zone” blog.

  6. I don’t know what he means exactly by ” that evolution isn’t computable…”.
    It sure as heck can and is used in computing to find novel solutions.

  7. Michael Fugate

    Marks’ view is that mathematical simulations of evolution are really intelligent design; all simulations require a “purpose” to achieve directional change. As with most work in the ID community, it is wordplay and nothing more.

  8. An older movie on the subject of AI and computer dominance is “Colossus: The Forbin Project” (a.k.a. “The Forbin Project”), a 1970 American science fiction thriller film from Universal Pictures. Pretty good movie for 1970.

  9. AR: “non computable” means that there is no algorithm for finding a solution. For example, the sum of the infinite series 1+x+x^2+x^3… is easily shown to be 1/(1-x) for -1<x<1, and that sum is therefore computable. However, none of the several infinite series for pi is computable, but one can enumerate pi perfectly easily by taking enough terms of the series to give acceptable accuracy.

    Perhaps Klugehappens and Marks should find out about Alpha Zero, almost certainly the strongest chess-playing entity in existence. It taught itself to play to this level in only four hours, by playing against itself millions of times, selecting legal but random sequences and discarding those that led to a loss. Sounds like evolution to me. It was programmed with no advice on chess, nor with any specific target; all it had to do was win (survive?).

  10. I just wanted to repeat this posting by @Michael Fugate:

    The DI’s mantra:
    If intelligence can’t do X, then only intelligence can do X.

  11. ““There exists no model successfully describing undirected Darwinian evolution,” as Marks puts it. But evolution IS controlled by many natural factors. That’s why it works over time to select those best adapted to thrive in any particular environment and pass on those genes to descendants.
    The statement is INTENTIONALLY dishonest.

  12. Michael Fugate

    Marks, Dembski and Ewert are playing on the ignorance of most people when it comes to math, engineering and science. They conflate using an evolutionary algorithm to search for the solution to a problem (engineering) with using one to explore how evolution actually works (biology). Because some searches using an evolutionary algorithm are directed by the programmer, then all searches must be directed by the programmer and therefore evolution only achieves higher fitness if it is intelligently directed. They misrepresent how selection works in populations.

  13. @Michael Fugate
    The biggest problem in dealing with anti-evolution arguments is often in trying to make enough sense of them to be able to take them seriously. (It took me a long time to make any sense out of “how come there are still monkeys?”.)
    In trying to make some sense out of “computable” – well, I kept coming back to the obvious point that “Intelligent Design” is intractable, so that I couldn’t think what they were trying to say is bad about evolution. Maybe that it uses the word “random” ? Of course the creatonists are not going to be helpful – whether that is simply because they don’t have any idea about what they are talking about (and thus cannot be articulate about it), or because they are being dleiberately equivocal.

  14. Michael Fugate

    It all comes back to mantra, doesn’t it?

  15. @och will: “That’s why it works over time to select those best adapted to thrive in any particular environment and pass on those genes to descendants.”
    I would go along with the creationists here, and claim that evolution does not usually (i.e. almost never) produce a species best adapted to thrive in any particular environment. Rather, it produces something good enough. Given time and a stable environment, what was once good enough will perhaps be superseded by something rather better, but we have no reason, and no need, to claim that the absolute optimum will ever be reached.

  16. And the only forms which are possible mutations, by the laws of chemistry etc. Only changes in amino acids. For example, there is no way that a form using helium or plutonium or mesons.