Gordon Bell is talking about the 1TB life. (side remark: i get google ads for a 1TB drive at $900). 1 TB gives you 65+ years of 100 emails a day, 100 web pages, 10 photos a day, 8 hours of audio, etc. why not store it?
mylifebits is about capturing this information in a database. You need the ACID properties of a database to manage this information.

mylifebits is about the automatic creation of metadata for rich media

Gordon talks of the need for accurate speech recognition and automatic transcription of audio, and image recognition to identify persons on photos. he scanned in all his old papers back from the 60s to the present. his team added more data capture over the years: browser history, phone conversations, GPS information over time.. Killer apps beyond search might be nostalgic screensaver, and eventually of course, the memex.

we feel good about putting our home movies in the database (and never look at them) compared to having them on tape (and never look at them) because we know we could. they are also looking at caloric intake vs output to plot health state.

why bother? 1) because we can 2) because we need to deal with digital media 3) it eliminates atoms (hello environment) 4) for business 5) preservation for historians 6) for the frail human mind
the storage structure behind mylifebits shares many of the winfs goals, of course: freedom from the folder structure, rich metadata, useful querying. gordon mentions timelines as a very useful ordering concept. spatial ordering by assigning photos with locations on a map could also become interesting. tagging with dublin core is mentioned, or personal taxonomies. “let me not go into this ontology hole” 🙂 gordon just mentioned deja view, a device attached to your glasses that records the last 30 s in short-term memory and allows you to commit interesting scenes to permanent storage.

to really do the metadata would take another lifetime. i hired an assistant.

sounds like an outsourcing opportunity to me.
alternative write up

• Uncategorized