Slashdot Mirror

← Back to Stories (view on slashdot.org)

Learn About the FRDCSA 'Weak AI' Project (Video)

Posted by Roblimo on from the let-us-run-your-life-for-you dept.

Today's interviewee, Andrew Dougherty, has a Web page that says he is "...an autodidact mathematician and computer scientist specializing in Artificial Intelligence (AI) and Algorithmic Information Theory (AIT). He is the founder of the FRDCSA (Formalized Research Database: Cluster Study & Apply) project, a practical attempt at weak AI aimed primarily at collecting and interrelating existing software with theoretical motivation from AIT. He has made over 90 open source applications, 400 (unofficial) Debian GNU/Linux packages and 800 Perl5 modules (see http://frdcsa.org/frdcsa)." Tim Lord says Andrew's project "brings together a lot of AI algorithms, collects large sets of data for those algorithms to chew on, and writes software to do things like ... guide your whole life." As you might guess, Andrew occupies a pretty far edge of the eccentric programmer world, as you'll see from this video (and transcript). He calls himself "a serious Stallmanite" (his word), and has chosen the GPL for his software in the hopes that it will therefore help the greatest number of people. (Speaking of help, he's looking for interesting data sets and various "life rules" that can be integrated with his planning software, and one of the reasons he presented at the recent YAPC::NA was to solicit help in putting his hundreds of Perl modules onto CPAN.)

14 of 52 comments (clear)

  1. I'm confused. by gl4ss · · Score: 2

    did the weak ai write the summary and the perl packages?

    --
    world was created 5 seconds before this post as it is.
    1. Re:I'm confused. by Half-pint+HAL · · Score: 2

      Why do you ask how I feel about the summary and the perl packages?

      --
      Got them moderator blues I blieve I walk out the do', With these mod-points I been gettin', I 'most never post no mo'
  2. Weak Al by Chris+Mattern · · Score: 5, Funny

    Weird Al's lesser known brother!

  3. What? by OverlordQ · · Score: 2

    So he can write hundreds of Perl Modules, but not a single script that can automate the upload process?

    --
    Your hair look like poop, Bob! - Wanker.
  4. What? by antifoidulus · · Score: 2

    Ok so maybe his past couple albums weren't the best, but he is still as "weird" as ever... What's that, oh weak "AI(as in artificial intelligence"), sorry /. Mobile's shitty font got me confused there for a second.

    --
    Monstar L
  5. Clearly weak AI: FRDCSA? by schlameel · · Score: 3, Funny

    Any respectable AI would suggest a better name.

  6. In English? Please? by X0563511 · · Score: 3, Interesting

    ... a practical attempt at weak AI aimed primarily at collecting and interrelating existing software with theoretical motivation from AIT

    <stares blankly>

    --
    For large sets, this will be our guide even unto death, for the LORD will work for each type of data it is applied to...
  7. Re:The girl with the green tshirt and black shorts by The+Cat · · Score: 3, Insightful

    Here's the lesson to take from this folks: Never listen to the Internet, because no matter what you do, there will be a gigantic dick on the Internet who responds with "meh."

  8. Re:Greatest number of people...? by Half-pint+HAL · · Score: 5, Insightful

    Sorry, there's one other good justification for GPL: if you're reusing other people's software that's already GPLed. Which is basically what he's doing. He didn't chose GPL, the authors of the software he's combining did.

    --
    Got them moderator blues I blieve I walk out the do', With these mod-points I been gettin', I 'most never post no mo'
  9. Where have all the nerds gone? by NoImNotNineVolt · · Score: 3

    Seriously?
    A fascinating interview, a fascinating person, a fascinating project.
    And this is all we can muster? A few jokes about Weird Al, complaints about not putting his modules up on CPAN...
    And not a single person is interested in an open source AI project that seeks to build up a comprehensive ontology from the crap scattered across the web?
    The deafening silence in the comments is hurting my head. It's like all the nerds left, and we're now left with a SlashBI crowd.

    The least we could do is imagine a beowulf cluster of these, shit.

    --
    Chuuch. Preach. Tabernacle.
    1. Re:Where have all the nerds gone? by Roman+Coder · · Score: 2

      I too was hoping for something more substantial, like a discussion on the merits of his approach to A.I., etc.

      --
      "The future can only affect the present if there is room to write its influence off as a mistake." - Yakir Aharonov
    2. Re:Where have all the nerds gone? by Anonymous Coward · · Score: 2, Interesting

      Part of the reason state of the AI art is where it is now is because there is so much work on it that hasn't gone anywhere. So I don't think any attempt at improving it is a step forward, as the field has already demonstrated that there is a lot of attempts that were not a step forward. I'm not trying to say the field is static, just it will take some combination of luck and insight/vision for further contributions to actually go anywhere.

  10. Had lunch with him at the conference by Krishnoid · · Score: 2

    One thing that struck me was the time and effort he put into it. He's been working on it for many years now, which is probably why he's been able to write 800 modules for the project.

    1. Re:Had lunch with him at the conference by Anonymous Coward · · Score: 2, Informative

      Hi Krishna!

      The closest thing I have to a paper is this:

      http://frdcsa.org/~andrewdo/writings/Temporal-Planning-and-Inferencing-for-Personal-Task-Management-with-SPSE2.pdf

      Although the data was lost, there is a substantial amount of projects of the data (i.e. metadata) such as file listings, software names, and so on. Also earlier versions exist. So a lot of it can be reconstructed. Also the disk containing the data could still be recovered. But the code which fetches and manipulates the data is in tact. I am going to make a push to release it in a little bit on git hub. I don't think there's anything particularly special about the code, except that it tries to solve certain problems that I do not see an existing solution for, such as iterating over SourceForge and GitHub and packaging them, things like that. A lot is not done, but there are substantial components. By making APIs available we expose the capabilities of the software, for future integration.