By Sean X Cummings and Adam Broitman
There has been an incredible buzz over the coming of
Sprocket technology, and many people we have spoken with agree that it will
fundamentally change the way we approach the interactive space at large. Much
in the way the explosion of Facebook enabled breaking down the barriers around
content and engagement as it relates to mining your casual friend network,
Sprockets will, for the first time, provide us with a holistic picture of what
our consumers are doing online, while providing the first truly intelligent
data-mining agent for ourselves.
More interactive than widgets
Sprockets are not widgets 2.0 any more than Facebook is
MySpace 2.0. They are, however, a technology that finally moves beyond the
two-way conversation with consumers, creating an enhanced, multi-directional
conversation among the consumer, company, site, data and -- more importantly
for advertisers -- among the Sprockets themselves.
Allow us to get technical here for a minute. Sprockets use
passive preference profiling (PPP) and interpretive semantics (IS), giving them
the ability to learn. What do we mean by that? Well, remember all the talk we
heard about "intelligent agents" or Google's dream of
"artificial intelligence?" Sprockets are the first applications to
truly approach that level of automated, interpreted content collection.
The beauty does not stop there. If all Sprockets did was
provide consumers with a way they could accomplish more online in less time,
and have it be more relevant, it would be a substantial breakthrough. The
beauty of Sprockets is that they provide advertisers, marketers and analytics
groups with the same type of automation through semantic interpretation; the
data, however, is passed along in the reverse direction.
Most companies are keeping their current strategy regarding
Sprockets fairly quiet, but by Q3 2008 anyone with a decent online presence
will be pushing ahead full-steam. If your company has not figured out how to
implement Sprockets, or you don't have a full-time Sprocket strategist on staff
by the end of the year, you are almost definitely going to be on the tail end
of this beast, and this is one place where you don't want to be riding the long
tail.
Benefits: real or imagined?
We are often asked, "Are Sprockets really going to
change the economics of the internet, or are they just another overhyped
technology?" and "What's so game changing about Sprockets?"
First: we have seen Sprockets work. We have seen them work
in the real world, not a demo. We have observed as Sprockets learn and grow in
complexity over the first two weeks of usage.
Second: Sprockets define simplicity. Anyone can develop a
Sprocket. The Sprocket Developer's Kit (SDK) has a simple drag-and-drop,
single-screen user interface. The user just connects tubes, called
"synaptic tubes," linking their basic interest points. This process
takes 15 minutes, tops!
From there the Sprocket takes over, filling in the gaps with
your internet persona while tweaking, augmenting or modifying the few data
points given. It does this by monitoring surfing habits, online purchase
activity and time spent on content, semantically absorbing the it. The genius
is that Sprockets do not store all this data. They retain meta-types of the
data in real time. In essence, every page you hit is crawled, indexed and
semantically related to your entire history. It is then meta-typed and
discarded, retaining only the meta-data.
An open-source legend is born
This is one of the most important developments to ever come out
of the open-source community. The initial kernel was programmed by Alexander
Dorsay III over the summer of 2006; however, community did not adopt the
concept initially. It wasn't until Anya Khait, an environmental architecture
student with a Ph.D in genetics decided to apply techniques and theories from
the Biomimicry Institute to the Sprocket kernel that the true power of
Sprockets was realized. She developed the "synaptic tubes" and the
core engine behind the algorithm modifier based on the growth of Ivy, which is
able to grip around any surface and adapt to its environment.
Without getting too deep into the science, let's say that
Sprockets, by design, have a "public you," and a "private
you" setting. By nature, most consumers exhibit a personal dichotomy --
their private and public selves. With some, those two personas are almost
identical; with others they are quite divergent. Other Facebook-style
properties treat you as one self, but you are never just one. The Sprocket
dynamically adapts to your activity, not the other way around. You do not
choose to tell it "I want to be private today." The Sprocket
dynamically adapts "on the fly" to activity that obscures your
secondary profile. They are not separate, but they are distinct. A single
synaptic tube links the two personae as it grabs information.
When a Sprocket shares any of the data in the outward
direction the entire profile is made anonymous, even down to erasing IP
histories, or what is referred to as digital public memory. The intelligence
sits in a data-hash in the Sprocket, changing the internal Sprocket algorithm.
That is what is different about Sprockets, the algorithm itself dynamically
adjusts; not just parameters, but the fundamentals of the algorithm.
The same Sprocket you develop for a website, can be used on
any mobile device, or converted with the Sprocket converter to any computer as
a standalone app -- although you will lose much of the semantically dynamic
algorithm modifying aspects of Sprockets that way. They are what the promise of
Java was supposed to be before Microsoft derailed it with their own version.
However, since Sprockets were built on the foundation of the web, that fear is
somewhat mitigated. In fact, it takes advantages of some of the same AJAX
technology framework that Web 2.0 does.
Sprockets are not AI
They are merely a highly sophisticated algorithm that uses
the collective intelligence as a decisioning engine. The advantage is they
don't make the mistake of using the collective intelligence of the web as it
exists now. The sheer volume and mess of the organization of website content
makes the collective intelligence there somewhat. well, stupid. If it wasn't,
you wouldn't need search engines like Google or Ask.com to make sense of it.
Instead, they use the collective intelligence of the
"Sprocketosphere," which uses only the intelligent decisions in the
data, and not the data itself.
Take a look at comparisons to other popular platforms:
- Standard website: one-way pull request from the consumer
- Rich media: one-way pull request from consumer, with
engagement - Widgets: customization by consumer and selective
identification of interest - Sprockets: multi-directional conversation with consumer and
website, customizable ad tracking, alignment checking and reporting, semantic
web gathering and inter-Sprocket communication.
The user no longer has to request the web pages; the
Sprockets talk amongst each other, among the web, among your email, SMS,
address book and social networking to dynamically deliver to you the zeitgeist
of you. When you choose to surf manually, the Sprockets learn your behavior and
communicate in-between the other Sprockets to adjust your "algorithm"
on who you are and what your interests are online. Best of all? They can do it
with the existing structure of the web, requiring no more tagging or special
codes to be inserted.
We have heard many promises of the next 'big' thing. We have
heard about many hyped technologies. Let's hope that Sprockets don't go the way
of Cold Fusion. In the end Sprockets can only be as intelligent as the user
they are learning from. Hopefully, the users reading this article will be one
step ahead.
Sean X and Adam were so taken with the potential of
Sprockets that they formed SprocketX.com. An open-source company that will
produce tools to leverage the potential of the Sprocket kernel.
Technorati Tags: sprockets, media, widgets, social media, Myspace, facebook, adam broitman, sean cummings