The ‘infinite monkey theorem’ suggests that a monkey, hitting the keys at random on a typewriter keyboard for an infinite amount of time will, at some point, type out William Shakespeare’s Hamlet.
A website called The Mary Sue recently calculated the actual odds of 100 monkeys typing Hamlet – taking into consideration the huge number of character sequences needed - are one chance in 36……36 to the power of 166,541 that is. Or as Google Calculator puts it, “one in infinity” - in other words not entirely impossible, but incredibly unlikely.
Enter artificial intelligence (AI). AI software can process extremely large data sets of information so quickly and efficiently that the time taken until one in infinity happens does not seem quite so long. By using large or deep neural networks, AI software can learn faster than any other technology - growing more knowledgeable with exposure to more data. AI is being applied to exceptionally complex issues. For example, AI can find the most optimal solution to issues that have more potential solutions than the number of grains of sand on earth – in a matter of seconds.
AI systems are now capable of producing music, art and software. The idea of something being created by a machine that would previously have been created by a human raises a myriad of complex issues, not least of which is intellectual property rights.
Man vs Machine
US patent law tells us that an inventor is “the individual or, if a joint invention, the collective individuals who invented or discovered the subject matter of the invention.” The US Copyright Office also employs a 'human authorship requirement'. Machines do not have a stake over IP rights and it is unlikely they ever will. Employers own IP created by its employees during their employment – if the duty to create IP was part of their employment duties. This could easily apply to the deployment of machines as well.
Jukedeck, for example, is a software tool that combines deep learning, music composition and audio production to create music for royalty free commercial use. It has produced more than 500,000 compositions for companies to use for their own purposes and the copyright for the music created can be purchased by the company that needs it.
The Future of IP
AI remains an area where there is frenetic IP activity. To date this has largely been based around protecting AI systems or applications. But, in a future where machines are deployed to create complex products, the question of IP ownership could become clouded. AI helps solve complex issues simply and elegantly; legal ownership on the other hand may not be quite so simple.
Interested in reading more about the future of IP? Read about the Fourth Industrial Revolution in a new blog by our Chief Data Officer, Tyron Stading.