september 2004 / june 2005
As with much in life this is something of an ongoing interest/concern/thought, and it didn't get thought about much more until august 2004 when I came across an article on "life caching" in the paper (for further links check the end of this article). It was interesting, if for no other reason than to get me to look over my previous thoughts about this subject and delve further into these ideas.
Their conception of life caching is not quite the ultimate aim I was contemplating in the hard disk as me, it is more an emerging trend based on:
- - blogging
- - digital cameras
- - cheap storage
- - ubiquity of the above
In other words it is only a small extrapolation of existing technology and practice that takes into account the increasing availability (and corresponding lower cost), desirability, ease-of-use, convergence and user-base of these tools.
The most basic of these that could be legitimately cited as a life cache is probably the photo-blog, effectively an online diary with pictures, or a digital photo album, depending on how one chooses to use it. To some it is no more than a glorified diary or scrapbook, but it does contain some of the basic principles I desire: longevity and searchability. Once these features are present we are definitely working toward the right area, the ability to retain and recall beyond our normal capabilities. What we lack with the photo-blog and other present-day implementations of digital memory ideas are the transparency and passivity of the process.
We have begun to have ubiquity, it is a trivial matter to own a digital camera now that they are small and cheap (not to mention available on many phones), and this ubiquity solves one problem: it is now easy to take a quick digital picture to document some aspect of life. What this doesn't solve is the effort involved in this process. It is not the passive memory recording that we dream of, merely a very small step towards it. We still have to actively decide to take a picture, still have to have the time and inclination to do so, still have to have the right equipment with us as some form of gadget that takes up space in our bag or pockets, still have to have free hands to handle the camera. In short we still need to make a conscious effort with all the ramifications this implies. Yes, it is becoming easier. No longer do we need to scan an (analogue) photo-print to digitise these moments, no longer is it necessary to carry exceptionally bulky equipment. But we still have to take the picture and sit at the computer to upload and comment on it, tools such as camera phones can remove the necessity of a computer but the effort is still there.
This lack of passivity reflects also the lack of transparency. By this I mean that not only does the digital memory process require a thought and physical effort, but also visible objects of technology. We have not yet progressed to the stage of distributed, truly ubiquitous computing where processors are in everything, all in contact, an invisible network of computation that any of us can access at any time to fulfil our needs; nor even the lesser stage of wearable computing where vital components are secreted in our clothes or worn habitually in jewellery, glasses, watches... Miniaturisation is a step towards this, but even a small camera-phone is not yet wearable in the true sense.
There are some people taking this work forward. A good example, one that builds on the principle of the photo-blog, is Nokia's Lifeblog. This is a system which takes content from your mobile phone – photos, texts, voice recordings – and creates a browse-able timeline of these events:
"Nokia Lifeblog is a PC and mobile phone software combination that effortlessly keeps a multimedia diary of the items you collect with your mobile phone. Lifeblog automatically organizes your photos, videos, text messages, and multimedia messages into a clear chronology you can easily browse, search, edit, and save. Nokia Lifeblog does the work of organizing the items you create and receive, and you can also add notes throughout the day, or tag and update your favourites so they're always on your phone."
(quoted from http://www.nokia.com/nokia/0,,54630,00.html)
The working name for this project was "memory prosthesis". The ideals are certainly here, but the implementation is still lacking. For one thing I hate phones, for another this capture is still limited to specific events during your day, a text message or taking a photo, events which are non-passive and also fail to encapsulate much of the average day (unless all you do is take pictures and text message).
mylifebits
A more exciting, far reaching and inclusive project is MyLifeBits:
"MyLifeBits is a lifetime store of everything. It is the fulfillment of Vannevar Bush's 1945 Memex vision [more on this later] including full-text search, text & audio annotations, and hyperlinks. There are two parts to MyLifeBits: an experiment in lifetime storage, and a software research effort."
(quoted from http://research.microsoft.com/barc/mediapresence/MyLifeBits.aspx)
This is a project far more in the vein of the hard disk as me, setting
out to record and store everything we see and do: pictures, text, web capture,
emails, video, TV capture, audio capture... everything. The project is run
by Gordon Bell at the Microsoft Research Labs, so has the necessary backing
to develop all the systems necessary. The software is built around several
databases that are used to catalogue each different type of media, these can
then be searched by key words, location, time, date and a whole host of other
factors, some automatically recorded and others entered by the user. Important
factors in this work are the variety of different data types that need to
be handled, the storage required, methods of capture, methods of annotation
and automation of annotation.
For example, photos uploaded to the database could automatically be cross-referenced with GPS data from the user to pinpoint the location they were taken. Provided the user uploads all the relevant information this could be an automated process. On the other hand a user may wish to link photos of a particular person to records of emails from the same person. Without good facial recognition software this could prove to be a laborious task. In one paper the team comment that:
"Marking up media with any kind of meta-data is usually considered a chore; something that users are not usually willing to do."
These are certainly issues of concern at the moment, but will no doubt become more trivial in the future as software and hardware becomes capable of these intelligent tasks.
Another interesting feature being explored in the MyLifeBits project is the creation of narratives or stories through the data. Ideas here include an automatically generated map of a journey showing the photos taken along the way, a replay of your day created by a time-lapse movie of the pictures taken, or a user-created narrative of personal highlights which could then be blogged. Contrast this creation of narratives with the raw data:
" ...there is also an element of passing the bits on for posterity, and it is questionable whether one’s grandchildren will be able to make any sense of such a mountain of data. Likewise, it is not always easy to share experiences with others. Jain has suggested that we need to move from an overwhelming log of raw data to a useful eChronicle."
(for a full discussion of MyLifeBits and storytelling see http://research.microsoft.com/barc/mediapresence/MyLifeBits.aspx and http://research.microsoft.com/~jgemmell/pubs/MyLifeBitsSubmittedJan2005.pdf)
This sounds like a promising start, an interesting and worthy project, but there is still a long way to go; evolution of hardware, software and our even own concepts of digital memory will be required before this becomes as awesome and life-changing as it could possibly be. Like most blue-sky projects this is, by necessity, rooted in current technology and this is where the ideals fall down. It's a necessary step, a big one, but only one of many.
sensecam
A project quite closely related to MyLifeBits is Sensecam. This is a project to develop a wearable and automated capture device. Sensecam uses various sensors (motion, temperature, light) to work out when the environment around the user is changing and take a picture. The pictures, along with sensor data, can then be incorporated into a life archive such as MyLifeBits. The key feature of Sensecam is the true passivity:
"Passive capture lets people record their experiences without having to operate recording equipment, and without even having to give recording conscious thought."
Meaning that the resulting pictures can be used as memory aids (who did I meet last week? where did I leave that book?), evidence of accidents, or an automated diary where each entry is the time-lapse film of the day's photographs.
The problems are the same as before, namely how to organize and make use
of all this data? In a recent
paper the group set up a year's worth of data to see how it behaved. It's
startling to think that this accounted for around 81000 images, requiring
133Gb of storage plus 19Gb for the corresponding MyLifeBits database. Although
the storage is getting cheaper all the time this is still a sizable amount
of data and requires a high-end computer to be searchable at speed.
And, as before, we're left wondering whether we could make use of all this data. The thing I'm most reminded of is the short film Salaryman 6, about an office drone who seems to wake everyday with no memory of yesterday. He modifies a digital camera to automatically take pictures throughout his day and soon finds the problem. Everyday is the same, the same boring office and the same repetitive tasks, does he suffer amnesia or just boredom? What kind of life would we record?
"After 4 days I had 4 films, I was curious to see what they were. In my lunch hour I got them developed.
Apartment...
Ticket Machine...
Office...
Convenience Store...
They were all the same, is this my life?"
Salaryman 6 by Jake Knight
development, formats and corporate muscle
OK, some of this is starting to sound like a Microsoft ad. I'm incredibly excited by the possibilities these projects are starting to grasp, but we should keep it in perspective. Nokia? Microsoft? All this stuff is being developed by commercial companies. Personally I'm not sure if I trust Microsoft to handle my life, if nothing else it implies an increasing stranglehold on the computing industry if a must-have product such as MyLifeBits were to be released and be useable only with proprietary Microsoft hardware. But what about bugs? What about hacking? What about having all your digital memories stamped with a Microsoft logo? What about all of your memories being prefixed with "my"? Oh dear.
I'm sure most of this wouldn't too much of a concern, after all we're talking about a hard disk with a load of files on it and some databases. Not too different to your average PC right now, just setup better for memory retention. It's not as though you'll be getting something from Microsoft implanted in your skull (not just yet anyway) and it's not as though you won't be able to keep backups of all this stuff.
A more worrying scenario might be a format war, the Betamax vs. VHS of memory archives. What to do if your memory is suddenly obsolete? Although this also sounds alarming it probably won't prove to be much of a problem, there's always the possibility of emulators and converters to allow you access to outmoded file types... assuming that these are within the grasp of the average users of course.
what kind of life?
Interest in the texture of life is another facet of this debate – not only do we need to address the issue of how to record and organise our everyday experiences, but what to do with them once they're stored. I'll come to some of the points raised by blogging later (primarily the ability to share your experiences), but something that I relate strongly to the promise of MyLifeBits and Sensecam are the 'everyday data' projects that many artists are involved in.
In particular Ellie Harrison has embarked on various data-based projects (such as documenting everything she eats for a year) and uses the resulting data to form objects ranging from bar graphs to sculpture. Here both the practice of recording and the final output constitute the art; with current technology the recording itself is a big challenge and requires continual effort (perhaps it could be classified as performance?), while the final output of a project shows some of the possible artistic uses we could put this data to. Perhaps most users of MyLifeBits wouldn't be particularly interested in this kind of project, but it would be an interesting narrative extension, something above and beyond a simple text & picture summary currently offered by the blog and, perhaps, with the ability to provide some quite useful information. How much do I drink/swear/travel/eat? My own small-scale project is the compilation of bus tickets into a scrapbook, something that reveals the depressing slog of gainful employment through the medium of fascinating repetition. My only problem is that I'm mostly too lazy to add to it!
Another interesting project is Christian
Nold's Biomapping which uses a custom-made device to record journey details.
It records not only your route, but your vital signs. Now you can build a
map of your excitement levels at different places around town, or your stress
levels during a day at work. Given the right technology, this idea could be
extended to create a biomapping commons with maps covering different types
of data – not just vital signs but pollution, noise levels, notes...
"How is the town today? It may be interesting to know generally how the town feels today- that is, a general measure of the whole city in general. Or maybe I want to know how the whole town changes during the day, whether its stress levels vary, for instance.
What areas are more stressed than others? By combining vital signs information with GPS in the database, we may be able to make a ‘broader’ map of how particular areas ‘feel’ at different types of day.We may be able to navigate by ‘feeling’- looking on our maps of the city, and our Biomap, and seeing how different areas ‘feel’.
Maybe we will become a different type of traveller: travelling to the most stressed, upset looking areas of the city, maybe we will begin to wonder why they are like that…
Or maybe we will look for somewhere in the city that is particularly mellow, where, judging from the colour of the map, people are generally having a good time
Maybe, when we look at our map, these areas might be rather surprising.
Maybe, when we navigate in this way, we will find new landmarks, new places to discover.
Maybe we will see our city in new ways."
Extract from Information Commons by Alan J Munro, http://www.biomapping.net/reflect.htm
These ideas are phenomenal and cross many fields. It's not just the recording technology that's of interest here, but the social, scientific and artistic results we may gain from these schemes once that technology is established.
current benefits
I'm getting a little ahead of myself again here. Although the above are in R&D it's going to be years before we see widespread implementation of these tools. What about the photo-blogging I began this section with? How does this fit with the ideals we're striving for? Despite the effort it's a good start, as is just a plain old blog. Although essentially not much more than a diary the digital element, and all that is implied by this, is present. There is also a benefit over storing files on your own computer and sorting them chronologically: the ability to easily share these things and the ability to make associative links between them.
Although a far cry from the uber-memory that we may eventually gain this added ability to easily share and organize is fantastic, and brings a new social dimension to memory. My blog serves as both a repository for the thoughts I want to keep from forgetting and for those I want to share and debate with a wider audience – not only does it fulfil the personal role of memory but also a social and cultural role. Flickr takes things a little further, by enabling the tagging and grouping of photos it aids my creativity as, over time, recurrent themes coalesce into coherent sequences and sets. Without the tagging it is likely that I'd forget about many of the pictures that I end up categorizing in this haphazard fashion (this kind of grass-roots system of classification is know as a folksonomy - see also IFTF for flickr & folksonomies - as opposed to a top-down taxonomy).
There is little transparent and ubiquitous technology here (beyond the ready availability of the necessary hardware) and little of the automation I crave to provide a true digital memory repository, but these are good tools with grand ideals.
brain/memory overload
There is, however, a downside to this easy social interaction – the ease of overload. There is much of interest in this world, and the internet not only brings more of it straight to you, but helps it grow exponentially. The same goes for the crap. The net result, for me, is a strange balance between exultation, despair and time wasting. On the one hand I find good stuff, I enjoy it, I'm pleased; yet I find it hard to let go in case the currents never bring me this way again. Thus I end up spending hours when I should be doing something else trying to take in everything I find, until overload hits. It's not like the sedentary world of books and magazines and CDs, when I can leave something on the side and browse it as and when the mood takes me and, because of the limited number of things in the house, be fairly likely to remember it if I want to find it again. This digital world is ever changing and I feel like I need to consume it while it's there, lest it have changed next time I log on (this is particularly true of a popular tool like flickr where panic sets in that I may not get to see every great photo that is posted!). I need to learn to let go, to take it easy – this is actually something that a system like MyLifeBits would help with as it would log my trail through the web (although that would probably end up looking like my equally overloaded browser bookmarks menu). The downside is the slog through crap, it becomes increasingly difficult to find a good, relevant or trusted source when researching a topic as the web is flooded with so many opinions.
I wouldn't classify either of these thoughts as negative views of the internet or ease of sharing, the point is more that I'm often lacking the correct tools and outlook to surmount these problems. I'm so overjoyed by the possibility that I forget about the practical side of things, the fact that I have a life that involves many other facets than just flitting through blogs and that, oft-times, these hours can feel remarkable unfulfilling: generally the promise outweighs the execution, generally there's never enough time full stop. These are behavioural and philosophical issues, not an error of the technology we're creating.
Still, it's no surprise that this rising network is challenging the way we
approach information, the information age is young and we've still got a lot
to learn and develop. But memory-wise I think we're heading in the right direction,
these tools bring the kind of associative linking and access to information
that Vannevar Bush dreamed of, they bring the kind of searchability and longevity
that I'm hankering for; all that's left is the natural progression of technology
and ideas to bring them into line with my hard disk dreams. It might
be a while though...
further links & related materials
- - trendwatching.com on life caching and generation C
- - the globe and mail: memory prosthesis (september 2004)
- - the guardian: how to save your life (august 2004)
- - wordspy tracking the term
- - christian lindholm and charlie schick: two main nokia bods and their lifeblogs
- - some thoughts from sunil vemuri at MIT
- - coining the phrase? the design of a human memory prosthesis, m. lamming, p. brown et al, the computer journal, 37(3) 153 1994
- - vannevar bush, as we may think (originally published in atlantic monthly, 1945)