Stephen Baxter’s short article in the science magazine, Focus, highlights some of the problems that arise in dealing with consciousness in a scientific forum. Baxter suggests that Turing made a correct, or at least insightful decision, in deciding to aim at a test as to whether robots/computers could think like humans, while avoiding the possibly unscientific or undefined area of consciousness. This is not strictly speaking true. Turing appeared to have believed that having a robot/computer that was perceived to think like a human was the same thing as it being conscious. Baxter points out that computing is based on algorithms, systems of rules for doing calculations. He regards it as a speculation that human thinking is based on algorithms, although for most neuroscientists this is more like an assumption.
Where Turing can be argued to have gone wrong is in eliding thought with consciousness. It is apparent that we are conscious of some of our thinking and of the content of working memory, on which thinking is dependent, but to what extent it is fully dependent on consciousness is questionable. Facts, ideas, even inspirations can be argued to spring fully formed from the long-term memory, or from processing in parts of the hippocampus. Logical or mathematical reasoning can be argued to proceed by at least partly unconscious rule application. Thus there is no great problem in viewing thought as the product of algorithms. If consciousness has a role here, it is more in applying the effort involved in carrying out this type of demanding process.
But consciousness looks to be mainly active in a different area. What we are primarily conscious of is the sensory input from the external world, and our emotional/evaluative preferences in response to the perceived reward/punisher qualities of these sensory inputs. This evaluative response colours rather than determines thinking. Areas such as the orbitofrontal and the anterior cingulate project to the main reasoning/planning area of the brain, and the outputs of both the evaluative and executive areas are also brought together in the basal ganglia, which lie directly upstream of action and behaviour. While thought can be understood in terms of a system of rules for calculations, it becomes much harder to see how this could be achieved for the subjective experience of sensory inputs or the evaluation of preferences for possible subsequent behaviour.
Why did a genius like Turing not spot all this. It has to be taken into account that he lived in a period that in the whole of history may have been the least favourable for thinking about consciousness. Behaviourism dominated thinking about the brain for much of the 20th century, and this doctrine entirely excluded consciousness and emotion.
Further to this, the neuroscience of the last two decades has thrown light on neural processes, particularly the separate but intertwined processing of thoughts and emotions, which was largely unknown to Turing. The potential scientific impact of the enhanced ability to scan the brain since the 1990s has been likened to Galileo’s turning a telescope to the skies in the 17th century.Tags: algorithms, anterior cingulate, basal ganglia, consciousness, emotion, orbitofrontal, thought, Turing, Turing test, working memory Posted by