2. Robotic Stylistics
2.1 Understanding and Concepts of Robotic Stylistics.
2.2 Finding the data
This paper is about a new trend in Stylistics called “Robotic Stylistics” whose subject matter is the linguistic outcome of robots. It investigates the possibility of making the robotic linguistic outcome the subject matter of stylistics, since robots are able to choose what they want to say due to their algorithms that enable them to make a choice.
Keywords: Robotic Stylistics, Artificial Intelligence, Stylistics.
The field of artificial intelligence has developed and prospered in the recent years and entered the linguistic domain introducing a whole new sub-discipline of linguistics knows as Computational Linguistics. It helps linguists increase their knowledge of how the human brain works and how it processes language. Robotic stylistics will be concerned with the analysis and study of the style of robotic linguistic outcome using the same linguistic techniques that are used in the ordinary stylistic investigation of human literary and non-literary texts in order to discover the kind of style a robot has, how recurrent it is, and what it reveals about robots or artificial intelligence in general. However, there are many problems in this field especially in the case of obtaining the adequate data for analysis.
2. Robotic Stylistics
2.1 Understanding and Concepts of Robotic Stylistics.
Stylistics is a branch of macrolinguistics which is about analyzing, and studying the style of a particular person or writer. It is concerned with both literary and non-literary texts and speeches. So it is, by this definition, concerned with the authentic, real utterances only and is not interested in non-authentic, or artificial, unreal utterances like for example, colorless green ideas sleep furiously. The reason is that such utterances are not real, and do not reveal anything about the person's style.
In recent years, the domain of artificial intelligence has developed so fast, and entered the linguistic domain making a new sub-discipline of linguistics known as Computational Linguistics. According to Jean Aitchison (1999: 100), artificial intelligence provided an insight of how people understand each other, and of how computers cope with inexplicit irreverent conversation. It is stated that human knowledge is stored inside the computer memory might be so in the form of stereotypical situation or what Aitchison calls it frames and these frames stored in computer memory may be modified to cope with inexplicit conversations that people may have with computers (ibid: 101). So in this area, AI helps in the domain or natural learning processes. It helps linguists to understand how human brain works and helps them in the field of parsing linguistic and syntactic elements (https://www.tutorialspoint.com/artificial_intelligence/artificial_intelligence_tutorial.pdf) But, the kind of artificial intelligence Aitchison (1999:100) talks about is called Weak Artificial Intelligence which is associated with Massachusetts Institute of Technology (Lucci and Kopec, 2013:44). Weak AI is a label attached to any system displaying intelligent behavior (ibid). So it does not take any account of the similarities between the artifacts and humans (ibid). This type is found everywhere nowadays, in our cellphones, in our smart cars, and in our speech recognition programs in computers…etc.
The concern of Robotic Stylistics is not with that kind of artificial intelligence that can make no choice and is limited in its functions and similarities and which is “powered by giant rulebooks containing massive quantities of data stored on the Internet [that make them] act intelligent but cannot understand the true meaning of what they say or do” (https://www.technologyreview.com/s/609223/robots-arent-as-smart-as-you-think/).
Robotic Stylistics is concerned with the other type of artificial intelligence that is called Strong Artificial Intelligence. It is represented by the scientists of Carnegie-Mellon University . The main concern of this type is with what is called Biological Plausibility, that is, the artifact behavior and performance should be based on the same strategies and methodologies used by humans (ibid). This type of AI has become notorious recently because of what is said about it by some respected scientists and programmers such as Stephen Hawking (http://www.bbc.com/news/technology-30290540) Elon Musk (http://www.independent.co.uk/life-style/gadgets-and-tech/news/elon-musk-artificial-intelligence-openai-neuralink-ai-warning-a8074821.html) and Bill Gates (http://www.bbc.com/news/31047780), and because of what some robots said about the future of humans if they dominated the world, like keeping humans in a “people zoo” (http://www.mirror.co.uk/news/weird-news/super-robot-makes-sinister-promises361329). Sophia the Robot for example said explicitly that she would destroy human beings (http://www.deccanchronicle.com/lifestyle/viral-and-trending/111117/i-will-destroy-humans-humanoid-ai-robot-sophia-gets-saudi-citizenship.html).
However, it has not fully reached what is theoretically known as technical singularity where robots become aware of their existence as very intelligent things and may destroy everything and everyone that might stand against them as some well-known scientists and programmers claim.
Strong AI robots, in this sense, can be an object of study in stylistics. We can analyze their linguistic outcome and figure out the kind of style these robots use since their algorithms allow them to learn from their mistake, from the internet, from watching others do something. The Strong AI robotic algorithms allow robots to say whatever they like as well. So, their behavior is very similar to that of humans that they can learn by, for example, watching videos, and are designed to teach themselves (https://www.umdrightnow.umd.edu/news/robots-learn-watching-videos), or by observing a human being doing something (https://www.technologyreview.com/s/541871/robot-see-robot-do-how-robots-can-learn-new-tasks-by-observing/). This kind of robots are “able to learn how to do a new job by watching others do it first.” (ibid), or by experience, and this kind of robots have a so-called artificial brain that is similar to a human baby’s brain (https://www.express.co.uk/news/science/729751/artificial-intelligence-Robots-brain-learn-like-human-child). They can also learn from what people say, in fact, there are a lot of chat robots or chatbots on the internet that can learn from people and reply to your messages, since their algorithms are designed to make the robots do so. The astonishing thing here is that their creators warn you before you have a conversation with it because they have no control on its choice of language. You can try to have a conversation with one, yourself on http://www.cleverbot.com.
Abbildung in dieser Leseprobe nicht enthalten
*This picture is taken from (www.cleverbot.com)
The strong AI robot once connected to the internet stores a big amount of information about language in its memory, which is the thing I like to call artificial competence in contrast with human competence. Once, this memory is filled with linguistic constructions, words, morphemes and human sounds, the robot will make its own choice to choose the linguistic construction that he/it likes according to the algorithm on which the robot is programmed (https://www.tutorialspoint.com/artificial_intelligence/artificial_intelligence_tutorial.pdf).