(The following is a draft of an essay I’m kicking around and will probably post over at the C3 blog. I’d appreciate your thoughts and comments – it’s less of a blue-sky thinking piece and more of a clarification and “this is what I’m doing in this space” piece, so it’s a little different from my normal fare.)
There has been much made lately of the tech sector’s newest favorite buzzword: cloud computing. Like many such newly-minted terms, there is some dispute about its actual definition; I wrote about one such permutation in a previous entry for the C3 Weekly Update when the MacBook Air was about to be unveiled at the Macworld conference in January. In it, I conflated the terms ‘cloud computing’ and ‘ubiquitous computing’, but in retrospect I should pull the two terms apart somewhat. They’re still linked at a very basic level – both cloud computing and ubiquitous computing hinge on the idea of decentralization, which I’ll get back to in a bit – but by attempting to distinguish these two terms, we begin to gain a clearer idea of where our digital culture is heading next.
Ubiquitous computing, or ‘ucomp’ for short, posits a world populated by reactive data points everywhere you look. Similar to the world put forward in Stephen Spielberg’s Minority Report, these digital interfaces react to your presence and present useful information often embedded in the very objects we hold. Ucomp is a world of intelligent objects, of RFIDs and spimes, and its patron prophet-saints are Adam Greenfield (Everyware: The Dawning Age of Ubiquitous Computing, New Riders 2006) and Bruce Sterling (Shaping Things, with Lorraine Wild; MIT Press 2005). This is the world where the early settlers are the Nike+ sneaker, the GPS-enabled iPhone 3G and Wal-Mart’s embedded inventory systems.
Cloud computing, on the other hand, is the result of users divorcing themselves from individual computers and moving their data onto the web. In a way, this is a return to the era of the public terminals in college libraries that represented the earliest exposures to the Internet for me and others of my particular generation – only it’s no longer just e-mail being stored remotely in Hotmail or Gmail or IMAP accounts instead of being downloaded to local hard drives, now it’s all of our data. While we have become used to, if not addicted to, the twin pleasures of amassing vast amounts of content and working with it anywhere through laptops and smartphones, the two pleasures simply don’t play nicely with each other – unless you can unshackle the content from the access device.
For example, while the initial draw of websites like Flickr was the appeal of sharing my photographs with others, now the big perk is becoming the ability to access my photo library from any machine I want, anywhere I want, anytime I want. I personally maintain a photo library that clocks in at over a quarter of a terabyte, so the idea of being able to fit said library on my laptop and pull up any photo I want anywhere I want becomes laughable. Add to that the additional bulk of MP3 collections and the staggering girth of digital video collections and the issue becomes clear.
This shift was crystallized with Apple’s introduction of the MacBook Air: the initial revelation that the machine only came with a relatively tiny 80B hard drive was shocking, but once you took into consideration that said drive was only supposed to contain your operating system, applications and a bare minimum of actual data, it began to make more sense. The reinvention of Apple’s .Mac service as MobileMe was even more telling – although the new name is fairly hokey, placing the word ‘Mobile’ in the service’s name points directly to the intended unshackling of the user from any fixed location. Even more telling is the service’s new iconography: MobileMe’s remote disk on your desktop is represented by a purple hard drive with a fluffy white cloud on it, and the service’s logo is another happy little cloud with your apps embedded in it.
That said, this cloud is obviously still forming. Logging into your MobileMe website gives you access to browser-based versions of Mail.app, iCal, Contacts and Gallery that Apple happily describes as “your desktop on the web”, but I have a confession to make: while I’ve been trying to make the move into the cloud like a good little early adopter, I have yet to ever find a use for these applications. Perhaps it’s because I am so hardware-heavy that this unshackling isn’t meant for me: I carry a first-generation iPhone in my pocket, I often have my MacBook Pro with me and I have one very high-powered Mac Pro in my office at work and another lower-powered but still quite hefty dual G5 in my office at home. If I were to rely more upon the public computing clusters here on the MIT campus or in public libraries, these applications might be really handy, but so far that scenario has never come up.
Instead, I’ve cobbled together a strange hybrid of traditional local computing and cloud computing. In my current transitional model, I’ve offset my paranoia of storing all my data on someone else’s server by installing a 2TB MyBook Studio Edition external drive on my home G5, which I’ve then striped as RAID I for redundancy. (I had a hard drive fail and nearly wipe out my entire music collection earlier this year, hence the paranoia.) Onto this drive I’ve moved all of my archived data – old projects, photos, music, videos, and so on. I’ve then made that home machine accessible via MobileMe to my other two machines from anywhere. This huge amount of data is on my machine, in my office, and is (relatively) secure. Meanwhile, I’ve moved my time-sensitive data, all my currently open projects, onto a 25GB MobileMe iDisk, so that this ‘hot’ content exists in the cloud and can be, again, accessed more quickly from anywhere. This iDisk is then mirrored on the desktop machine and Mac OS X’s Time Machine application backs it up to a second external 1TB drive. Once a project is done, I remove it from the iDisk and archive it on the home machine. InDesign files, Word documents, images of quick sketches, all this new stuff exists in the cloud until it’s no longer active, and then gets shunted off to the archives. Meanwhile, more group-centric documents are handled through web applications like Basecamp and Google Docs. While I’m intrigued by online replacements like Mint for my old standby financial app Quicken, the only thing I’ve been convinced to rely upon for security reasons is the online banking suite offered by Bank of America. While it’s entirely possible – even probable – that apps like Mint are perfectly safe, in some respects I’m definitely a conservative old codger, thankyouverymuch.
This is still a relatively new experiment, and anyone interested in following its progress should let me know. It definitely has its drawbacks – for one thing, keeping massive photo and music libraries on a RAID-striped external drive can add some serious time drag to read/write speeds because it’s essentially writing everything twice, and for another, I’m still running into serious issues syncing simple things like Safari bookmark files between multiple machines. Still, it has the electric tingle of What’s Next about it, which is certainly a lot of fun, and the promise of someday upgrading my relatively heavy MacBook Pro to a MacBook Air has an appeal all its own.
Like I said before, though – this cloud is definitely still forming. In his terrific lecture to the 2007 EG Conference (made available online in a terrific TED podcast), “Predicting the Next 5000 Days of the Web”, Kevin Kelly describes the future of computing as shifting away from a mass of individual machines and towards a horde of tiny portals into one singular machine: the Internet. This is another linking node between both ucomp and cloud computing – as both our iPhones and our Nike+ shoes become I/O ports for this one singular machine, and all of this data is combined and made interchangeable, some very exciting shit will be going down. Right now, though, my iPhone doesn’t do a great job of doing much computing – but the promise is there. It’s possible the greatest buzzword of the 21st century so far is abstraction: the abstraction of content away from presentation through XML and XSLT, transmedia stories and shifting media formats, data and computing hardware. These ideas are all connected, just like these terms, just like this one machine – and that’s where our culture is most definitely, yes, converging.
Bring it on.