memory lane

The fragment below was written on the 21st of September last year. I don’t think I can prove it any way or another, not that it would really matter. It was written in a haze, after the idea hit me like a steamroller.

On September 22nd, Facebook announced Timeline, which is the closest thing to the idea below which actually has an implementation. It’s kind of close to what I envisioned, in a way, even if less ambitious, but it may be just the beginning of what they’re doing. Anyway, made me wonder.


Writing this down before I forget all about it.

This is the kind of idea that can define a life. With it, the crippling fear of work, a bad implementation, wasting my life.

When the time is right, the whole world will see it. Until then, just jolt it down.

The idea came to me while watching my last.fm profile, specifically, the “most listened track” per month part of my profile. That section brings back a lot of memories. Looking at the most listened track for a month can make me remember what I was mostly listening to, in general, what was the general mood, a specific moment when I was listening to the track, etc.

Starting from this, one can become lost in reverie pretty soon. But this is just a data-point, even in relation to what my last.fm profile holds. But then, it’s just a drop in the bucket compared to all the data my online profiles have on me.

Facebook, Google+, Twitter, Chat archives. These are just some of the other data sources which might contribute to my memory lane. But that’s just me – Flickr, Foursquare, Goodreads, Shelfari, Amazon, whatever-the-last.fm-for-movies-equivalent-is (Mubi?), blogs, from all these data points one could almost recreate what happened to a person in a certain period.

But gathering all the data in one place, this is just the first step. A very important one at that, as even just getting these data points together can be a good place for a user to recreate his memories, if it’s presented in a memory-juggling way. But what I want to create is a step over that. I want to create a system that will not just gather all this data, but, uhm, understand is probably a word which would make me look like a mad-scientist/singularity-nut. So, not understand what happened, bu at any rate, create certain correlations. Some photos posted to Flickr have meta-data which say that they were taken in another city than the “Current City” listed on FB or Google+? twitter update locations confirm a location chage? Someone’s been on a trip!

In the end, the system would accumulate enough data on the user to create his narrative. Feed it enough writing samples, and that narrative is written in the user’s own style and words. Re-creating the user’s memories, with the user’s voice, from his online data in all the siloes he has data in.

Ana, bless her heart, said I’m trying to do what Caden tried to do in Synecdoche New York, in a digital medium. I hadn’t thought about it, but she is, if I think about it, annoyingly obviously right.

This is not a business idea. I could/would never monetize this, were it to be built. But this is the kind of idea one can dedicate his life to, and that’s fucking scary. The sheer amount of things I’d have to learn in order to create something of this magnitude scares THE SHIT out of me.

The ultimate mash-up awaits.

Mihai Chereji

Twenty-something web-dev & part-time self-saboteur with no sense of direction on the street or in life. Will write when bored, about music, programming, civic duties etc. @croncobaurul

Romania

Subscribe to cronco

Get the latest posts delivered right to your inbox.

or subscribe via RSS with Feedly!