Human computer programmers beat human Go players; press misreports it

Here: "With this defeat, computers have bettered people in the last of the classical board games, a game known for both depth and simplicity."

This is about equivalent to saying that "Shovels have bettered people in digging dirt out of the ground."

Folks, this is a machine built and programmed by us humans, the we employ to better our performance at a task we decided upon.  It is the crudest sort of magical thinking to attribute what happened to our tool as if it were an autonomous being.

15 comments:

  1. If the programers who wrote this computer program played Go against the program, who would win?

    ReplyDelete
    Replies
    1. That's like asking if the guys who built a backhoe try to dig a hole faster than the backhoe with their hands, who would win?

      Delete
    2. If it's just the guy vs a backhoe the guy wins because the backhoe needs someone to operate it.

      Delete
  2. http://www.terrybisson.com/page6/page6.html

    So what really happened is that the guy who taught Kasparov beat the guy who taught Karpov?

    ReplyDelete
    Replies
    1. No, Ken. No one "teaches" computers anything. We build machines (using software). Period.

      It is like you digging a hole with your hands versus me with a shovel. The shovel did not beat you, I beat you, using a shovel. Or like you arguing this point with your brain, versus me arguing... oh, never mind.

      Delete
    2. Ken, I just taught a piece of metal how to pick up food! I call my AI a fork. It is smarter than you at food.

      Delete
    3. No. There are different algorithms for the symbolic manipulation so needed to play Go. Some meat machines instantiate one of them, and some silicon machine has instantiated another. The latter seems a better algorithm. It is mere pro-meat prejudice to count one of these as thought and not the other.
      How after all do we even know that other meat machines think except by observing behavior we think indicative of thought. Again, it is pro-meat prejudice to exclude machines from this. The point of Bisson's story of course.

      Further no programmer taught the machine to play well. The machine learned through pattern recognition by playing itself zillions of times. This I take to be Josiah's point.
      All those programmers die. Who wins the game then?

      Delete
    4. Ken, the philosophical arguments here are airtight, and date back to Leibniz, and you are either ignorant of them or don't comprehend them. I don't know which, and don't really care, since you can't be be bothered to even think them through.

      "Further no programmer taught the machine to play well."

      It seems like you can't remember that I build these machines myself. It was neither "taught" to do anything nor did it "learn" anything: someone built a machine, and it did what it was built to do.

      "All those programmers die. Who wins the game then?"

      Hey, the guy who invented the fork is dead. (And the fork, by the way, "instantiates" an algorithm for eating.) Now we have to give all credit to the fork, and none to the inventor.

      Delete
    5. And Ken, let me grant you this: IF we already had good reason to believe that materialism is true, THEN your view would be plausible. But we have no such reason!

      Delete
    6. "Meat machines": a machine is a mechanism built by someone to achieve some purpose the builder has. Who built humans, and what is the purpose for which humans were built?

      Delete
  3. No Gene, the arguments are not "airtight". You are simply missing the whole point of the Turing test.

    And cells are filled with machines. You beg the question when you say only "someone" -- not a machine or a process -- can build a machine. Evolution has built many machines.

    Materialism is the best theory we have, and as you admit, adequate for this issue.

    ReplyDelete
    Replies
    1. Right, turn evolution into an agent building machines. Claim that others "miss the point" of a simplistic test that any 5-year-old can comprehend. Hold up a piece of complete incoherence like materialism as "the best theory we have."

      This is childish, cargo-cult level thinking. ("If we build a box that looks like a helicopter, we will get food!") It certainly is not going to convince me of anything. So why do you keep on going?

      Delete
    2. How could you seriously think that anyone "misses the point" of the Turing Test? It is about as simple (minded) an idea as any that has gained fame in the last century.

      If someone passes the algebra test, we must grant they know algebra! Even if it turns out that their father stole the test the night before and simply had his kid memorize all the right answers: No! We're not allowed to consider anything else. The kid passed the test, the kid knows algebra!

      Delete
    3. Here's a nice piece: "And most people in the computer age understand the distinction between living intelligence and the tools men make to aid intelligence — tools that preserve the fruits of the human intelligence that went into building them, but which are in no way intelligent themselves."

      But, of course, he's speaking of people outside the cult of materialism.

      Delete
    4. From the above: 'In discussing the “system” argument against his Chinese Room thought experiment, Searle once said, “It is not easy for me to imagine how someone who was not in the grip of an ideology would find the idea at all plausible.” The AI champions, in their desperate struggle to salvage the idea that computers can or will think, are indeed in the grip of an ideology'

      Yep. And ideologies render us unable to think on the topic of the ideology.

      Delete