Artificial intelligence has been an attraction between researchers for exceptionally long and is currently a field that is being worked on for its probable benefits. AI like everything else, has its advantages and disadvantages. This article describes Artificial Intelligence, its types, its pros, and cons, the risks of forming a bias due to limited datasets and why it should be regulated, how it can lead to the breach of privacy on a large scale due to its need for collecting enormous amounts of data and further fall into the wrong hands risking the safety of millions.
Artificial Intelligence (AI) Is a branch of computer science that aims to build smart machines capable of doing not only physical human labour but also have the capacity to substitute work that requires human intelligence. The earliest development towards AI began with Newell and Simon. They designed the “logical theorist” in 1955, the first AI program. The person who coined the term “Artificial Intelligence” however, was John McCarthy, also known as the “Father of Artificial Intelligence”. He created the LISP Language[List Processing] in 1958 which was adopted by many researchers and is still in use today.
Types of AI
Narrow AI/Weak AI
Narrow or weak AI is the most popular AI at the moment. It focuses on one task primarily and operates within a limited sphere. It relies on machine learning and deep learning. Siri, Alexa, Google Search are a few common examples of Narrow AI.
Artificial General Intelligence
Artificial General Intelligence refers to creating a machine with human-level intelligence that can be applied to any task. It has infatuated researchers for ages and continues to do so. It is however still a goal far from being achieved.
Emergence of AI in India
On 12 October 2018, NITI Aayog made an agreement with Microsoft India that it will establish a national program to direct efforts towards the area of Artificial Intelligence. The strategy has been termed as #AIForAll. According to the National Strategy for AI discussion paper the main sectors of focus are:
- Health care:
AI can help remote areas and rural areas to access health care better since they currently are located far from it and lack adequate healthcare professionals. AI driven diagnostics, personalised treatment, etc can help counter this problem.
AI can help meet the increased demand for food. It can address issues such as inadequate demand prediction, lack of assured irrigation, and overuse or misuse of pesticides and fertilisers. Some use cases include improvement in crop yield through real-time advisory, advanced detection of pest attacks, and prediction of crop prices to inform sowing practices.
- Smart Mobility, including Transports and Logistics:
Potential use cases in this domain include autonomous fleets for ride-sharing, semi-autonomous features such as driver assist, and predictive engine monitoring and maintenance. Other areas that AI can impact include autonomous trucking and delivery, and improved traffic management.
The retail sector has been one of the early adopters of AI solutions, with applications such as improving user experience by providing personalised suggestions, preference-based browsing, and image-based product search. Other use cases include customer demand anticipation, improved inventory management, and efficient delivery management.
AI can help in checking manufacturing quality, through flexible and adaptable technical systems to automate processes and machinery to respond to unfamiliar or unexpected situations by making smart decisions. Impacted areas include engineering, supply chain management, production, maintenance, quality assurance, and in-plant logistics and warehousing.
AI can help in energy system modelling and forecasting to decrease unpredictability and increase efficiency in power balancing and usage. In renewable energy systems, AI can enable storage of energy through intelligent grids enabled by smart meters, and also improve the reliability and affordability of photovoltaic energy. Similar to the manufacturing sector, AI may also be deployed for predictive maintenance of grid infrastructure.
- Smart Cities:
Implementation of AI in newly developed smart cities and infrastructure can help meet the demands of a rapidly urbanising population. It can help in traffic control to reduce congestion and enhance security through improved crowd management.
- Education and Skilling:
AI can help improve quality and access of education through personalised learning, automating, and speeding up administrative tasks thus reducing the load on teachers and predicting the need for student intervention to reduce dropouts or recommend vocational training
Advantages Of AI
- Digital assistants: Digital assistants like Siri, Cortana, Google assistant help reduce the time and energy taken to type words and directly convert spoken words to text.
- Translation: technological developments like Google translate have reduced the handwork required in translating sentences from language to another. Although not yet perfected it is still handy and impressive.
- Facial recognition: This is an especially important feature that enhances security to a whole other level compared to pins and passwords.
- Medicine: AI can help in diagnosis remotely and suggest treatments.
- Agriculture: AI can help farmers in analysing the weather better and accordingly improve their harvest. It can also help to detect diseases , pests, and poor nutrition in plants helping them to measure the amount of pesticide and fertilizer they have to use.
- Business sector: AI can help in doing tasks that are repetitive and with fewer errors as compared to manual labour. Chatbots for customer service also helps to provide instant service to customers.
- Education: AI can reduce the load on teachers that takes place in grading, rewarding marks, etc. It could also establish a way for personalised curriculums based on differing capabilities and understanding levels of students.
- Financial sector: AI can help in compiling personal data and detect fraud.
- Legal sector: AI can help in speedier disposal of cases as it will take less time while analysing cases.
- Cyber security: AI can analyse threats to security by accessing its huge database.
- Mining: AI can reduce the threats mining poses to humans and also help reach depths that human limitations prevent.
- Defence sector: AI can provide logistics support, help in situational awareness, location of goals, conducting surveillance missions in unsafe areas.
Disadvantages of AI
- Job security:
While AI is attractive and looks promising for substituting manual labour-intensive areas it also poses the risk of increasing unemployment. If every area is covered by AI, there will not be any job lefts for humans. Overdependence can diminish human creativity.
- High Cost:
This goes without saying that for creating such high-tech machines a large amount of money is required. The software needs to be upgraded to keep up with the changing needs.
- Cannot replicate humans:
Machines can only work according to how they have been programmed. They will not know how to respond to unfamiliar situations that do not exist in their database
With the digitalisation of so much information online, it poses a threat to privacy especially in the event of a cyberattack on the AI Database.
Need for Regulation
While using AI Technology if a person’s personal details get shared it is important to know what happens with this data. It would also be difficult to know who is liable when someone’s privacy gets breached by AI. Similarly, if an AI Robot gets citizenship, we would have to decide who would be liable for its actions in case it breaks the law or causes harm to a person or property.
Striking a Balance
To prevent the fear of unemployment and a fully automated world devoid of creativity, artificial intelligence should be used in areas where human specialisation is not available or in areas where it is dangerous for humans to work like mining and detonating bombs, etc. Thus, creating a balance between AI and Human employment.
Artificial intelligence should be developed in India and it definitely has a lot of advantages. Although the drawbacks are fewer compared to the advantages, they should not be ignored, especially the cost factor and the unemployment risk. These two barriers are huge in a country like India that is already overpopulated and severely stricken by unemployment hence not striking a balance can worsen the situation for the majority.
What is machine bias and how can it be prevented?
Machine bias occurs when due to a limited database, the machine makes wrong assumptions. AI is a software and if an underlying algorithm has flaws the AI will inherit them. This includes how humans interact with the AI. To deal with this issue it is important to use a representative dataset and ensure AI models go under frequent real-world tests based on real-world datasets.
What are the ways in which AI currently can pose a threat to privacy in India?
AI needs to collect a lot of information to enhance its algorithm. If not monitored properly it could lead to information getting in the wrong hands. India lacks personnel highly trained in maintaining cybersecurity at the given moment and this could lead to a lot of security issues. There is also no surety that the data being collected is being used for the purpose it states it will. It would diminish privacy and there would be no sphere where an individual has autonomy over their words and actions, it could be abused by the government and in a country like ours it could lead to the punishing of thoughts or expressions the government would feel is against it that could have been exchanged privately between individuals. It could unintentionally lead to a slow transition into totalitarianism
In what ways can AI help in the field of law?
AI can help in the field of legal research. Currently, it takes a lot of time to research, with AI it would speed up the time taken and increase the profit of firms. It would help in developing innovative ways of approaching clients. It could help in deciding minor cases like traffic violations for which only a fine can be imposed. It can replace secretarial staff by using software that converts voice to text with high accuracy which not require the intervention of a stenographer thus improving efficiency and reducing the time taken, it can assist in bringing greater transparency in the court’s registry where the filing and listing of cases before judges can be done by AI.
In what ways could AI have helped in the COVID-19 situation?
In hospitals, it could have been of use for testing patients through advanced robotics reducing the risk of spread to doctors/paramedical. Automated sanitizing machines could be deployed in areas thus preventing humans who currently do this from contracting the virus. Machines that could detect signs of COVID-19 in a person could be installed in places frequently visited by people. With the large databases available to AI it could easily analyse ways to find a cure faster than humans would.