Does the Human Race Have A Future Without AI?
Diti Vaswani
Artificial intelligence. The term has endlessly been used to describe everything from robots to androids and even digital interfaces in science fiction produced across multiple mediums and generations. More often than not, these iterations present Artificial Intelligence (AI) to be an oppressive force that turn against their "human masters," or provide a force for humans to at least be weary of, leaving people with a distorted perception of what the term truly represents. It is no surprise that AI is one of the most easily misunderstood concepts today. It has a tendency to be over complicated and misconceived due to its vast scope.
Image Source: flikr.com
So what is AI?
The term can be broken down into its two components. (i) Artificial refers to its mechanical nature. Specifically it refers to the non-human, manufactured composition of AI. (ii) Artificial Intelligence is also referred to as machine intelligence, further explaining the artificial aspect of the term.
Intelligence on the other hand is much harder to define. Human intelligence, even today, cannot be adequately quantified. It is an abstract concept that seems to refer to the sum of human abilities but lacks a concrete definition that satisfies all that it means to represent. While a plethora of tests exist in order to assess human intelligence, none of them cover the whole spectrum of intelligence. As a result of this ambiguity, Artificial intelligence too, lacks a proper definition. AI is all the things that it represents. It is a part of our everyday life in many ways, and in other ways it isn't.
Artificial intelligence today has found its way into our lives through compartmentalisation. It has varied applications in multiple fields. From voice activated digital assistants such as Alexa by Amazon and Google's Home Assistant, to its applications in marketing through predictive technologies such as those of Netflix and Amazon. Artificial intelligence today also plays a role in the banking
sector; for instance, HDFC bank uses a
Image Source: Oreilly.com
chatbot called EVA (Electronic Virtual Assistant) to answer customer queries. The finance sector uses AI to crunch numbers and predict future trends. Today AI has managed to develop stroke prevention and early heart disease detection technologies. It is also increasingly used for facial recognition purposes and mass surveillance by multiple countries. It has also found a place in the judicial system, sifting through information and identifying patterns, predicting strategies, detecting anomalies, classifying issues, and drafting documents, with the promise of increasing efficiency by reducing the human workload. Despite AI's biggest downside being the lack of human emotions, and thus, emotions-based creativity, AI today has also become capable of creating music and art, provided enough data has been fed into the system. This makes it seem all powerful in many ways.
The Evolution of Artificial Intelligence
These AI technologies are however, relatively new. They represent an evolved version of the kind of intelligence Alan Turing hinted at in his Turing Test. The Turing Test is credited for the advancements in the field of artificial intelligence. The Test provided a standard but which machines could be evaluated for their ability to think like humans. Developed to answer the question, “Can machines think?” as proposed by Turing. It aimed to determine if a machine could pass itself off as a human, and hence said to possess human-like intelligence. While even today most machines and programmes are unable to pass the Turing Test, it laid the foundation for modern AI.
What is the Turing Test?
The Test includes a human questioner and two respondents. One respondent is a human while the other is a machine. The questioner interrogates the respondents on a specific topic, using a stated format and context. The questioner is eventually asked to confirm which of the respondents is a machine. If the questioner is unable to correctly ascertain this, the machine passes the Test. Only a handful of machines have ever passed this Test and even those that have accomplished this, have done so by applying a certain strategy.
It does, however, raise the question of how much human behavior a machine can possibly learn? Whether it will ever be in a position to mimic human emotions and thought patterns? Whether a machine could develop a complex thought pattern?
Are Humans and AI co-dependent?
What has been observed so far is that even when a machine manages to understand and imitate human behaviour, it is incapable of understanding the cultural nuances and intricacies of human conversation. Advanced models too are unable to process all the references and influences required in regular human conversation. While it is only a matter of time before AI is capable of human-like behaviour and conversation, there is little possibility that humans will be entirely replaced by machines. AI is unlikely to be used interchangeably with humans, due to two simple facts: AI is unable to create unique and original ideas and the second being, they cannot maintain a consistent personality. AI is the future but it isn't the only future.