The human curiosity and desire for communication often inspire them to transfer their knowledge and speech patterns to other organisms and artificial creations. The rise of artificial intelligence (AI) and robots was not an exclusion – to talk to robots, scientists began designing the Robot Interaction Language (ROILA). It is a spoken language that was developed to communicate with robots. The difference of ROILA from other languages for computing is that it should be simple for both machines and humans to understand. For people, natural speech is the easiest way of sharing information, in contrast to programming languages and algorithms.
For robots, however, the fluidity of human communication is challenging as it often does not follow the rules that can be analyzed or deciphered. For this purpose, ROILA was created – it is a language that combines the learnability of natural languages and the optimization for machine recognition patterns.
The rules of ROILA were designed to mimic the standards of existing spoken languages. Thus, its alphabet is based on the main letters from the English language, but the pronunciation excludes all diphthongs to remove inconsistencies that occur due to vowels’ position in a word. All words are taken from Basic English, and the vocabulary is separated into three types. ROILA words have to follow a structure, including CVCV, CVCVC, and CVCVCV, where C is a consonant, and V is a vowel.
Finally, the language’s grammar is isolating; affixes and inflections are replaced by word markers or additional words to specify the meaning. For example, a verb in the past tense such as “walked” turns into “walk jifi,” where “jifi” implies the past tense. While the rules of ROILA may be new to English speakers, its intuitive use of letters and words should simplify the learning and speaking experiences.
Robots, however, need to use human languages for several purposes. In business, research, and other spheres, machines have learned to recognize words, including their meanings and connection to each other. Nonetheless, the creative field is still in a major debate about robots possessing an ability to make art of any form. A question of whether a robot can simulate poetry or another creative text persists. Historically, it was believed that robots would be able to replace people only in spheres that were connected to monotonous work, construction, physical labor, or areas related to analysis and calculations. Creativity was a field completely separate from machines, being driven by humans’ innate originative skills.
Recent developments put this belief into question, with advanced AI technologies mimicking human behavior and thinking patterns with precision. Machines can learn using images, databases, social interactions, and algorithms. They can create pictures and images that are challenging for a person to differentiate from a human-made object. Several attempts were made in the field of poetry, in which a creative AI wrote a poem based on the information it examined.
Moreover, the flexibility of the term “poetry” blurs the standards for what can be perceived as creative. Such projects as botpoet.com provide examples of poetry written by both people and robots that, sometimes, makes the task of guessing the authorship difficult. All in all, the question of robot creativity remains open as the research in this field is scarce. Nevertheless, the continuous stream of innovation and the quickly improving machine learning make the idea possible.
Reference List
‘About’. 2015. Web.
‘Language Guide’. 2015. Web.
Monet, M., ‘Can a Robot Write Poetry?’, The Startup, 2019. Web.
Pease, A., and A. Jordanous. ‘Report on the Eighth International Conference on Computational Creativity’, AI Magazine, vol. 39, no. 1, 2018, pp. 62-65.