The May 2010 Communications of the ACM has a nice review and critique of the concept of lifelogging.* Beyond Total Capture: A Constructive Critique of Lifelogging (full text!) by Abigail Sellen and Steve Whittaker. I appreciated their bringing in the perspective of psychology and why-do-it questions.
Until now, most of the research in this area has focused on ways to pull in more and more data, information and artifacts that represent "everything" people are doing. This article suggests that the group needs to step back and consider the purpose behind the capability to do this. In their discussion, they talk about the "Five Rs" of memory-related tasks: recollecting, reminiscing, retrieving, reflecting and remembering intentions. The article states clearly that while the various lifelogging tools can do many of these things, none of them can do it all and none of them are designed with these specific tasks in mind - the designers just believe that these things will become possible.
It's time to test this belief and start moving into goal-oriented research, now that the tools have the wide range of capabilities they have. The article suggests several items: Do selective capture to enhance the specific memory-related tasks; lifeloggers should create cues to memories; memories are complex concepts, and the tools should be clear about what elements they are addressing; lifelogging can only be a support for memory tasks, it can't replace them. I assume this last is much to the dismay of Gordon Bell and the proponents of replacing human thinking with computation.
* Lifelogging is the concept of capturing everything you see and do - pictures, emails, receipts, even ambient conditions. The best known project is Gordon Bell's MyLifeBits , and the article references a ESPRC-funded project called Memories for Life that is geared around the same idea.
[Photo: "'As We May Think' Collagist Summary" by Derek Mueller - a modern reference to the Memex article by Vennevar Bush.]