Main Main Main Main

This area does not yet contain any content.

    Search:

     Subscribe in a reader

     Subscribe via email

    Entries in meaning (3)

    Sunday
    17May2009

    time, clocks, legnth, and rulers

    DEEP METAPHORICAL ANALYSIS via http://icanread.tumblr.com/

    While I admire the attempt to understand the true nature/meaning of time here, this is like saying "legnth doesn't exist, rulers exist."

     

    Tuesday
    05May2009

    memories, tangibility, and the nostalgia project

    So I've sent out some beta information this weekend on the nostalgia project I've been working on. For some reference, it's a simple submission-based database webapp that tries to get at the idea of nostalgia and meaningful memories by asking the question: "what's something you miss?"

    Chances are that a lot of the readers here already recived an email from me about it, but if you'd like to check it out over the next week as I integrate feedback and smooth the rough edges, leave me a message in the comments or email kylestudstill@gmail.com. I've limited these original invites to 100 handpicked people, but mostly because it's running off of an old iBook of mine until I move it to a real server next week. If you're here, I'd love to have you join in the fun.

    Already some interesting thoughts coming in: how necessarily connected are our memories to tangible things? We've definitely developed a kind of "on the cloud, we can store anything, forever" mentality, as if this kind of archive is some ultimate human goal. Will digital memories continue to satisfy us 20 years in the future? I think in some way we'll miss the tangibility of things that represented intangible memories. Not quite sure how that will develop just yet.

    Friday
    05Sep2008

    The Turing test, search, and meaning.

    Occasionally I enjoy talking with Jabberwacky, the artificial intelligence project/chat bot. I've had plenty of amusing conversations, and generally feel like I'm contributing my small bit to the development of AI models. I got to thinking of Jabberwacky today because of a fascinating article about the Turing test and some insights into the nature of search.

    Search in general is a fascinating thing to me, because at the core of search is the very nature of how people come to obtain information they have a true desire for. Being that search and user experience is in the global limelight these days, the new millenium has been full of insightful thinking on what that process looks like. This article in particular explores the problems of search from the role of the user and the role of the AI/search engine.

    On the user's end, search often runs into problems because we as users tend to 1) not word our queries very clearly or intuitively, and 2) rarely have a perfect idea of what we're looking for in the first place. Our most current solution? Suggestion panels that match most common querries and common popular phrases. The article goes into depth about this idea and how relatively revolutionary this kind of functionality is, but clearly there's a lot of work to be done in regards to beginning a search.

    Obstacles on an engine's end are a bit more complicated, but what I found particularly interesting was the problem artificial intelligence has with ontologies of words and their meaning. In short, computers are great at recognizing words and patterns, but fail horribly at understanding how those words relate to each other to produce any meaning. While we are by no means close to solving the incredibly elaborate puzzles wrapped in the nature of meaning, Ferrara talks about how projects like Princeton's WordNet are building databases of ontological references, and suggests that the future of search may be found in similar wiki-like public projects.

    All-in-all, absolutely fascinating. It's interesting to think of how search has come into the spotlight in a historical sense too, thinking through how what we as users expect to be able to do with the internet has changed over a few short decades. Ferrara makes an interesting point in noting that while Turing was wrong in his predictions of what AI could do by this point in our history, we have made leaps and bounds in providing online and search experiences that are incredibly rich and complex. In many other was we have gone far beyond Turing's most fantastical musings.