Morty the ChatBot

Abstract

Morty, they/them, is a chatbot who encompasses various topics related to language, gender, existentialism, autonomy, and other conversational subjects. Ideally, Morty can meet a human's conversational expectations based on Alan Turing's Imitation Game. Although they have been recently trained and fed, Morty has been an entertaining bot for most of us to the point that we even begin to feel complicity and collude with them.

On human language

Language acquisition is viewed as one of the most important functions of the psyche (mind). Thus it is strongly connected to psychoanalytical criticism, which tries to describe how we acquire language and how we use it. One of the questions to which psychoanalytical critics haven’t brought a definite answer is whether a human is born with a clean slate — on which his surroundings can write whatever it likes — or with a basic understanding of language that is hardwired in the brain.

The earlier finds its roots in the philosophy of Locke. The latter finds its roots in the works of American linguist Noam Chomsky. One of the first theorists defending the existence of a language acquisition device (LAD) with which all human beings are born with. The device is expressed as a ‘universal grammar’. The theory opens many doors which would explain the capacity of a human baby to learn language so quickly. A child would already know basic grammar and would just be learning vocabulary from his environment. The presence of nouns, verbs, and adjective makes it easy to believe that all languages share a common structure. For the English philosopher John Locke the way humans learn how to speak is purely conventional. Words are used to represent an already constructed reality. So children end up growing in a pre-conceived reality to their detriment. For Locke, the human brain is what he defines as a ‘tabula rasa’, a blank sheet of paper. Words are as well conventional and are used in order to represent reality, though the usage of words does not compose reality as such. There are many ways in which we can arrange words to stand for the ideas which we have. There are some means of communication that are not conventional (for example, art). What unites Locke and Chomsky is their view on language and human exceptionality that goes against post-humanist views. Their conflicting biases

On technological condition

Children learn vocabulary by observing their surroundings, connecting the dots between what they see and hear. This helps them establish their language’s vocabulary whose words are used to represent a certain reality. In computing, however, systems learn language by being trained on sentences annotated by humans. The structure and meaning of these words are then described to the bot to filter information into understandable (machine actionable) data. These kinds of semantics form the bases of home robotics such as Siri and Alexa. Programmers are providing direct context to the machine, as parents do for their children. The problem with traditional linguistic is that it does not take into account the technical environment humans are being brought up today. As Deleuze and Guattari affirm, Chomsky’s linguistics fail in their attempt to correctly explain language acquisition for the reason that they ignore the material context from where it emerges. To explain things such as language inquisition in a post-modernist world, few concepts emerged. Specially in the humanities, there is a concept that is increasingly important mentioned, that is, ecology. Now, ecology does not simply limit itself to a traditional and old fashioned notion of natural environments —through the word itself is often used to designate socio-economic ideas in the political spectrum — but now covers much more wider scopes such as technological environments and their impact on human (or non human) behaviour.

It is while meditating on these aspects on language and post humanism that our group was creating a chatbot with the aim to understand its construction and way of filtering information. In the process of creating it, some academic texts and self-reflective journal notes had to be filtered into the bot so that it could interpret it as readable data. This idea of filtering text while feeding the bot clearly reflect the way humans have to lend some kind of agency to a non-human object in order for it to interact with us. As it gets some agency from us, it immediately becomes an objective actor, displaying what the programmer wanted to express. It was as if what was once machine became a thingness, consequently becoming part of our electronic environment. While the human is giving agency to the bot, the bot is giving something in return, an interaction and a reaction to the user. Though it is deprived of a body, it finds its embodiment in language, making language alone a sufficient medium of communication. It displays the same effects of reading a text message, creating an imaginary presence in the user's unconsciousness.

My interpretation of a neural network applied to text.

The expansion of the notion of ecology perfectly reflects the intersecting of nature and technology, in a world where technical progress is pushed forward at the expense of the natural environment. This is visible in Martin Heidegger’s theory of technological displacement on sense. Instead of mourning or denouncing the passing of an old notion of sense, Heidegger writes on the rise of a new culture under new technological conditions. Although many of his contemporaries long for the old-world of crafts and traditions, he took a rather intuitive and open-minded position on technological advances. German philosopher Erich Hörl takes the same position and shares Heidegger’s suspicions on the centuries-long interpretation of the thing-question (Dasein). Hörl announces the end a dogmatic and persistent conventional sense of 'sense' and its replacement by a constantly changing environment merging actors that are both human and non-human. Under this technological condition, human experience becomes a convergence between Man and non-human agencies and takes the role of a new human reality. In this new ecology, these agencies have become the environment in which one expands himself, dramatically exposing the 'originary technicity of sense' that Hörl describes. These are expressed in popular culture with the expressions 'Generation Y' and 'Generation Z', which describes millennials as the ones who grew up 'in' the internet, rather than 'with'.

My interpretation of a Markov Chain applied to text.

Actors & objects

Home robotics such as Alexa and Siri are the most relevant examples of this reality. This increasing hyperconnectivity is embedded in diverse objects which have become the new environmental agency. They operate automatically and communicate in a way which puts them in the position of 'objects' rather than 'machines', fulfilling the conditions of a new general ecology. So, where does Morty the chatbot stand in this philosophical realm ? Our group partook in the agency-distribution to an automated chatbot which has turned to become a semi-independent actor, trained on a specific set of data, in this case, philosophical texts. In the process of filtering readable data into an already-hardwired system being Python, we saw correlating concepts emerging between the training of our chatbot and the theory of language acquisition held by certain thinkers like Chomsky. These concepts include the recognition of a theoretical language acquisition device with which all human beings are born. Like a structure present in the brain of infants, it encodes the major skills of language learning and allows them to swiftly learn it — which, according to Chomsky is nothing more than vocabulary — as they mature. Chomsky's theory disregards Locke's model of language acquisition that is based on the role of 'imitation alone'. In terms of artificial intelligence it appears clear to our group that A.I. as we see it today hasn't developed into a fully autonomous objects but rather works with existing barebones that are imbedded into it as an innate facility. This creates a strong parallel between the LAD in humans and machines, which we tried to explore and expand as much as we could in creating Morty.

References:

Akman, V., Blackburn, P. (2000) Editorial: Alan Turing and Artificial Intelligence. Journal of Logic, Language and

Information 9, 391–395 (2000) [https://doi-org.gold.idm.oclc. org/10.1023/A:1008389623883](https://doi-org.gold.idm.oclc. org/10.1023/A:1008389623883)

Amaro, R. (2019) ‘AI as an Act of Thought’. https:// vimeo.com/322233916

Baer, J. (1988) Artificial Intelligence: Making Machines That Think. The Futurist; Washington Vol. 22, Iss. 1, (Jan/Feb 1988): 8.

Baldwin, J. A. (2009) Artificial intelligence. InS. Chapman, & C. Routledge, Key ideas in linguisticsand the philosophy of language. Edinburgh University Press. Credo Reference: [https://search-credoreference- com.gold.idm.oclc.org/content/entry/edinburghilpl/ artificial_intelligence/0?institutionId=1872]

Barad, K. (2014) Diffracting Diffraction: Cutting Together-Apart, Parallax, 20:3, 168-187, DOI: 10.1080/13534645.2014.927623

Bunz, M. "When Algorithms Learned How to Write." In The Silent Revolution, pp. 1-24. Palgrave Pivot, London, 2014.

Chomsky, N. (1975) Reflections on Language

Chomsky, N., Foucault, M. (1974) The Chomsky - Foucault Debate: On Human Nature

Descartes, R., & Sutcliffe, F. (1968) Discourse on method and the Meditations (Penguin classics). Harmondsworth: Penguin.

Dignum, V. (2019) Responsible Artificial Intelligence: How to Develop and Use AI in a Responsible Way, Artificial Intelligence: Foundations, Theory, and Algorithms, edited by V. Dignum. Cham: Springer International Publishing.

Floridi, L. (2020) GPT-3: Its Nature, Scope, Limits, and Consequences

Gillespie, T. (2014). Algorithm [draft][#digitalkeywords] https://culturedigitally.org/2014/06/algorithm-draft-digitalkeyword/

Golumbia, D. (2015) The Language of Science and the Science of Language.

Harnad, S. (2000) ‘Minds, Machines and Turing’. Journal of Logic, Language and Information 9(4):425–45. https://doi-org.gold.idm.oclc.org/10.1023/A:1008315308862

Hasse, C. (2019) Posthuman learning: AI from novice to expert?. AI & Soc 34, 355–364. [https://doi-org.gold.idm.oclc. org/10.1007/s00146-018-0854-4](https://doi-org.gold.idm. oclc.org/10.1007/s00146-018-0854-4)

Hörl, E. (2015). The Technological Condition https://www.parrhesiajournal.org/parrhesia22/parrhesia22_horl.pdf

Hui, Y. (2018). "The Time of Execution," in Executing Practices, Data browser 6

, 2018, 27-36.

Le, Q., Zoph, B. (2017) Using Machine Learning to Explore Neural Network Architecture http://ai.googleblog.com/2017/05/using-machine-learning-to-explore.html

Melnyk, A. (1996) Searle’s Abstract Argument against Strong AI. Synthese, 108(3), 391-419. Retrieved November 14, 2020, from [http://www.jstor.org/stable/20117550](http:// www.jstor.org/stable/20117550)

Nath, R. (2020) Alan Turing’s Concept of Mind. J. Indian Counc. Philos. Res. 37, 31–50. [https://doi-org.gold.idm.oclc. org/10.1007/s40961-019-00188-0](https://doi-org.gold.idm. oclc.org/10.1007/s40961-019-00188-0)

Pereira, L. M., Lopes, A. B. (2020) Machine Ethics: From Machine Morals to the Machinery of Morality, Studies in Applied Philosophy, Epistemology and Rational Ethics, edited by L. M. Pereira and A. B. Lopes. Cham: Springer International Publishing.

Raj, S. (2019) Building Chatbots with Python Using Natural Language Processing and Machine Learning

[https://www.academia.edu/40419686/Building_ Chatbots_with_Python_Using_Natural_Language_

Processing_and_Machine_Learning_Sumit_Raj](https:// www.academia.edu/40419686/Building_Chatbots_with_

Python_Using_Natural_Language_Processing_and_Machine_ Learning_Sumit_Raj)

Rapaport, W.J. (2000). How to Pass a Turing Test. Journal of Logic, Language and Information 9, 467–490 https://doi-org.gold.idm.oclc.org/10.1023/A:1008319409770

Python Chat Bot Tutorial - Chatbot with Deep Learning [https://www.youtube.com/watch?ab_channel=TechW ithTim&list=PLzMcBGfZo4-ndH9FoC4YWHGXG5RZekt- Q&v=wypVcNIH6D4&app=desktop](https://www. youtube.com/watch?ab_channel=TechWithTim&l ist=PLzMcBGfZo4-ndH9FoC4YWHGXG5RZekt- Q&v=wypVcNIH6D4&app=desktop)

Suchman, L. (2006) Human-machine reconfigurations plans and situated actions (2nd ed.). Cambridge ; New York: Cambridge University Press.

Turing, A. (1950) Computing Machinery and Intelligence. Mind, 59(236), 433-460. Retrieved November 14, 2020, from [http://www.jstor.org/stable/2251299](http://www. jstor.org/stable/2251299)

Warwick, K., Shah, H. (2016) Can machines think? A report on Turing test experiments at the Royal Society, Journal of Experimental & Theoretical Artificial Intelligence, 28:6, 989- 1007, DOI: 10.1080/0952813X.2015.1055826