The Imagination Era
The development of AI began with dreams. Pamela McCorduck has traced several routes to AI: the route of imagination – what might be; the route of philosophical inquiry – the bridge between imagination and what is; and the route of science – AI as it has been realized since the development of computer programming. In the ‘imagination era’, AI was reflected in the concept of gods that could create non-humans to protect or threaten humans. Stories in Greek and Egyptian mythologies stand as testaments to humanity’s imagination. However, such imaginative exercises were only the beginning. From the sixteenth century on, a population explosion of automata took place and what people had perceived as the power of gods (or magic) came to life in the form of toys for rich children who could afford them.[1] The invention of the mechanical duck in 1738 by Jacques de Vaucanson symbolizes the automata stage. The mechanical duck did many things its live counterpart could do, like swing his wings, drink water and eat grain.
In the seventeenth and eighteenth centuries the art of mechanical statues flourished and, by the nineteenth century, the AI ‘that penetrated and dwelled in people’s imagination were composed of the printed word rather than wood and metal and cloth.’[2] Through books and stories, people learned more about the potential of machines and robots in the near future. Olimpia and Frankenstein are only two examples.[3]
Now, we are climbing the steps toward the next route: the singularity route. As indicated by scientists, soon after AI reaches human-level abilities, there will be no limit to their continued progress. This route is unique since, in many ways, humans may not be able to control the pacing or be able to turn back. From reaching human intelligence, AI is projected to move on to super-intelligence within 2 to 30 years.[4] As Frodo and Gandalf learned when they entered The Doors of Durin[5] no one can tell what we might find: humanity’s dreams or nightmares might come true.
The Singularity is Near!
In his popular book, The Singularity Is Near: When Humans Transcend Biology,[6] Ray Kurzweil reflects on the future of humanity in the upcoming decades of meta-ideas and mass information. Kurzweil defines this as a singularity era: a period during which the pace of change will be rapid and deep, in a way that human life will be transformed irreversibly. Since the 1997 victory of IBM’s Deep Blue computer over chess champion Garry Kasparov, we have witnessed significant achievements in the progress of AI abilities. In 2011, IBM’s Watson won the trivia game show Jeopardy! and DARPA invented CALO (Cognitive Agent that Learns and Organizes), which led to Apple Inc’s Siri. In early 2016, an AI computer defeated the human master in the game of GO,[7] and, in January 2017, the AI computer program Libratus beat the best poker players in the world, winning more than 1.7 million dollars’ worth in chips.[8]
Furthermore, machines outperform humans not only in games but also in the labour market. Algorithms are increasingly used in different fields, such as insurance, finance, human resources, and medicine. In light of those developments, Kurzweil argues that information-based technology will engulf all human knowledge and skills in the upcoming decades, ‘ultimately including the pattern-recognition powers, problem-solving skills, and emotional and moral intelligence of the human brain itself.’[9]
In a recent paper, Vincent C. Müller and Nick Bostrom embarked on an intriguing and difficult journey to determine when AI will reach singularity.[10] Müller and Bostrom questioned 550 experts on their prediction for the future of AI. In order to avoid deviation in the results by using biased AI terms, Müller and Bostrom defined AI as a high-level machine intelligence – ‘one that can carry out most human professions at least as well a typical human’. [11] Surprisingly, the majority of the experts asserted that AI is likely (over 50%) to reach human ability by 2040-50, and very likely (with 90% probability) by 2075.
What Lies Ahead
AI presents a challenge to intellectual property law. Consider, for instance, Hegel’s personhood theory: for Hegel, ‘property is the first embodiment of freedom and so is in itself a substantive end.’[12] Hegel’s philosophy might advocate for the legal standing of AI as part of accepting its personhood and autonomy rights. Incentive theory is a different challenge. Granting exclusive rights to authors imposes social costs on the public that can only be justified ‘to the extent that they do on balance encourage enough creation and dissemination of new works to offset those costs.’[13] In the AI era, incentive theory might lose its substantive grip since it, allegedly, will not apply to AIs. Additional studies have claimed that ‘psychological and sociological concepts can do more to explain creative impulses than classical economics. As a result, a copyright law that treats creativity as a product of economic incentives can miss the mark and harm what it aims to promote.’[14] Clearly, the internet has brought new and different variables into play. Although these differences themselves might not be relevant to AI, the AI era may potentially herald similarly divergent reasoning that could change, diminish, or extinguish existing IP theories.
Aviv Gaon is a PhD candidate at Osgoode Hall Law School. His research explores the question of whether AI creation deserves to be protected by copyright law and, subsequently, to address the current legal discussion considering the standard of copyright protection.
[1] P McCorduck, Machines Who Think (2nd ed. 2004) 13-14.
[2] McCorduck, ibid, 15.
[3] E. T. A. Hoffman’s story – The Sandman (1815), which introduced the Robot (or automaton) Olympia; Mary Shelley Frankenstein (1818); See also – Isaac Asimov, The Rules of Robotics (1950).
[4] V C Müller C and Nick Bostrom, ‘Future progress in artificial intelligence: A survey of expert opinion’, in Vincent C. Müller (ed.), Fundamental Issues of Artificial Intelligence (Springer Berlin 2016) 553-571.
[5] The Doors of Dorin, known also as the west door of Moria forms the entrance to the Dwarf city of Khazad-Dum in the J R R Tolkien’s, Lord of the Rings trilogy (1955).
[6] R Kurzweil, The Singularity Is Near: When Humans Transcend Biology (Penguin 2005).
[7] Go is an abstract strategy game invented in China 5,500 years ago. For further reading see – https://en.wikipedia.org/wiki/Go_(game). B Alarie, A Niblett & A H Yoon, ‘Focus Feature: Artificial Intelligence, Bog Data, and the Future of Law’ (2016) 66 UTLJ 423.
[8] ‘Oh the humanity! Poker computer trounces humans in big step for AI’ (31 January 2017) TheGuardian, available online <https://www.theguardian.com/technology/2017/jan/30/libratus-poker-artificial-intelligence-professional-human-players-competition> (last accessed 7 February 2017).
[9] Ibid 8.
[10] V C Müller C and Nick Bostrom, ‘Future progress in artificial intelligence: A survey of expert opinion’, in Vincent C. Müller (ed.), Fundamental Issues of Artificial Intelligence (Springer Berlin 2016) 553-571.
[11] Ibid.
[12] M Lemley, P Menell and R P Merges, Intellectual Property in the New Technological Age (fifth ed., Aspen Publishers 2010) 7.
[13] Lemley, Menell and Merges, ibid, 14.
[14] Rebecca Tushnet, ‘Economies of Desire: Fair Use and Marketplace Assumptions’ (2009) 51 Wm. & Mary L R 513, 515.