Different views on what AI is
According to John McCarthy, the father of Artificial Intelligence, “the science and engineering of making intelligent machines, especially intelligent computer programs” is what AI means in broad sense.
AI being highly interdisciplinary field, the focus of artificial intelligence shifts depending on the entity that provides the definition, while the foundation generally remains the same. AI can be viewed from different perspectives as follows:
-
The definitions based on approaches that measure success in terms of fidelity to human performance,
These definitions aim at imitating the human thought process and behavior. The measure of accuracy of conformance to human performance can be measured with the help of following 2 aspects:
- Thinking humanly: Newell and Simon’s, the ‘‘General Problem Solver’’, was aimed at comparing the trace of its reasoning steps to traces of human subjects solving the same problems.
- Acting humanly: Sophia is social humanoid robot which imitates human gestures and expressions and also converse like human. The above examples of AI illustrate the goal of creating systems that think and act humanly. The pursuit of this goal mostly involves theoretical work such as defining hypothesis and empirical experimentation. This human-centered research is mainly contributed by philosophers, psychologists, neuroscientists and psycholinguists wherein they try to understand how learning, perceiving information and reasoning are done by our thoughts.
-
The definitions based on perspectives that measure against an ideal performance measure, called rationality
These definitions aim at creating systems which are solely based on reasoning, logical thinking and decision making. Again, such systems are measured in terms 2 aspects:
- Thinking rationally: Google’s chess program AlphaZero and AlphaGo
- Acting rationally: Industrial manufacturing robots The contributions made keeping in mind above aim are mainly by mathematicians, computer engineers, and control theory and cybernetics specialists. The main goal of such rational systems is try to perform better than humans and surpass the humans by thinking and acting in the way which is beyond human capability.
-
The definitions focused on defining AI in terms of its usefulness to mankind
This approach uses human reasoning as a model but not necessarily the end goal. The bulk of commercial industry today, focus on the objective of using human thinking pattern as a guide to deliver better services and products, instead of running behind achieving the perfect replica of human mind. For example:
- Google AI defines its research as the “AI provides new ways of looking at existing problems, from rethinking healthcare to advancing scientific discovery.
- Similarly, IBM defines its AI Consulting as “services that help you leverage AI to drive smart reinvention of your workflows and technology”
Each direction has yielded valuable insights such as human-centered goal lead to fascinating discoveries whereas rational and commercial goal led to progress and betterment of mankind.
The ‘hype’ associated with AI 
AI has become a very real obsession in all corners of the tech world. Gartner’s latest Hype cycle for emerging technology is model used to depict new technologies in terms perspective of markets. It suggests that key artificial intelligence (AI) categories such as deep learning, machine learning and cognitive computing are at their “peak of inflated expectations.”
This hype in AI has manifested due to several reasons. Firstly, massive amount of money that is being invested in research of AI by almost every other top tech company in the world. Google has invested the most in artificial intelligence (AI) out of the tech giants, investing $3.9 billion, followed by Amazon and Apple. Also, there is explosion of startups which began in 2012 and continues today. Around 42% of AI startups have been acquired by big tech giants in 2017. Secondly, the AI products which are often over-rated leads to big impact on the perception of how AI is seen. Often, use of AI is heard almost every day and in all kinds of contexts. With self-driving cars, people are getting even more excited about what awaits in the future. Thirdly, it’s the opinions and predictions of people that make it all seem even bigger. For instance, Ray Kurzweil predicted, “artificial intelligence will reach human levels by around 2029”. Also, in 2013 two Oxford professors, Frey and Osborne, estimated that 47% of the job market could become computerized within the next twenty years. Such predictions have sent ripples through many businesses and governing bodies of various nations.
Last but not the least; one cannot ignore the contribution of media in raising the bar of expectations even higher. The media coverage is often being a bit too enthusiastic about any small scale development if the field of AI without validating its veracity.
The question remains, will this hype bubble burst or will scientists and engineers be able to advance the AI technology up to expectations? From previous experience of first hype cycle and current progress made so far, one cannot expect immediate results. This bubble of hype is likely to burst but still there is hope according the hype cycle graph provided by Gartner, which indicates “slope of enlightenment”, only if product are focused on delivering productivity.
My views on positive and negative aspects of AI
Although some approaches to AI would focus on replicating the human mind, the real success of AI lies in creating the system that is able to think beyond human capabilities and is accessible to everyone so as to enhance their lives.
AI has applications in variety of fields and has numerous benefits which can impact people’s lives positively. Artificial Intelligence helps in dealing with the mundane tasks through automation which will give people freedom from dull repetitious tasks, creating more time for creativity. Unlike humans, AI system can work without any break. This mainly improves the productivity and makes everyday lives of the people faster and easier. In critical tasks like performing surgery, AI system can be extremely useful to avoid errors and work with precision and accuracy. Expert systems can be deployed to support decision making which helps to arrive at rational decisions quickly. Also, hundreds of lives can be saved from incorrect diagnosis as AI can enhance accuracy of diagnosis and predict the trends in epidemics. For example: deep learning can diagnose the irregular heartbeats better than a cardiologist.
AI can show significant impact in educational systems today. It can help tailor courses based on child’s abilities and also give insights of reasons behind the poor academic performance. Also AI systems can be deployed to take risks on behalf on human beings in expeditions such as sea exploration or space travel.
While AI has many positive aspects, I think one should consider mitigating the risks and threats that comes along with it in order to utilize this technology to its full potential. Automation of manual systems might lead to and in some case have led to many job losses. If the use of AI becomes uncontrolled, there is no doubt that humans will unnecessarily be highly dependent on the machine losing their creative minds.
The aspect of judgement, a human ability that changes according to circumstances, remains unaddressed in case of AI systems which is found to play a crucial role in some of the critical tasks.
Furthermore, one cannot simply turn a blind eye to pollution and waste disposal problem that is reckoned to increase due to manufacture of digital products and electronics which goes hand-in-hand with development of AI. Also data centers for AI requires massive amount of energy which might put stress on current energy constraints.
A critical appraisal of AI
There is no doubt that Artificial Intelligence can play major role in enhancing everyday life and enabling humans to get better insights before taking decisions. One can think of numerous benefits it can offer to the mankind but one has to be aware and careful in not getting misled by false hopes and promises. As evident from many examples, some the aspects of AI are often exaggerated and hyped by media and other sources, hence one should critically analyze them before starting to expect results. For example: AI driven autonomous vehicles with route optimization and ride sharing techniques are said to reduce number of vehicles on the road. But questions such as what problem it is aiming at, is it necessary to solve it this way or an alternative can be found, what is the status of technology behind it and how far is this feasible etc. can help get better picture of degree of positive impact.
We should not forget that AI is still its initial stages and will require time and money to fulfil the expectations. It can be a boon if properly used and not become entirely dependent on it.
The biggest concern of today with regards to AI is the issues of unemployment due to automation which is true but at the same time chances of creating more skilled jobs are also high as technology augments the human capability. Even though specific positions may disappear, people can continue working alongside machines in new roles. In regard to addressing the fear of losing jobs, the proposal of a basic universal income is also gaining support among businessmen and politicians.
Additionally, the environmental impact of AI is also a hot topic of discussion nowadays. Many companies are coming with the solution to tackle the problem of pollution, e-waste and energy consumption. For example: Google is implementing AI to optimize energy efficiency at its own data centers which has resulted in cutting around 40% of energy used for cooling. So as long as we are successful in keeping technology beneficial and mitigating the risks associated with it, we will be able to help the human civilization. Therefore one should not fear Artificial Intelligence but embrace it.