Letters to the Editor

Letters posted here are associated with the following article:
The administration's best friends in Congress come to its defense. The false defense from Rockefeller and friends. How the "concision" requirement of television limits debate.
The letters thread is now closed.
  • War can be good...

    I sent this email to the UMASS republicans.

    Will you be posting results of your cost-benefit analysis of the Iraq war? I mean, how many people know that the computer mouse was a result of cold war missle control systems? (Remember the video game 'Missle Command')

    War can be good!

  • Demon Seed

    What's new is the practice. Lots of variables, lots of computer power, and an increasing number of real-world applications. What interests me about it in a political context is that the deus ex machina might be turned on its head, might become, if you like, the machina ex deus. It has profound philosophical implications, I think, and even though we may continue for the foreseeable future to stumble ahead, stepping into cowpies along the way just as we've always done in the past, anything which promises to relieve us eventually of the problem of scale seems worth pursuing.

    The movie (not the book) version of Demon Seed had the computer Proteus going off the rails and invading the home of one of its creators because it was going to be shut off -- because it had prescribed a method for averting nuclear holocaust that was politically unacceptable to the humans.

    I don't think genetic algorithms, genetic programming, computer evolutionary design, cognitive computing or many of the other "if we only scale it up enough it will start to think" designs right now would be capable of solving our problems of scale for us. What they do best is optimization and sometimes pattern recognition. Really good for reducing carbon emissions or data mining netizens in small rooms in San Francisco. Solving the problem of how to limp in a jumbo jet with major equipment failure in a way a human couldn't, great. Solving major world problems or creating a new filtering mechanism that avoids the problems of representative government, not so much. I saw a chart of progress in artificial neural computing, for instance, with curves indicating scale, and showing the motion upward from brainpower of a slug to brainpower of a mouse. Hope to be at brainpower of a mouse some time next calendar year. When did you say you wanted that war solved? Would a semi-slug-mouse do for now? That's assuming that the only thing wrong with the algorithms and methods is scale. A fly can elude you for hours if it's lucky, some of those problems it's doing are solved by as few as 30 or 40 neurons. This isn't to pour cold water on radical new computing methods, I love them, I do them for a living. But please try to find a bunch of humans to do the really difficult stuff, they think better.

    I did an internal paper on using the internet as a thinking device once, got good reviews. Have even tried it a few times. Works really well. But the unit CPU in that case is a human brain. Walter Freeman thinks that the purpose of the human brain is to interact with other human brains, the internet is a medium in which this occurs.

    I'm not sure that the characterization of the brain as having two brains a rational one and an emotional one works well unless you allow each to operate on the other, and allow a state where the rational is being regulated and stimulated by the emotional one. And it would be difficult to characterize the subconscious in such a model. But then there's so much Skinnerism still in these fields that it wouldn't surprise me if no one cared (warning: my ideas on this last are not considered mainstream even when they account for facts parsimoniously -- private gripe).

    That said, William has hit on an especially critical problem. Many of our governing problems are getting worse just because of too many people. And a representative that can't hear clearly what the constituents want can't be expected to do the will of the constituents. And a selection system that allows a ruling class rather than constantly purging itself of deadwood suffers from the same problems that the intelligence agencies were combatting in coming to that NIE last week.

    Bill Clinton was asked his opinion on pretty much this core problem on Jon Stewart, his take was that there is so much legislation to vote on, and so much time and pressure to raise funds and meet lobbyists, that most members of our congressional delegation are suffering from stress caused by lack of sleep. If true, it could indicate that any closed door session with no note taking is risky because those who were to be informed may have been unable to concentrate on what was being delivered.

  • Maybe Jonah

    Should bring along that beast of a mother of his. If anyone starts any trouble, she could eat them. Or look at them. And frighten them senseless.

  • I Was Hoping

    Mr. Greenwald would have cleared this up by now, but I think Amity is correct in that if/when the loathsome Sen. from Tel Aviv decides to caucus with the Repubs. that would change the majority/minority situation (provided Sen. Sanders voted with Demos., I think Darth Cheney would be given a vote to break the tie).

    I actually wouldn't put it past Lieberman to do that if it looks like nothing is going to happen on Iran (that he wants to happen). He knows that (God willing) after 2008 we should have enough cushion to tell him to suck ass & strip him of all seniority, etc.

  • Now one of Jonah's supporters

    has offered to "menace the crowd with an AK47".

    Looks like police presence may indeed be necessary to counter the threats of violence from Jonah's *supporters*!

  • Paul in KY

    I Was Hoping Mr. Greenwald would have cleared this up by now, but I think Amity is correct in that if/when the loathsome Sen. from Tel Aviv decides to caucus with the Repubs. that would change the majority/minority situation (provided Sen. Sanders voted with Demos., I think Darth Cheney would be given a vote to break the tie).

    I'm no expert in Senate procedure, so I only know what I've read, and I've read a lot -- and heard a lot from people who are more or less experts in Senate procedures -- that Lieberman's defection would NOT result in a change of control in the Senate. Here's is a good discussion of that argument - I don't know if this is true or not, but I've heard it in many reliable places -

    http://www.dailykos.com/story/2007/1/18/211259/764