Benefits of Anticipations in Cognitive Agents

Abstract

This book proposes the anticipatory approach for the design of cognitive agents. That is, we propose to implement anticipatory mechanisms in artificial cognitive systems to generate highly flexible and adaptive systems that can efficiently cope with dynamic environments. While the first two chapters identified several different facets of anticipatory mechanisms, this chapter focuses on the benefits and potential drawbacks of using these types of anticipatory mechanisms in cognitive systems. We focus hereby on the explicit forms of anticipation defined in the previous chapter.

Thus, this chapter intends to answer questions such as: “What are the specific advantages of anticipatory mechanisms and capabilities?”, or “Why should we endow artificial cognitive agents with anticipation?”, as well as, “Are there any disadvantages in doing so?”

This chapter first identifies three general potential beneficial aspects of anticipations and then provides concrete examples of the impact of anticipation in several cognitive functions such as motor control, learning, attention, and social interaction. Section 3.3 points out that anticipations might not always be advantageous and may sometimes interfere with effective behavioral execution—especially when the involved predictions are inaccurate or not synchronized well.

Finally, we summarize and draw conclusions.

Man shoots an arrow into the future with a chord attached. The arrow fixes itself in an image, and he hauls himself toward it. Paul Valery