We believe that, over the course of the next decade, the global economy and the Internet will merge into an information economy bustling with billions of autonomous software agents that exchange information goods and services with humans and other agents. Agents will represent--and be--consumers, producers, and intermediaries. They will facilitate all facets of electronic commerce, including shopping, advertising, negotiation, payment, delivery, and marketing and sales analysis.
In the information economy, the plenitude and low cost of up-to-date information will enable consumers (both human and agent) to be better informed about products and prices. Likewise, producers will be better informed about and more responsive to their customers' needs. Low communication costs will greatly diminish the importance of physical distance between trading partners. These and other reductions in economic friction will be exploited and contributed to by software agents that will respond to new opportunities orders of magnitude faster than humans ever could.
The agents we envision will not be mere adjuncts to business processes. They will be economic software agents: independent, self-motivated economic players, endowed with algorithms for maximizing utility and profit on behalf of their human owners. From other agents, they will purchase inputs, such as network bandwidth, processing power, or database access rights, as well as more refined information products and services. They will add value to these inputs by synthesizing, filtering, translating, mining, or otherwise processing them, and will sell the resultant product or service to other agents. In essence, these agents will function as miniature automated businesses that create and sell value to other agents, and in so doing will form complex, efficient economic webs of information goods and services that respond adaptively to the ever-changing needs of humans for physical and information-based products and services. With the emergence of the information economy will come previously undreamt-of synergies and business opportunities, such as the growth of entirely new types of information goods and services that cater exclusively to agents. Ultimately, the information economy will be an integral and perhaps dominant portion of the global economy.
Today's Internet is inhabited by a relatively modest assortment of software agents, some of which play useful roles in electronic commerce. Two familiar examples are shopbots, such as those at mySimon.com and DealPilot.com, which collate product and price information to aid shoppers, and recommender systems, such as the one at amazon.com, which automatically suggests potentially interesting items to shoppers. Almost all such agents interact directly with humans, and almost none of them behave as economic decision-makers in their own right. Given this state of affairs, how and why do we expect the information economy to emerge?
The transition to the information economy will begin as an evolutionary step. The tremendous pressures that have fueled the rapid growth of electronic commerce in the last few years will continue to drive automation, and some of this automation will be encapsulated in the form of software agents. As they grow in sophistication and variety, software agents will begin to interact, not just with humans, but with one another. Interactions among agents will be supported by a number of efforts that are already under way, including standardization of agent communication languages, protocols, and infrastructures by organizations such as FIPA [9] and OMG [21], myriad attempts to establish standard ontologies for numerous products and markets (CommerceNet [5] being one prominent player in this arena), and the development of various electronic payment schemes such as IBM MicroPayments [13].
When interactions among agents become sufficiently rich, a crucial
qualitative change will occur. New classes of agents will be designed
specifically to serve the needs of other agents. Among these will be
``middle agents'' [6] that provide brokering and
matchmaking services. By and large, middle agents (and in fact
virtually all agents) will charge for their services, both to
prevent abuse of their valuable resources and to provide income for
their human owners.
Even functions that are not directly associated with commerce, such as
search engines or more sophisticated computational services offered by
application service providers, may be represented by agents that
charge other agents or humans.
Just as today's global economy functions as a decentralized mechanism that adjudicates and satisfies the myriad conflicting needs of billions of human economic agents, it seems plausible that the information economy could provide coherency to an even larger population of economic software agents. However, it would be dangerous to assume that theories and intuitions based on centuries of experience with human economic agents will be directly applicable to understanding, anticipating, and controlling the behavior of markets in which software agents participate. In effect, we are contemplating the release of a new species of economic agent into the world's economy -- an environment that has heretofore been populated solely with human agents. This new species may be created in our image, but it will not behave exactly as we do. Economic software agents will make decisions and act upon them more quickly than humans. While they may be more expert at certain narrowly-defined tasks, they are likely be less generally capable and flexible than we are. Before unleashing them on the world's economy, it is essential that we consider their economic incentives and behaviors very seriously, and we must use every available means to anticipate their collective interactions with one another and with us.
This survey reviews several recent results obtained by
collaborators affiliated with the Information Economies project at IBM
Research on collective interactions among agents
that dynamically price information goods and services. It describes
how we have used the tools of modeling, analysis and simulation to
provide insights into the interplay among dynamics,
optimization, and learning in agent-based economies. This work
represents one aspect of a more comprehensive effort to understand and
foster the transition to a global agent-based information economy. The
Information Economies project as a whole encompasses theoretical work
on agent-based economies, development of software infrastructure
and componentry for supporting
economic agents, and cooperative work on designing industry-wide
standards for interoperability of software agents and multiagent
systems.
Section 2 gives an overview of recent work on dynamic pricing by agents, and sets the stage for a detailed investigation of the dynamics of an economy of shopbots and pricebots, presented in section 3. In section 4, we move beyond simple pricing, showing how agents might base decisions concerning product attributes or configurations on economic considerations. We shall conclude in section 5 with a summary and a brief preview of our future research on information economies.