icosilune

Archive: June, 2009

Pandora

[General] (06.28.09, 11:02 am)

The past few weeks have been taken up doing writing and coding for a couple of projects. The writing has been slow, but ultimately, I hope, productive. The coding has been for Statics and my self-motivated crash AI course. The Statics stuff is nice because I recently made a few very useful usability fixes, and fixed some really insidious bugs that should lead to much better stability.

In the meanwhile, I have Pandora to keep me company.

For those uninitiated, it is a radio station that allows users to build radio stations, consisting only of music that the user likes. They are an offshoot of the music genome project, which employs an interesting classification system to categorize songs. The player will find songs with characteristics similar to what the user reports to like and dissimilar to ones that the user doesn’t. This is very interesting to me from a technology perspective. Evidently Pandora has been around for a while, but I only just heard about it a few weeks ago. I really love internet radio, and I think that, despite a few annoyances, Pandora has managed to find some miraculous way to keep listeners happy, the music publishers happy, and themselves afloat. Which, given all of the nonsense that internet radio has needed to put up with recently, is pretty miraculous. I’m in there under “ashmore62“.

Minor news

[General] (06.12.09, 10:31 am)

Not much has been happening this summer. I’ve been doing some writing, some project work, and the like. Hopefully the writing will live to see the light of day and the project work will come out smoothly. At some point, I want to write a bit about games and entertainment, based on some productive discussions with Audrey, also referencing “player-centric design” which I have been reading about recently. I am going to be taking an AI class in the fall, the first AI class that I have ever taken, but it is actually an advanced one: specifically on narrative and AI. A friend of mine has described his recent disenchantment with AI, but I remain hopeful. Since I haven’t taken an introductory course, a few others and I are going to be doing some prep work before the class. The textbook we are reading is “Artificial Intelligence: A Modern Approach”, which is a bit stuffy (and curiously written, given all of the AI critiques I’ve read), but nice and thorough. A new edition of the book is coming out in October, so I’m borrowing a copy for now, and will probably get a real one later. One thing that surprised me, though, is after doing a cursory search online, there don’t seem to be many good open APIs for AI. At least not in the three minute search that I did. A lot of the algorithms described lend themselves very well to good abstract design principles, so maybe I’ll see about whipping something up later on.