Website design
HIT164 COMPUTING FUNDAMENTALS RESEARCH ESSAY ASSIGNMENT OPTION 6: Artificial Intelligence
Kazi Intisar Adeeb s284814
Table of Contents Introduction 1 Learning 1 Speech Recognition 1 Prediction 2 Natural Language Processing 2 Reasoning 3 Dialogue Technology 3 Conclusion 4 References 5
iii
Introduction
The relationship between humans and technology has really evolved. This has been due to the process of purchasing devices and machines that are supposed to fulfil certain needs to humans (d’Arnault, 2015). Artificial Intelligence has since seen its way in today's world, especially with the increased technological competition. Samantha is an example of such technology in "Her" the movie. Samantha, an operating system develops an emotional relationship with Theodore, her user (Her - Official Trailer (HD) Joaquin Phoenix, Amy Adams, 2013). Samantha shows competency as she transforms from an assistant, to like an actual agent who is seen arranging publications of letters belonging to Theodore, and to being a girlfriend. She is characterized by being a planner, she reasons, can have a dialog, speech generation, natural understanding of language, and speech recognition. The following paper discusses some benefits of computer intelligence evident in Samantha and comparison to possible disadvantages.
Learning
Machine learning is one of the major elements used in the movie "Her". Machine Learning offers machines the knowledge or ability for them to learn without the need for an all programmed program. It involves computer algorithm that automatically improves via experience (Issa, 2015). One of the learning techniques evident in the movie is Deep Neural Networks (DNN). This technique utilizes matching of patterns that contain a lot of interconnected layers. This technology is inspired by the neural networks of the brain. The technique relies on examples and that explains how Samantha is almost knowing how she should respond, do or act. This has a benefit of the fact that she will never forget what she has been trained about but a disadvantage on the end. She will only adhere to what she is trained on or what is in the training exemplars, unlike the human who would be able to handle a completely strange issue.
Speech Recognition
Speech Recognition is also another attribute that the movie uses to bring out the role of Samantha. Speech recognition develops technologies and methodologies that allows spoken language to be recognized and translated into text by the use of a machine or computer (Cheng & Day, 2014). The resulting system is able to analyse the voice of an individual and with the fine tuning of the voice, it's able to recognize the particular speech. This results in better accuracy. There is no one instance that Samantha is unable to recognize what Theodore intends to say. She is always having a response and even seems to understand the user. This is an advantage to Samantha over humans. Humans are limited to certain knowledge and to only the knowledge they acquire through schools and experience. Samantha, on the other hand, uses a large pool of data that will increase as vocabularies increase. This data is the one that statistical models use during training through the use of machine learning techniques that are powerful. With the evolution of deep neural networks, the performance of both the language and acoustic models even boosts speech recognition. Further, there is also the application of DNNs to the issue of matching meanings the words already recognized (Becerra-Fernandez, 2000). This is done through the transformation of sequences of a word into vector representations. Compared to the recognition of patterns, these are more amenable.
Prediction
In the movie, one can tend to think that Samantha has learned her user in and out. She seems to understand what Theodore wants. She is able to plan meetings for him, delete inboxes and don't just delete but delete with an understanding of the impact or importance. This is majorly through prediction another element of machine learning. Compared to simple dialogs that virtual assistants engage in, they are only able to produce scripted conversations (Kowalski, 2011). Prediction in this is limited. Samantha is hyper-efficient and shows a lot of eloquence in whatever she does, making Theodore's life pretty enjoyable. In comparison with humans, they can hardly predict accurately and if the prediction is successful a lot of analysis is involved. The machine would offer a better prediction. This is a benefit compared to possible disadvantages.
Natural Language Processing
Natural language processing is the technology that enables the successful processing of natural/human language by computers. Its aim is to ensure that computers understand human speech, by understanding the meanings of the words uttered by humans and recognizing the habits of human speech (Fankhauser, 2015).
In the construction of "Samantha", natural language processing would be a very important technology. This technology involves understanding human language, translating it to machine-readable format, a format that the machine can understand, and generate the natural language.
“Samantha” would first have to understand human language. This would involve training her through machine learning. Machine learning would ensure that Samantha would learn without explicit programming. This continuous learning means that if exposed to new data, Samantha would respond appropriately. Learning can be achieved by passing both written and spoken word to Samantha, and then train her on how to separate the words maybe into individual morphemes, and the identification of the morpheme classes. Semantic Indexing would be another natural language processing method to use for these intelligent machines to extract meaning from the human language.
While using natural language might have limitations in areas of machine translation, reliability issues especially where the interface is not programmed and the huge training requirements; using this technology would ensure increased processing speeds, improved consistency, efficiency and accuracy of the documentation process and hence the results, and would lead to the identification of the most pertinent information in the databases, as opposed to using humans, whose answers or results to the same question might vary every other time. Also, more human resources would be required to perform the activities that would be easily by an AI such as Samantha.
Reasoning
Reasoning refers to the technology and algorithms used to enable machines to imitate the step-by-step reasoning used by humans to solve problems or to make logical deductions and conclusions. With this technology, it is possible to understand high-level goals, fill in any blanks and missing information implicitly without receiving the explicitly stated step by step instruction (Brodie & Mylopoulos, 2012).
In Her, Samantha is able to perform tasks such as those of a personal assistant, arranging of the publication of her boss’s letters and acting as a girlfriend among others. With reasoning, machines are able to perform complex tasks and reason flexibly without any pre-determined responses or complete information by manipulating mental concepts, logically deducting information and matching patterns. That is why Samantha is able to be a really good girlfriend and personal assistant, and in the end, is able to realize that humans are unsatisfying companions and thus loses interests in them.
Reasoning's reliance on observation to collect information and assumption of uniformity of nature are its main drawbacks. However, it works to quickly make conclusions with incomplete information and has many application areas such as predictive analytics, robotics, scheduling among other areas.
Dialogue Technology
This technology enables a computer system to coherently converse with a human. Technologies such as dialogue modelling, speech technology, language processing and the availability of powerful computing systems have contributed to the successful implementation of the dialogue technology.
In artificial intelligence such as Samantha, this technology allows for dialogue between the human and the computer system (Guardian, 2014). It allows for the human to speak coherently to the system, the processing of the input, management of the conversation flow and keeping of the dialog history and state by the dialog system, and the generation of the output by the intelligent system and rendering of the output by an output renderer.
A dialogue system allows for intelligent systems to behave like humans, imitating the human behaviour and emotions and basically behaving like humans (McDermott, 2007). This dialogue technology would be required to successfully construct an intelligent system like Samantha, who is able to have a coherent conversation with her boss in a very graceful manner, exactly the way a human would do. These systems are able to also adapt to situations that they had not earlier anticipated and recover from any incorrect inferences, use the right tone and speak coherently. With machine learning, intelligence systems are able to adjust their dialog strategies on the basis of the users' feedback, discover new concepts and continuously learn and adapt to the new spoken form of the existing concepts.
While it is a bit expensive and complex to develop a fully functioning dialog system, dialog technology has many applications in today’s world, including usage in customer service, technical support, help desk services, call centres, website navigation among other areas. They bring about efficiency in service and reduces the cost of labor in case humans were to do the same work.
Conclusion
As demonstrated above, artificial intelligence possesses both benefits and disadvantages but considering the above, benefits outweigh the shortcomings. Samantha is an indication that application of technology is yet to come to an end.
References
Brodie, M. & Mylopoulos, J., 2012. On knowledge base management systems: integrating artificial intelligence and database technologies. s.l.:Springer Science & Business Media.
Cheng, S.-M. & Day, M.-Y., 2014. Technologies and Applications of Artificial Intelligence: 19th International Conference. s.l.:Springer.
d’Arnault, C., 2015. Medium.com: On The Relationship Between Humans and Technology. [Online]
Available at: https://medium.com/adventures-in-consumer-technology/on-the-relationship-between-humans-and-technology-ebad02e0c2c0
[Accessed 12 April 2017].
Fankhauser, W., 2015. Artificial Intelligence Applications: Natural Language Processing. s.l.:CreateSpace Independent Publishing Platform.
Guardian, T., 2014. Her- Review. [Online]
Available at: https://www.theguardian.com/film/2014/feb/13/her-review
[Accessed 12 April 2017].
Her - Official Trailer (HD) Joaquin Phoenix, Amy Adams. 2013. [Film] Directed by Spike Jonze. s.l.: joblo.
Issa, T., 2015. Artificial Intelligence Technologies and the Evolution of Web 3.0. s.l.:IGI Global..
Kowalski, R., 2011. Artificial Intelligence and Human Thinking - IJCAI. Journal of Information Systems, pp. 203-500.
McDermott, D., 2007. Artificial Intelligence and Consciousness. s.l.:Yale University.
Becerra-Fernandez, I., 2000. The role of artificial intelligence technologies in the implementation of people-finder knowledge management systems. Knowledge-Based Systems, 13(5), pp.315-320.
5