1.3 Background: Autonomous and Intelligent Systems
Here is some background on Autonomous and Intelligent Systems.
- Automaton: A machine designed to operate either wholly or partly by itself. Historically, such machines incorporated control mechanisms, which were designed to automatically follow a sequence of operations, perhaps in response to some predetermined input (e.g. instructions of some kind), and in this way these machines appeared to be able to perform quite constrained kinds of actions, such as answering ‘yes’ or ‘no’ to specific questions. Well-known examples include the bellstrikers in mechanical clocks, which appear to be self-guiding, whereas they are simply following a specific set of sequences (admittedly at opportune times, such as the beginning of every hour). Importantly, particularly for this course, is that the notion of automaton is often associated with machines built to resemble human or animal motion, typically as a form of entertainment.
- Robots: These machines are typically more complex in their design and construction, but also, they can often do far more complex tasks. Moreover, they have an additional dimension of capability when made to be programmable. By incorporating this generalising capacity of modern computers, robots are in principle endlessly extensible in terms of the behaviour they can be programmed to perform, potentially enabling them to model all manner of complex sequences of actions automatically. Note that the name derives from the Polish word robota, meaning ‘slave’ – recall the discussion immediately above about concerns around the use of AI technology to build so-called ‘intelligent’ assistants, where one possible historical connection to our current conceptions of AI stems from the vision of artificial creatures, which are essentially servants or slaves in earlier cultures (e.g. Homeric Greece).
Now that you have some historical context for understanding the place of AI technology in society and culture, let’s step forward to the more familiar, relatively recent history of AI, as it arose in the early decades of last century. While AI resulted from activity within and across a number of disciplines over many years, particularly Mathematics, Engineering (including foundational work in what would eventually emerge as the discipline of Computer Science), and even Psychology, there were a number of milestone events which were seminal in determining how the field itself eventually formed.
Watch the followingon the early history of AI in the twentieth century, from 1 minute 42 seconds to 16 minutes 30 seconds. Now choose only one of the questions below about key events and people in the history of AI presented in this video. Write an answer to your chosen question, making sure to justify it (e.g. with examples, case studies).
Make sure you watch the complete video excerpt before writing your answer, as you may find some of the later examples and cases studies useful.
- What was the question that Turing asked? Try to justify your answer to this question based on you current understanding of how much progress has been made in AI since Turing devised his ‘test’ of AI.
- Do you think ELIZA was a good demonstration of intelligence? Explain your answer. You might like to use your answer to reflect on the similarities between ELIZA and modern chatbots e.g. does this comparison help to think about how much progress we have made since ELIZA?
- Spend some time trying to find out what you can about the ‘AI Winter’. Do you think we are headed toward another such winter? Justify your answer (perhaps comparing recent progress to that around the last AI Winter).