Artificial Intelligence (A) and Patentable -Creativity
Artificial intelligence (AI) is rapidly evolving. It is capable of reading, talking, learning and generating its own work. [1] Some scholars and computer scientists are of the opinion that AI is some sort of ‘humanoid’ [2] with a higher potential than the human brain with its ‘slow-acting and wasteful’ energy. [3] Further they argue that there is no such thing as ‘true human creativity’ since neither AI nor human beings are capable of generating information. [4] In line with the above, there are no obstacles to define the work generated by AI as creative. One may argue further that both humans and AIs have been programmed, one by genes and the other by software programs, to accumulate experiences. [5] Arguably, AI can exhibit the qualities of a human mind. But is it possible for an AI to generate a work that is creative? To find the answer to this question, one need to first consider the meaning behind the word ‘creativity.’
Creativity
Creativity is a feature of intelligence related to the association of ideas, analogical thinking, or searching a structured problem-space. As such, it makes more sense two define creativity in terms of processes. [6] A creative idea, on the other hand, might be defined as novel, non-obvious/inventive, and industrially applicable. [7] Within this framework ‘inventive’ has to different meanings. The work generated might be ‘non-obvious/inventive’ with respect to the entity that created it (e.g.: AI) or, so far as one knows, to the whole of previous creations. The ability to create the second type of work for the reasons of this note will be called P-creativity (P for patentable).
P - Creativity
There are many kinds of creativities (political, artistic, advertising and so on) but this paper focuses only on the unique brand of creativity that produces patents (P-creativity). Patent is a ‘creature of law that allows inventors to apply for and obtain a right to exclude others from making, using, or selling invention if the application and the invention meet certain standards’. [8] This note suggests the term P-creativity to describe the kind of technological breakthrough creativity that leads to patents. Kennedy and Vatkins proposing following definition: ‘patentable creativity is the capacity of the prepared mind to recognize and satisfy a need in a new, useful, and unobvious manner for society’s benefit.’ [9]
The first part of this definition covers the intellectual input of the inventor, whereas the second half connects it to the requirements of patent law. Section 101 of the US Patent Act sets forth the general requirements for patent protection in a single sentence:
‘[w]hoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvements thereof, may obtain a patent, subject to the conditions and requirements of this title.’ [10]
Additionally, under US patent law, an invention is patentable only if it meets the following four requirements: (1) the invention must be statutory (2) novel (3) useful and (4) non-obvious. [11] Having a satisfactory expression and a working definition of P-creativity, one can now proceed to ask: whether it is possible for the AI to be P-creative. One may argue that if only an AI is capable of generating a patentable invention the AI is P-creative.
AI as Creator of a Patentable Work
System such as genetic programming (GP), which is an evolutionary algorithm-based methodology inspired by biological evolution to find computer programs that perform a user-defined task, [12] has already automatically created a computer program that is competitive with a human, produced results that are patentable.’ [13] This is to indicate that AIs known as ‘genetic programming’ or ‘evolutionary engineering’ have arguably proven to be the most effective at replicating human P-creativity. [14] Proving the opinion that an AI does not depend only on a linear set of programming instructions—rather, it has some capacity to reason for itself. This is also evidence that AIs are capable of generating something unpredictable, non-obvious and therefore P-creative.
In conclusion, AI can be creative.
References:
(OSCOLA style of referencing)
[1] See my posts "Thinking Machines and Patentable Inventions" & "Fully Autonomous Artificial Intelligence - An Agent of No One"
[2] S N Lehman-Wilzing, ‘Frankenstein Unbound, Towards a Legal Definition of Artificial Intelligence’ (1981) 30 Futures—IPC Business Press 442, 444.
[3] AC Clarke, Profile of the Future (2nd edn, Hachette UK 2013) 28.
[4] J Rose, Progress of Cybernetics (Gordon and Breach 1969) 28.
[5] HA Simon, ‘The corporation: Will it be managed by machines?’ in M Anshen and GL Bach (eds) Management and corporations (McGraw-Hill 1985) 17-55.
[6] Margaret A Boden, ‘Creativity and Artificial Intelligence’ (1998) 103 Artificial Intelligence 347, 347; Margaret A Boden (ed), Dimensions of Creativity (MIT Press) 1994.
[7] Patents, Trademarks, and Related Rights: National and International Protection, Volume 1 (Harvard University Press 1975) 1988.
[8] Alan R Thiele and Judith R Blakeway and Charles M Hosch, The Patent Infringement Litigation Handbook: Avoidance and Management (American Bar Association 2010) 9; Rachel Gader-Shafran, Intellectual Property Law Dictionary (Law Journal Press 2004) xxviii.
[9] Joseph P Kennedy and Wayne H Watkins, How to Invent and Protect Your Invention: A Guide to Patents for Scientists and Engineers (John Wiley & Sons 2012) 172.
[10] The United State Patent Act 35 Section 101.
[11] Ibid.
[12] Pandian Vasant, Handbook of Research on Novel Soft Computing Intelligent Algorithms: Theory and Practical Applications (IGI Global 2013) 326.
[13] John R Koza, Genetic Programing III: Darwinian Invention and Problem Solving 5 (1999); Jonathon Keats, John Koza Has Built an Invention Machine, POPULAR SCI., May 2006, at 66, 72 (describing how a genetic algorithm independently and autonomously generated several different circuit designs that had been previously patented by human designers).
[14] Ibid.
(Image: Getty Images)