Skip to content →

Category: real time web

The Form of the Question: The Form of the Answer

Published December 6, 2008 by cgerrish

To what extent does the question establish the possible ground from which an answer can emerge? Does the shape of the question determine the shape of the answer? What happens when the question doesn’t match the subject?

Search is a query against a fixed set of data. To achieve depth the volume of data must be enormous. What happens when you search a real time stream? It’s a batch query against a stream of data. There are two common examples:

  • Getting a quote on a stock during market hours on a 15-minute delayed basis
  • Getting a quote on a stock during market hours on a real-time basis

Each is just a snapshot;  a moment in time. The 15-minute delayed quote isn’t information you can trade on. The moment for action has long since passed. The real-time quote is almost time you can trade on– but it’s still just a snapshot. A trader has a live quote that changes as trades hit the consolidated tape. The quote changes in real time without an additional query. The live quote gives additional color, one has a sense of the volatility and direction of price.

Map of the Market

Now think about the difference between search and track. with regard to Tw*tter and the micro-messaging stream. If you’ve ever used track via IM you’ve experienced the difference between a snap quote and a live quote. Imagine if you had a watchlist of your track terms that you could see change in real time. A trader can transact on any ticker she tracks– that means both reading and writing. This gives you a sense of some of the possible user interfaces, as well as the economics, of the micro-messaging stream.

Comments closed

A Note for Karoli on the Production of Value

Published December 3, 2008 by cgerrish

High heel

Karoli Kuns and I exchanged some comments about a recent NewsGang Live where the idea of “value” was discussed. In particular, how one might decide to purchase a pair of shoes. The conversation on the show swirled around the functional and exchange value of shoes. Let’s assume we all buy shoes that fit. Sometimes we buy shoes that are comfortable. We assign value to shoes, or almost anything, based on a number of factors.

For an interesting angle on the production of value, we can turn to Jean Baudrillard. He wrote that there are four ways of an object obtaining value. The four value-making processes are as follows:

The first is the functional value of an object; its instrumental purpose. A pen, for instance, writes; and a refrigerator cools. Marx’s “use-value” is very similar to this first type of value.

The second is the exchange value of an object; its economic value. One pen may be worth three pencils; and one refrigerator may be worth the salary earned by three months of work.

The third is the symbolic value of an object; a value that a subject assigns to an object in relation to another subject. A pen might symbolize a student’s school graduation gift or a commencement speaker’s gift; or a diamond may be a symbol of publicly declared marital love.

The last is the sign value of an object; its value within a system of objects. A particular pen may, whilst having no functional benefit, signify prestige relative to another pen; a diamond ring may have no function at all, but may suggest particular social values, such as taste or class.

Take a pair of shoes and assign a percentage of the total price to each the categories of value. The largest number will probably be next to the sign value. The shoes may have a value within the fashion system of shoes; within the fashion magazine system; within the designer system; and most importantly shoes have a signaling value within our social group. All of these things play in to the price we’re willing to pay.

Now think about the role of a social graph through Tw*tter as a method of signaling value. Compare it to using Search. PageRank uses citation as a method of deriving the value of a link. Citations are painstakingly extracted from spidered and indexed web data. Think about a real-time market where value is established through citations (gestures) across overlapping rings of social graphs. Sign value is largely produced as a social process. The key is: the gesture market needs sufficient volume and market makers.

One Comment

Track is Expensive

Published November 18, 2008 by cgerrish

Risk

Saying that “Track” is expensive is another way of saying that it’s risky. It’s saying you haven’t solved enough of the technology problem to proceed. Given the circumstances, the current price is too high for the projected revenue.

However, the absence of track isn’t a static problem. A moment will arrive when a player will take the plunge and dominate the market. They will have put enough of the pieces together to see how the story ends. The first one in will have access to the most information about what track means, how to implement it and how people use it. That means they’ll be able to optimize the service faster than followers in the space. As time passes, the price of the technical solution will go down, the revenue opportunities will go up and cash flow will move from investment to revenue.

The question is, at what point do you enter the game? Can you wait for a set of economics that are cash flow positive? Of course, by then, usually the game is over. Unless of course, there’s only one player in the game. Then the cost of waiting is zero. Riskless competition isn’t competition.

2 Comments

Tracking Voices: Attack, Sustain, Decay

Published November 16, 2008 by cgerrish

Radar read out

In trying to understand something like “Track,” I find that as a new angle is uncovered I need to make note of it before it slips back into the aether. Part of understanding is explaining something to yourself, and then trying to explain it to someone else. It’s turning a shard of a mental image into a story. Understanding the signal-to-noise ratio in that transmission is one measure of success. Sometimes a transmission can carry the payload of a dense and ambiguous metaphor— something that is neither signal nor noise.

Noise is something we can’t or don’t want to understand. Signal is communication for which we already have a framework for understanding. Ambiguity is a different kind of payload in a signal. Sometimes it’s important to drive toward clarity, other times it’s important to let something remain in an ambiguous state and allow for the meaning of play and play of meaning to unfold. The usefulness of track is something largely undiscovered. The tools we use to track the idea of Track are both primitive and highly sophisticated. We talk to each other; we listen; and then we talk to each other some more.

Measuring the decay of sound

The small piece of the picture that came into focus for me today was the distinction between “who” and “what.” Distinguishing “track” and “search” seems to have some conceptual value. Search is more associated with what; track is more associated with who. Either can be used for the purposes of the other, but there’s some value in making this distinction.

There’s a sense in which track can be used to understand the current presence status of a person on the Network. We use a status indicator on IM to indicate to our personal network of reciprocal connections our level of availability. Tracking a person or a topic keyword tells you who is currently speaking on the Network. Who, not what. Speaking, through microblogging (tweeting), is a form of indicating your presence and availability.

An essential component of track is its basis in the real-time stream. One way we make conversation is through making sounds– and sounds have a physics. Finding the presence of speakers must occur within the context of the sound envelope— track must do its work in the period starting at the end of the sustain and finishing at the end of the decay.

The decrease in amplitude when a vibrating force has been removed is called decay. The actual time it takes for a sound to diminish to silence is the decay time. How gradual this sound decays is its rate of decay.

Once the sound envelope has completed its decay, the presence of the speaker can no longer be assured.

A directed social graph, or affinity group, can be followed to understand current presence status. Track can also be used for that purpose, and additionally to discover new speakers on the subject of one’s affinity. Condensing value out of that stream returns us to the beginning. A story emerges, a melody emerges– from the attack, sustain and decay– of the voices in the stream. A thousand flowers bloom in an eternal golden braid.

Comments closed