In considering Alchemy as it relates to the Internet, I have been spending my time trying to reconstruct my college readings of Walter Benjamin, the beautifully melancholy German essayist from the 30's. I took a class on the Frankfurt School, a group of intellectuals who distinguished themselves with their critique of the emerging modern consumer culture and the technology that enabled it. Benjamin was the imaginative outlier of the bunch, writing about "Hash in Marseille," the psychology of the collector and the shopping arcades of Paris at the turn of the century. I cannot recall an instance of Benjamin discussing computers, although his commentary on art in the age of mechanical reproduction as well as his citation of Kempelen's chess-playing dwarf machine, the Turk, in important ways pre-figured our contemporary situation of extreme automation.
The reason I have been rummaging through college essays and running down to St Mark's Bookshop at 11p on Sunday night is because I have this quote stuck in my head like a persistent Hall & Oates tune. I remember Benjamin stating that "Great categories are monuments in the history of civilization" and I thought that this perfectly summed up the alchemical act that relates to everything from my decision to describe a link with a certain tag on del.icio.us to the broader power of naming markets and establishing value. For the life of me, I could not confirm the quote online, and so I located the original source, which was the dense prologue that Benjamin wrote to the only book he was able to publish when alive, the Origin of German Tragic Drama. The introduction is less about any particular subject (ie Trauerspiel) than it is about the theory of knowledge in general, and the specific powers of naming. When I found the text, it actually read as follows:
The great categories which determine not only the shape of the systems, but also philosophical terminology- logic, ethics, and aesthetics, to mention the most general- do not acquire their significance as the names of special disciplines, but as monuments in the discontinuous structure of the world of ideas.
Not quite as catchy as what I remembered, but it does get the same point across: how we describe something is in itself an act of creation, beyond simply representing some external object. Something happens during the act of naming. It could be a company, an index, a domain, a category, or a database class. This is the modern alchemist, consistent with the history of the solitary materials scientist who mixed various elements into something whose value was far more than the sum of the parts.
A fine example of the Internet alchemist is Joshua Schachter, the inventor of del.icio.us. The seed of his idea, like that of many of his peers, was simple enough: I have a problem, nobody else has solved it to my satisfaction, and so I might as well figure it out and share the solution with others. As he commented to a blogger:
I started a website in 1998 or so; that kind of thing would later get named a "blog." Anyway, in the running of this site, I collected a lot of links so I could hopefully organize and post them. So I started keeping them in a file, which grew and grew. To be able to quickly find things, I started leaving one-word notes at the end. Tags. I later built a system, around 2002, that was a web-based database for storing my bookmarks. It also had tags, but was not multiuser. I'm not sure if it was 2002 or 2001, anyway. It first shows up in archive.org at 2002 though. Then later, I rewrote it multiuser, as del.icio.us. That was 2003 or so. So the motivation was mostly because I was solving a problem I had, and then I solved it for everyone."
The burning question for me in understanding Josh's unique contribution, and by extension the art of Web alchemists in general, is to what extent he figured "multiuser" behavior into his initial system? I doubt he consciously imagined a service that would feature thousands of simultaneous users posting links. Still, few developers have been able to create such a massive self-organizing with total focus and deep resources, much less when working on it part-time while employed as a developer of trading systems on Wall Street.
Ironically, Josh's citizenship in finance is useful here as it relates to the experience of George Soros, who almost 20 years ago wrote about his concept of reflexivity in the Alchemy of Finance. Soros claims that one's understanding of a situation changes the situation, and that the secret to his investing success was understanding his (and other investors') impact on what had otherwise been seen to be efficient markets.
This is consistent with what I see happening online, where meta-data (information about information) is creating significant economic value, from the many millions of Google and Overture keywords to the emerging class of Flickr, Del.icio.us and other tag-driven systems. Our browsing, clicking, searching and tagging behavior are the base metals which alchemists like Josh are turning into precious datastors. Rather than perform this magic in some sort of black box (ie like our friends in the Googleplex), Josh is transparent about his agenda:
Hmm. Mostly right now I have a big deck of cards with all the todos ("hipster pda") which I need to move to some sort of online thingy. I'd set up a tadalist with RSS feed, but it appears that the RSS feed is empty.
Mostly there's three major categories of stuff to do; Server Architecture, Things Involving Schema Changes, and Other Stuff.
For those of us living on the West side of Manhattan, Josh is our (mostly) friendly neighborhood Internet architect. He has the vision (and the perfectionism) of Louis Kahn and does not suffer fools lightly. Little is important to Josh other than preserving the simple scalability of del.icio.us as a system of collective meaning. In the spirit of my incantation of Benjamin, when I first met Josh late last year (through another emerging Internet alchemist and pal Mike Frumin) he described his project as "crystallized attention." What is del.icio.us after all but the traces of what people do with their time? These memory links, and the tags that go with them, are themselves a form of media. For me, this is the essence of Internet alchemy, and relates the complementary functions of automata, algorithms and APIs. It is also the reason why I took to Josh, encouraged him to take friendly venture money so that he could quit his job, and recently became involved in del.icio.us as an angel investor and advisor.
If we look back over the past ten years, there are handful of similar alchemical moments in the history of the World Wide Web (and many other minor moments that may have generated significant financial profit but were otherwise inconsequential in terms of establishing long term structural value such as Broadcast.com). The most memorable are probably Andreesen's browser at University of Illinois, Yang and Filo's "Yet Another Hierarchical Official Oracle" at akebono.standford.edu, Bezos' online bookstore, Omidyar's Pez-trading application, and Page and Brin's simple search box. (It is strange, no, that Andreesen and Bezos are direct investors in del.icio.us?).
In his 2002 commencement address at Tufts, EBay founder Pierre Omidyar, described what he considered to be the secret of EBay's success:
I can tell you, without the ability to prepare for the unexpected… …There wouldn't be an eBay today. The key is recognizing that no matter how convinced you are in the power of your own ideas… …Sometimes, ideas have ideas of their own.
That's certainly true in terms of system design. Almost every industry analyst and business reporter I talk to observes that eBay's strength is that its system is self-sustaining -- able to adapt to user needs, without any heavy intervention from a central authority of some sort.
So people often say to me - "when you built the system, you must have known that making it self-sustainable was the only way eBay could grow to serve 40 million users a day." Well… nope. I made the system self-sustaining for one reason: Back when I launched eBay on Labor Day 1995, eBay wasn't my business - it was my hobby. I had to build a system that was self-sustaining… …Because I had a real job to go to every morning.
I was working as a software engineer from 10 to 7, and I wanted to have a life on the weekends. So I built a system that could keep working - catching complaints and capturing feedback -- even when Pam and I were out mountain-biking, and the only one home was our cat.
If I had had a blank check from a big VC, and a big staff running around - things might have gone much worse. I would have probably put together a very complex, elaborate system - something that justified all the investment. But because I had to operate on a tight budget - tight in terms of money and tight in terms of time - necessity focused me on simplicity: So I built a system simple enough to sustain itself.
By building a simple system, with just a few guiding principles, eBay was open to organic growth - it could achieve a certain degree of self-organization. So I guess what I'm trying to tell you is: Whatever future you're building… Don't try to program everything. 5 Year Plans never worked for the Soviet Union - in fact, if anything, central planning contributed to its fall. Chances are, central planning won't work any better for any of us.
Build a platform - prepare for the unexpected... …And you'll know you're successful when the platform you've built serves you in unexpected ways. That's certainly true of the lessons I've learned in the process of building eBay. Because in the deepest sense, eBay wasn't a hobby. And it wasn't a business. It was - and is - a community: An organic, evolving, self-organizing web of individual relationships, formed around shared interests.
Clearly, Omidyar created EBay within an architecture of public participation. Listening to him, as well as to Josh, it would seem that there is simply no way to create long term sustainable value online without engaging consumers in the act of media production. And yet one could argue that Yahoo! and Amazon do not necessarily do this. Yahoo! is working hard to incorporate community production and meta data into their core platform (cf. Flickr acquisition, Yahoo360, etc) and Amazon leverages ratings, reviews, wish lists, associates and other community features to create a richer experience for users. Still, these are not as fundamental to the browsing and shopping experience respectively.
With regard to Google, it is hard to figure out exactly to what extent they leverage the community at large and and to what extent they are simply providing this community with cheaper storage, better tools and faster search. Community in Google is not Orkut, their attempt to compete with Friendster which was aborted mid-stream when the threat evaporated. Rather, the Google community folds in on itself as user search behavior drives keyword purchases by advertisers (AdWords), which suggests which content offers the highest effective value in the form of context-sensitive ads (AdSense) which users can leverage profitably for their own blogs (Blogger). One of the first examples of this was Michael Buffington's Asbestos Blog. I have not met Michael, although from reading his blog I have a lot of respect for his alchemical curiosity and ability to execute interesting tests. He expresses melancholy with the knowledge that these strategies are, as in any advanced liquid market, moving into the hands of speculators who are far more focused on the gold than they are on the magic of base metal transformation.
"We'll take ad inventory that costs 50 or 75 cents, buy it in bulk, and turn it into gold by targeting $6 or $15 precision ads there," he said. "We'll be the alchemists."
Without further commentary, this is the natural segue into a discussion of Arbitrage which will complete the Media Futures series.