Friday, October 17, 2014

[PHI 1000] Smartest Machine on Earth

If the mental (or mind) and the physical (or the body/brain) "are, in some sense, radically different kinds of thing," as dualism says, and what characterizes creatures that have a mind is the ability to think, then we should expect that purely physical things, like machines, would not be able to think.

Now, meet Watson.


Watson is able to answer Jeopardy questions by testing hypotheses. If that counts as thinking, then the following argument against mind-body dualism can be made:
  1. If the mental (mind) and the physical (body) are radically different kinds of thing, then machines cannot think.
  2. Machines can think (e.g., Watson).
  3. Therefore, it is not the case that the mental (mind) and the physical (body) are radically different kinds of thing.
Is this argument sound?

15 comments:

  1. In order to determine whether this argument is sound or unsound, we must look for truth in the premises. The premise here that I would like to focus on the most is "Machines can think (e.g., Watson)". It can be argued that due to Watsons limited database of information, and his inability to freely use that information, that is he really thinking? It can be said that humans dig into a database of information stored in not hard drives but their brains to answer questions and critically think. Watson can do this when prompted with a question, but can he ask himself a question? Can he use the answers he comes up with an apply it in situations that are not in front of him? I think that this is a big deal when it comes to making a judgment on whether or not he is actually thinking or just making a pre set association given data.

    ReplyDelete
  2. This argument is most definitely valid, as the premises given justify the conclusion. However, as much as it pains me to, I must disagree with it. While I personally side with materialism, I just don't think this is an appropriate argument for defending it. Machines cannot think. Although I've said it numerous times in class, I'll say again that I feel thinking requires a deeper ability than just that of data processing. A calculator can process data given to it in an efficient manner and provide an answer. Yet, we do not say that a calculator can think. Where do we draw the line? The machine is an extremely effective computer, but not a thinker. Therefore, I must conclude that the argument is unsound, much to my displeasure.

    ReplyDelete
    Replies
    1. From the video it looks like Watson is able to do more than just "data processing." He is testing hypotheses.

      Delete
  3. This argument in my opinion is unsound. The premises do lead directly to the conclusion which makes it valid deductive. And since we assume that Watson's ability to answer questions by testing hypotheses is considered thinking, then premise 2 is true. However we can not consider premise 1 true because it is just so general. What does it mean to be a radically different kind of thing, in what ways is it radically different. This vagueness leads to the inability to prove the premise true, which means that the argument becomes unsound.

    ReplyDelete
  4. This argument is not sound because one could argue that the first premise is false. Even if the mental and the physical are different things, that does not necessarily mean that machines cannot think. The mental and the physical could be completely different things but still interact with one another. This reasoning does not guarantee that machines cannot think.

    ReplyDelete
  5. I believe that the argument is sound, should the definition of mind be altered. A mind is conventionally believe to stem from the brain and retains the ability to think. But, machines seem to exhibit most conventional criteria for the ability to think. So, if the definition of 'mind' is changed to contain both bodily minds and computer minds, then the argument is sound. Is there any explanation to how a mind arises in the first place? Because, if a mind can be created from flesh, why not electronics?

    ReplyDelete
  6. I think that the argument is unsound because the accuracy of premise #1 is questionable to me. I don't see how whether or not the mind and body are similar relates to whether or not machines can think.

    ReplyDelete
    Replies
    1. The idea here is the following: if mind is thinking stuff, as mind-body dualism says, and physical stuff is not thinking stuff, then machines, which are made of just physical stuff, would not be able to think.

      Delete
  7. Sung Soo (Mike) MunOctober 21, 2014 at 2:29 AM

    This is a sound argument because if first two premises are true then it will definitely guarantee the conclusion. Although there would be a way to prove that all machines that does not only encompass Watson have abilities to think; however, it would not interfere with the validity or soundness of the argument

    ReplyDelete
  8. I don't think this is sound because the words think is still vague. It needs to be defined what thinking is and what respects the machine thinks like a person if it even thinks at all.

    ReplyDelete
  9. While this argument is valid I don't believe it is sound. I personally believe that there is a difference between thinking and data processing. When I take a test I'm not truly thinking i'm simply processing data that I acquired earlier. I only really think when I have to answer a "deep" question, such as this one. I believe Watson is incapable of answering these "deep" questions and therefor can't truly think.
    - Kyle Degen

    ReplyDelete
  10. I believe this argument is valid. However, I do not believe it is sound due to the premises being false. Thinking requires having a personal belief or opinion. It is personal and comes from within. Thinking does not require a “correct” answer. However, the only thing Watson is concerned with is finding this “correct” answer. This is not thinking, but rather problem solving with the use of probability as an indicator of the “correct” result. Also, Watson cannot function without a command. Take a tv for example. It can give you a movie recommendation based on existing data and specific algorithms. However, both watson and the tv are not thinking. They are merely functioning according to the way in which they’ve been programmed to.

    ReplyDelete
  11. In my opinion, this argument is unsound. While in class I supported the idea that machines can think, I cannot agree with this argument, because it implies that they think in the same way that humans do. It was established in class, that machines lack things such as emotions, imagination, and the ability to think critically and abstractly. This, among other reasons, is what separates machines from humans. Machines, even one as high tech as Watson, cannot think in the same way humans do, but rather in the most basic form of thought, processing data. So yes, while I believe machines can think, I cannot accept this argument as one that is valid, sound, and properly defending materialism.

    ReplyDelete
  12. I think that this argument is valid deductive, because the truth of the premises guarantee the truth of the conclusion. However, I agree with Erfana's point. The vagueness of the word "think" makes it so I can't confidently say that it is a sound argument. We had trouble during class as well, defining what "think" is in the context of the argument. Overall, I would say that the argument is unsound.

    ReplyDelete
  13. Testing hypotheses is an important part of human cognition (see the SEP entry linked to in the post). In other words, testing hypotheses is a kind of thinking. If so, the following argument supports the second premise of the argument in the post:

    P1. To test hypotheses is to think.
    P2. Watson can test hypotheses.
    C. Therefore, Watson can think.

    ReplyDelete

This is an academic blog about critical thinking, logic, and philosophy. So please refrain from making insulting, disparaging, and otherwise inappropriate comments. Also, if I publish your comment, that does not mean I agree with it. Thanks for reading and commenting on my blog.