Swarm intelligence is an interesting and potentially useful area of AI. With the current interst in community and social websites, crowd sourcing and collective intelligence any new idea in the subject of swarm intelligence is likely to have applications. In particular you really should be interested in academic research on the topic.
The only problem is that if you aren't part of an academic establishment it can be costly to read the lastest research. In an effort to spread the news about its journal Swarm Intelligence Springer has made it free to read until the 30 June 2010.
You can simply go to the website Swarm Intelligence, click on current issue and browse through the contents of all past issues. When you find a title you want to read you can click on it and be transferred to an abstract page. This is the point where normally the question of payment would arise but if you click on the PDF symbol you will find that the article downloads without even a registration being necessary.
The journal is only at Volume 4 so you won't be overwhelmed with a manual data mining task. If you are new to the idea of swarm intelligence then a good place to start is the article "The biological principles of swarm intelligence" in Volume 1, Issue 1.
As well as algorithms and computer approaches to implement swarm intelligence, the journal also includes coverage of simulation and analysis of biological systems. For example try Trail formation in ants. A generalized Polya urn process which is in the current issue (Vol 4, Issue 2).
Some of the most interesting papers from a programmer's point of view are the "cross over" subjects where biology has something to say about how algorithms should be implemented. For example, Ant colony optimization for multicasting in static wireless ad-hoc networks in Vol 3 ,Issue 2 or Ant colony optimization for real-world vehicle routing problems in Vol 1, Issue 2 - and yes ants do feature a lot!
On the hardware side it also covers algorithms and simulations of swarms of robots or intelligent particles. For example in the current issue (Vol 4, Issue 2) we have Energy-efficient indoor search by swarms of simulated flying robots without global information. See also Biologically inspired redistribution of a swarm of robots among multiple sites in Vol 3, Issue 4.
Notice that if you have a Kindle you can send the PDF to your reader using Whispernet or via a USB connection and read the entire set of back issues on your way to work.