Bryson was a PhD student in the MIT Artificial Intelligence (AI) lab in the late 1990s. The paper is a highly personal view of the Cog project at that time. In addition to various observations on AI in general, and behavior-based AI in particular, the author offers brief observations on science, philosophy, Rodney Brooks, robotics, and, in particular, the influence of Daniel Dennett on the Cog project. Apparently, Dennett was quite involved in the project, and interacted with both students and faculty. The title of the paper is based on a metaphor that Dennett used to describe the ability of consciousness to focus on selected aspects of the environment.
There are statements in the paper that are debatable, like, “Philosophy is the least mechanistic, but perhaps the most honest means of enquiry,” as compared to experimental science and logical reasoning. There are many comments on intelligence (the topic of the author’s dissertation and current research), but the word is not defined here. Even the statement, “Intelligence is what defines humanity,” is questionable, since it excludes animal intelligence. The author also states that readers of the paper probably don’t believe that computers will become intelligent in our lifetime. Again, without a definition of intelligence, this is a highly debatable statement.
In the last paragraph of page 22, I believe that the words “expect” and “except” are interchanged, which makes it meaningless. One more observation: page 24 contains the observation, “But, unfortunately, the brain is actually very complicated.” This we knew before reading the paper.
In summary, this collection of personal observations on the development of Cog, and the effect of Dennett on the process, is an important part of the history of humanoid robot development. It is recommended reading for students of robot intelligence.