This commentary has two main aims: First, it aims to reconstruct the major important predictions and claims Eliasmith presents in his target article as well as his reasons for endorsing them. Second, it plays its own version of “future games”—the “argumentation game”—by taking some suggestions presented by Eliasmith maximally seriously and then highlighting problems that might arise as a consequence. Of course, these consequences are of a hypothetical nature. Still, they are theoretically relevant for the question of what will be needed to build full-fledged artificial cognitive agents.
Chris Eliasmith discusses recent technological, theoretical, and empirical progress in research on Artificial Intelligence and robotics. His position is that current theories on cognition, along with highly sophisticated technology and the necessary financial support, will lead to the construction of sophisticated-minded machines within the coming five decades (Eliasmith this collection, p. 2). And also vice versa: artificial minds will inform theories on biological cognition as well. Since these artificial agents are likely to transcend humans’ cognitive performance, theoretical (i.e., philosophical and ethical) as well as pragmatic (e.g., legal and cultural laws etc.) consequences have to be considered throughout the process of developing and constructing such machines.
The ideas Eliasmith presents are derived from developments in three areas: technology, theory, and funding; and I will demonstrate the background assumptions underlying these. In this way, I want to demonstrate that if we read Eliasmith as defending a formal argument (rather than a thought experiment), this argument has the form of a petitio principii. To illustrate this very clearly, a formal reconstruction of the (not explicitly endorsed, but implicitly assumed) arguments will be conducted. I then argue that even though they are constructed as arguments, and Eliasmith’s claims fail, his suggestions provide an insightful contribution to the philosophical debate on artificial systems and the near future of related research. I further want to stress that we should perhaps confine ourselves to talking about less radical alternatives that do not necessarily include the mindedness of artificial agents, but have some element of biological cognition (architecture or software) in them. A number of subordinate questions have to be looked at in order to arrive at a point where a justified statement about the possibility of phenomenologically convincing artificial minds can be made. These considerations include more possibilities than simply the dichotomy of human-like vs. artificial. This is due to our having to think about possibilities that lie between or beyond these two extremes, such as fragmented minds and postbiotic systems, since they might soon emerge in the real world. The way in which these will be relevant to philosophy will be largely a question of their psychological make-up—most notably, their ability to suffer.
To start with, the following two sections will present some relevant aspects of the position expressed in the target article. They will summarize, and highlight some of the article’s many informative and noteworthy suggestions. I shall also bring in some additional thoughts that I consider important. Afterwards, I will play a kind of future game of my own: I take Eliasmith’s predictions very seriously and point at some of the problems that might arise if we were to take his suggestions as arguments. To be fair, Eliasmith himself says that what he presents are “likely wrong” predictions (this collection, p. 3). So on a more charitable reading, his claims are not intended to be arguments at all. Yet the attempt to reconstruct them as a formal argument has the advantage of showing that his claims are based on a reasoning that is itself problematic.