Sunday, March 8, 2015

On Turing tests, robowriters, and typewriters as symbol and reality

 
 
 
 
    

From one of the most popular poets in the world:

 

 Real typing or unreal?

  


New York Times story:


21 comments:

  1. Like 'artificial intellegnece', 'smart'phones and the like. They are devicess. Dumb, stupid, and will only work if the embedded code is correct. None will ever do more than the code. So a typewriter is amachine, but it is a direct extension of the writer using it.

    ReplyDelete
  2. For some reason, it really bothers me when people don't capitalize "I" or the first letter of the first word at the beginning of a sentence. It amazes me that this "poet" doesn't even seem to understand basic english grammar!

    Sorry, off topic mini rant. I do sometimes wonder how far the computer will advance by the time I am an adult. Will computers replace writers in my lifetime? I fear that it may be a possibility.

    ReplyDelete
    Replies
    1. Blame e. e. cummings. And, um, I can't resist pointing out that "English" should be capitalized. ;)

      I expect to live to see people curling up with their kindlenooks to pore over a romance or thriller that a computer spat out in one second. I'm reminded of the sappy machine-written love songs that the proles enjoy in "1984."

      Delete
    2. I can imagine in another (couple?) hundred years people seeing computers with keyboards, screens, and mice/trackpads as nostalgic and pure as we see typewriters today—when those computers have been replaced with technology that interacts with our own thoughts and the lines between input and output are blurred.

      Delete
    3. A horrifying prospect, to me at least!

      Delete
  3. "The target of the Jihad was a machine-attitude as much as the machines," Leto said. "Humans had set those machines to usurp our sense of beauty, our necessary selfdom out of which we make living judgments. Naturally, the machines were destroyed."

    "Thou shalt not make a machine in the likeness of a human mind."
    -Orange Catholic Bible

    ReplyDelete
  4. I'm afraid you're mistaken on the robot's machine,not to be nitpicky. Looks to be a Royal KMM, not a Remington Model 17/KMC.

    -An adoring fan with a former drawband problem.

    ReplyDelete
    Replies
    1. Thanks for commenting. I think you're right, that is another KMM!

      Delete
  5. One of the things I like about writing on a typewriter, is the fluidity of mistakes. Some people seem to feel that it destroys the fluidity of language, but to be honest - in an era of machine corrected grammar and spelling, the mistakes return the humanity to writing.

    The 1919 Australian silent film 'The Sentimental Bloke' - based on the C.J. Dennis's poetry 'The songs of a sentimental bloke'. Featured heavy use of Australian slang in its intertitles, which was far from the standard 'Queens English' at the time. It was a voice of the people. In a lot of ways Typewriters give a real voice, rather than an artificial or augmented voice - in an era of photoshopped do perfection photographs and videos that are unreliable in their authenticity.

    ReplyDelete
  6. If a musical scholar were to write intricate code for a machine, code that mimicked the writing style of Mozart, I believe that composition could fool a great many people. If the machine managed to write a symphonic piece that hit most of the "broad strokes" of Mozart's style and the scholar tweaked the end result for performance we would see that as a collaboration. Finally, if I went to the local theater and listened to our local orchestra perform the piece, it really wouldn't matter to me who (or what) wrote the music. Just because I was emotionally moved by the piece doesn't make me feel "cheated" because it was written by a machine. In this case as with all other cases involving computers, we simply used the machine to do the "heavy lifting" of the project.

    believe most people get there panties in a twist when it comes to having machines do work that we deem ‘emotional’. Just because a machine barfed it up doesn’t mean the machine had any experiences that triggered those emotional cues, nor is that important. To my way of thinking, if I later learned that a poem I enjoyed was written by a stadium full of fine poets, and each one wrote one word down, then passed the pad over to the next poet until it was done- that would be cool! Refining computer code IS that stadium full of poets.

    Applying that same idea to the writing stylings of Elizabeth Barrett Browning, Hemingway, or any other recognizable writer's style, the same applies. I'm not dehumanized because I was bamboozled by a machine. Hell, I know adults who still run around espousing the genius of Dr. Seuss, I love those books too (for what they are). If kids listen to contemporary pop stars and deem them genius, fine. The argument should only ensue when anyone begins to compare one 'genius' (like Mozart) to another 'genius' (like a machine). The machine will only model the genius, it will never reflect on it.

    Now, let's see if someone can write code that helps a machine feel as if it's listening to puffed up, self-righteous political discourse while being loaded on wild turkey and barbiturates, then writing about it. I'd be hip to reading the next (albeit posthumous) Hunter S. Thompson book.

    Great post, and great comments- type on! ~Tom~

    ReplyDelete
    Replies
    1. Thanks, Tom, this is all nicely stated.

      The idea of finding meaning in the accidental or unintentional is very much a feature of modern art. I'm all in favor of it. The point where I get off the train is where people make the confused claim that machines can "think," just because they can generate material that makes *us* think.

      Of course, computer-generated material is never completely accidental or unintentional, because the human programmers had a hand in it. It's a collaboration, as you say.

      Delete
  7. By the way, I'm 94.8% sure that the "The more real..." poem is unreal: that is, it was created by a human being, but not on a typewriter. Instead, it was made with the app The Amazing Type-Writer.

    ReplyDelete
  8. Intriguing post, Richard. I reckon neither of those poems are "by" a "human" poet - for a start, they'd date and locate their precious words and give them a title; "Unititled" is a favourite one. And if little lower-case a.e.were a true modernist in touch with muse and typewriter he'd have made his h e a r t more o p e n.
    As for the sports report - surely it's not Vladimir Guerrero but Jesus Nijinsky ?

    ReplyDelete
  9. What's interesting is can you hold a computer responsible for libel? Whom is it that has accountability for what a machine writes about someone? It seems that there will have to be sets of "I, robot" rules to discourage ad hominem, slander, libel, etc. I guess the publisher is the likely bearer of ultimate responsibility, but after what court battle or account-draining quest for justice? I can see these anonymized personal attacks employed like so many land mines strewn about a third-world countryside - left with impunity, and the victims thereof literally left holding the bag of trying to put the pieces of their life back together, with no clear legal or moral recourse. Sigh.

    ReplyDelete
    Replies
    1. I hadn't considered these issues. They're partly analogous to the question of who will be responsible when a self-driving car gets into an accident. (Come to think of it, "automobile" already means self-driver ....)

      Delete
  10. Assuming these hard-done publishers pay them, what's Tommy "Robot" Eliot going to with all those royalties... ? Oil himself to-death, probably.

    ReplyDelete
  11. Very interesting. A lot has been said already. Just to share that I concur with your three conclusions.

    ReplyDelete
  12. If, as I used to understand Wittgenstein, the question is not what a word means but how it means, then you might say computers can in some fashion understand our languages. But these days I think Wittgenstein was after something different: my view of his discussion of language games is on the games, not the language. He was perhaps attempting to get at how we play and how play is involved in speaking and meaning. Which makes the case of computers thinking and speaking and writing and meaning things a bit more muddled. What is the difference between Deep Blue making all of the moves in chess and being able to beat the best human players and Deep Blue appreciating the deep beauty of chess and being able to share a chuckle with Magnus Carlsen or marvel at the drunken audacity of Alexander Alekhine?

    ReplyDelete
    Replies
    1. Good questions. In my view, a computer can't PLAY games; it can only follow rules. If playing a game simply consisted in following rules, then the best basketball players would never commit fouls.

      One could reply that there are meta-rules for when to obey the rules of basektball and when to violate them; but I think that at some point, good playing transcends any rules.

      Wittgenstein, for one, would argue that practice can't be reduced to rule-following. You have to be initiated into a way of life before you can understand any rules or play any games. This argument fits neatly with Searle's "Chinese room" thought experiment, which makes the point that simply following syntactic rules doesn't create semantics (understanding of meaning).

      Delete
  13. Lovely post! I agree with you. Computers can never replace humanity, not in the important ways.

    ReplyDelete