This is the published part of my presentation. I will make no efforts to keep this page updated. The Internet is in constant flux. Links that this document refers to might have moved, changed or vanished. Still I hope that you will find it useful. Please feel free to comment by e-mail to email@example.com
This documentation only summarizes the main points put forward in my presentation. I am currently working on a full academic paper that will provide more detail. Please check my research site for announcement and availability.
e-Commerce is the latest buzzword in the business lounge(s). Predictions vary between 4 and 10 billion USD for the business to consumer markets, and 2 to 150 (sic!) billion USD for the business to business markets at the year 2000.
My prediction today is that e-Commerce will be filled with artificial markets actors. As Nicholas Negroponte said:
most people have a hard time understanding
that the viewership of broadcast will be largely machines in the future,
Commerce Before the e. In the old days we had real markets, with real booths, real crop, real horses, real handicraft and real people. These physical places was filled with man, sound and smell. I presented a video clip that illustrated that time.
|During the marketing evolution we developed
models and a language to manage the exchange of products. In its simplest form the market
Figure 1 - A simple market model with Sellers, Buyers and interMediaries
On the way towards the e. The original markets developed into modern forms. Such as department stores, super and even hypermarkets. Of course we have preserved a few traditional market places, often for nostalgic reasons. It seems that many marketers are nostalgic when they go electronic. They design www-shopping sites that resemble original markets and physical shopping.
What they all have missed...is that e-markets might be populated by artificial actors besides real people! I illustrated this by showing how these non-human players already look like.
|Of course commerce is forever still in
e-transition. The following is but one of the attempts to catch what is going on.
My model now contains Artificial Actors in addition to the real ones outlined above.
Yes, it is hard to see who is what. And it seems that sellers look like byers, intermediaries look like sellers or buyers etc. But it is beyond the scope of this presentation to explain that here.
Figure 2 - A not so simple market model with Real and Artificial Actors
easyT is a Flight Finder found at www.resfeber.com. EasyT and its cousins found at sites operated by a number of airlines and commercial flight recommendation systems, simply shops for flights following different criteria.
I also presented a story around the evolution of shopping softbots. I demonstrated how Bargain Finder, ShopBot and Jango worked. The two latter have been shut down respectively remade.
I showed a video with Mary, a Virtual Actor being trained in Virtual Reality by a real person in a Swiss laboratory.
|Olga is a sales agent developed at the Center for Speech Technology (CTT) at the Royal Institute for Technologyl in Stockholm. Olga is focused at multimodality, and it features a full-bodied animated robot-like lady, capable of multimodal speech synthesis and body gestures, as well as a graphical direct manipulation interface||
Figure 3 - Olga, an Artificial Sellers
Then I demonstrated Shallow Red is a kind of advanced Eliza-system that informs customers about stuff in a natural written language. Collaborative Filter agents are found at bookstores such as Amazon Books and Barnes&Noble. We took a look at how the latter works.
Kasbah is an agent mediated electronic market place developed and operated by the MIT Media Lab.
Kaplan is a Constraint Satisfaction Problem agent that help its users to find colleges.
Take a look at what I think is the WWW“s next to the most wonderful site: Gabocorp. So nice for humans, useless for artificials. Who is most important, man or machine? Should the travel company spend money on site layout and graphics, or make sure that flight finder agents, such as easyT, knows about and can interpret the company offerings.
I did not have time to go into details on what artificial actors are, or how agents are defined. For introductions to agents, see: Intelligent Softare Agents, amEC, UMBC agent Web or BotSpot,
Most e-Commerce applications are developed with an old mind-set in marketing. That mind-set does not take artificial market actors into consideration. E.g. while humans appreciate wonderful graphic design, that is not understood by the artificials. Or when developers are getting high on Java, generally that type of implementation are toxic to agents. Advertisers pour money into brands, agents could not care less. You have to Imagine Artificials in e-Commerce!
For more insight, please check out
Interactive Media face Artificial Consumers and marketing theory must re-think,
published in the Journal of Marketing Communication and written by myself and Anders Lundkvist
I would appreciate any comments to firstname.lastname@example.org