Biotic interfaces imply relationships with electronic machines that are, to varying degrees, invasive. Of course, ubiquitous, lightweight biosensors that are no more cumbersome than a wristwatch or eyeglasses will be used for the most common applications. But there will still be a need for additional specific equipment. Hence the importance of the new "humanized" interfaces with computers and networks - "intelligent agents," autonomous, user-friendly, personalized expert programs, rather like dematerialized robot programs.
The cyberspace of the future will offer a plethora of information: a multiplicity of databases and interactive networks, a thousand interactive multimedia television networks, virtual libraries, audio-visual catalogues, guides, electronic reservation systems for hotels, car rentals, etc., all at the best price. We are moving towards complete saturation! It will be impossible for a poor isolated biological brain to find its way in these networks, to access these services, to use the countless passwords, keys, and codes that will be required. It will need a guardian angel, a guiding spirit, an untiring major domo, a zealous and faithful intellectual assistant - in other words, an "intelligent agent" able to navigate through the twists and turns of interconnections, sort out and select the relevant information, propose strategies for accessing knowledge, find and classify the mass of data generated by computers in their digital conversations, and negotiate with other "agents" to defend the interests of its boss.
What will these "intelligent agents" be like? This curious term designates expert programs that provide ongoing electronic assistance with all the functions of computers and networks. One paradox of computer technology is that the more computers improve, the more they require of their users. Today, pulldown menus, icons, and dialogue boxes are part of the familiar landscape of the computer screen. But the software waits passively until the user decides what he or she wants to ask of the machine. Intelligent agents, on the other hand, try to anticipate the most probable actions of the user. After a period of breaking-in and shared experimentation with their owner, they learn how to carry out routine tasks automatically. When messages are received in an electronic mailbox, an agent can decide to sort them in order of priority or transmit them to other agents. It can find information in a spreadsheet and fax it to a correspondent. Agents can negotiate with each other on the network to determine the best times to set appointments between busy managers. They can access stock-market services and follow the investments in a portfolio, selectively collect information according to the interest profile of a user, or assist the user in finding the best price of a product by looking through hundreds of catalogues. Agents will also be able to select film or television programs and read newspapers and choose interesting articles for their boss.
Intelligent agents will quickly become essential for reaching correspondents when they are needed most. Studies show that in emergencies, one call in four actually gets through, and the others get lost along the way, resulting in wasted time. Large telephone and computer companies such as ATT, Motorola, and IBM are developing intelligent messaging systems that draw on various means of communication in order to locate a person wherever he or she is; to reach the person, all one will have to do is type a message into a microcomputer. The agent then takes charge, trying all the connection possibilities: telephone, fax, electronic messaging, portable phone, or beeper.
Equipped with speech and simulated human expressions, represented as explicit icons, agents will become true intellectual assistants, sometimes amusing, ironic, or critical, always familiar, and often essential.