The artificial intelligence debate
HSV Public Lecture by Professor Carl Mahoney at Balwyn Library on 22 October 2015
Artificial intelligence is now well embedded in many areas of the economy and society. It is evolving at a rapid pace and is controversial because of its involvement in crime and warfare. Carl indicated some of the areas where artificial intelligence is currently being used and some of the problematic aspects.
A brief history of computers
Plants, animals and man all have a very sophisticated computer system in their genomes. Capable of developing and running all living cells, the genetic ‘computer’ is still only partially understood. We now know that parts of the genome, once called ‘junk DNA’, and now referred to as epigenetic material, play an important role in controlling the function of genes, switching some on and others off. It is possible that in the future we will discover how the epigenetic material is controlled and perhaps have a quite different and expanded understanding of the roles that genes themselves play.
A few of the notable innovators in this field include;
- Charles Babbage (1791-1871) designed the first man-made computer, a programmable calculator. Ada Lovelace, a mathematician, devised the program.
- Alan Turing (1912-1954) wrote a seminal paper in 1936. This stimulated intensive interest in developing a digital computer in Britain, Germany and America, especially at the outbreak of the second world war.
- John van Neumann (1908-1964), a mathematician and Hungarian emigré, worked in the USA on the architecture of computers.
- Norbert Weimer (1894-1964) published a book called Cybernetics in 1948, showing widespread applications in medicine and biology.
- In the 1960s the internet was set up as a collaboration between universities. To have your computer linked up you had to apply to the Vice-Chancellors Committee! In the 1970s Tim Berners-Lee wrote a memo foreshadowing an international collaboration, or ‘world-wide web’. The development that followed was phenomenal, as we know.
The human brain
Artificial intelligence aims to copy the function of the brain as well as the function of the mind. The brain has an intricate network of neurones which branch and interconnect, transmitting information and storing memory. The sensory input from our eyes and ears can be analysed in engineering terms as a set of finely discriminated set of frequencies. By contrast, consider the semantic network of the mind, which is all the time developing associations between ideas and data and testing it deductively.
‘Big data’ systems operating today have encroached on our privacy, with many of the sophisticated modern devices which we rely upon becoming de facto surveillance machines. Our mobile phones transmit a constant stream of data about our activities, and modern cars are equipped with computers which are linked to a central system. Other areas in which large data systems are operating include the stock-market and astronomical laboratories. Quantum computers, which are even faster than digital electronic computers, are virtually ready for commercial use.
Science and technology are now so closely entwined that many scientists think and work like engineers, with very close collaboration as well, frequently. Mining is a field which relies heavily on automation as well as artificial intelligence. Although some drilling is still done in order to sample ore deposits, much prospecting can now be accomplished by remote data analysis. Many mines also use driverless trucks to transport loads weighing several hundred tonnes. These are monitored by a manned central command system which may be thousands of kilometres away from the mine. Other mine truck-drivers wear caps fitted with sensors able to detect fatigue and the driver’s risk of falling asleep. In the future it is likely that we will have miner-less mines and also driverless trains.
Robots are becoming increasingly sophisticated and are being used in many fields, such as factories and personal care. They can be programmed to function independently or, as is the case with surgical robots, be controlled by the surgeon who is nearby directing the procedure.
Transhumanism, superintelligence and singularity are founded on the idea that we are moving towards a point where computers will be as intelligent as people, thus capable of eliminating aging and disease, controlling mood and motivation and be self-replicating in some situations. However, although computers roughly follow Moore’s law or observation, more strictly speaking – i.e. that their capacity doubles every 18 months, they are not necessarily becoming more intelligent. Furthermore, human intelligence is increasing, in step with technological advances.
The ethical issues surrounding artificial intelligence are widely debated, mainly about the uses to which it is put and the unintended consequences, for example, in warfare, crime, punishment and human implants. A group, Campaign to Stop Killer Robots, led by Nobel Peace Prize winner, Jody Williams, made a submission to the UN in October this year, seeking a ban on the development of armed autonomous devices which can be programmed to kill without human oversight.
In summary, there are many exciting, wonderful innovations in this field. However, they bring a host of ethical issues in their train and we need to be ready to address and debate these.
Report by Jennie Stuart