We use Siri every day to carry out mindless tasks such as responding to a text, calling a friend, and searching for a random fact. Though Siri doesn’t provide much new utility, it makes it easier for the user to carry out daily tasks with their voice. First of all, this proves the laziness of mankind which these massive technology companies, such as Apple, respectably take advantage of. However, we neglect the amount of effort that goes into training algorithms responsible for programs such as Siri, Alexa, and even Google translate. All these algorithms utilize a branch of computer science known as natural language processing (NLP) which basically concerns the machine interpretation of written or spoken language.
The future demands for computer interactions considering the growth and development of artificial intelligence. NLP makes the transition from simple AI to complex machine learning possible. The greatest challenge in NLP isn’t actually the machine itself but instead the inconsistencies in the human language. It is known that we make simple errors such as spelling, grammar, punctuation, and syntactical errors. There are some known algorithms, such as autocorrect and Grammarly, that attempt to resolve these errors. Despite these various attempts, the inconsistency in these errors makes it difficult to train these machines with data consisting of the same random errors. With the lack of proper training, current algorithms will continue to fall for these inconsistencies resulting in undesired output. Another example of faulty training leading to biases is Siri not recognizing the vocal input due to a strong accent. It is very likely that Siri was not thoroughly trained on a dataset that included that accent leading to a noticeable bias that proves inconvenient to the user.
As society progresses into the future, interactions between humans and computers are inevitable. Therefore, this branch of computer science is an increasingly essential field as a foundation for the growth of humanity. However, it is essential to take into consideration the inconsistencies and unpredictability in human nature when approaching the future of computer science.