Ethics in AI

Niti Aayog in collaboration with Nasscom rolled out artificial intelligence modules in schools.

Background:

• Project: The National Association of Software and Services Companies (Nasscom), in collaboration with Niti Aayog's Atal Innovation Mission (AIM), launched an artificial intelligence (AI) based module for students of Indian schools.
• Scope: The AI-Base Module will be implemented across 5,000 Atal Tinkering Labs (ATL), empowering 5 million students. The module is likely to be introduced to ATL students on 27 February.
  • The AI module is targeted at young children.
  • The module contains activities, videos and experiments that enable students to work with and learn the various concepts of AI.
  • Its is the first ever industry government academia initiative on such a scale to keep school students abreast of latest technologies.
• Objective: The module combines play and academics, in order to make things interesting. The idea is to make artificial intelligence fun for children to enjoy it, so they can evolve, learn and take India forward.
  • The ultimate goal is to make students industry ready for a AI led Digital future.
• AI potential: By 2030 market for AI is likely to be around $15 trillion, out of which India’s share will be close to $1 trillion.
  • AI has become a strategic lever for economic growth across nations and will continue to be one of the most crucial technologies of future.
  • According to NITI Ayog, use of machine learning and AI can up to 1.3% to India's GDP on an annual basis.

Artificial Intelligence

• Artificial Intelligence is intelligence exhibited by machines. The term was coined in 1956 by John McCarthy at Dartmouth conference, Massachusetts Institute of Technology.
  • Today it has gained prominence due to its multifaceted application ranging from healthcare to military devices.
• Processes in AI: It is a simulation of human intelligence processes such as learning (acquisition of information and rules of it), reasoning (using the rules to reach approximate or definite conclusions), and self-correction by machines, especially computer systems.
• Philosophy and ethics around AI: R&D of AI started with the intention of creating intelligence in machines that we find and regard high in humans. While AI offers much potential in bettering human lives, it also poses many threat to it.

Different forms of AI

• Difference between automation and AI: Automation is basically making a hardware or software that is capable of doing things automatically — without human intervention.
  • AI on the other hand is the science and engineering of making intelligent machines.
• Robotic process automation: Robots can now be programmed to perform high-volume, repeatable tasks normally performed by humans. It supersedes IT automation in agility and adaptability to changing circumstances.
• Natural language processing (NLP): Detecting and processing human language. For example, spam detection, which looks at the subject line of a text/email and decides if it’s junk.
• Speech/facial Recognition: Hearing and grasping different human languages in terms of sentences, while talking to humans.
  • Facial recognition can be used for unlocking phone, detecting intruders etc.
• Pattern recognition: Identifying patterns in data.
• Machine vision: Capturing and analysing visual information using a camera, analog-to-digital conversion, and digital signal processing. It is not bound by biology and can also see through walls. Its application ranges from signature identification to medical image analysis.
• Machine learning: Ability to learn by computers without being explicitly programmed. It involves use of algorithms to parse data, learn from it, and making a predictions.
• Deep learning: It is a subset of machine learning and can be thought of as automation of predictive analytics. It was inspired by the structure and function of human brain.
  • It uses Artificial Neural Networks (ANNs) algorithms which are based on the biological structure of the brain.

Benefits and uses if AI

• Healthcare Sector: Faster, cheaper and more accurate diagnosis and thus improving patient outcomes and reducing costs.
• Business Sector: Applying AI to perform highly repetitive tasks faster and more effortlessly than humans. Provision of better customer service.
• Education Sector: Smart teaching can make learning more engaging. Personalized learning pathways can take education beyond classrooms.
  • Educational processes like grading, rewarding marks etc. can be automated.
• Financial Sector: Better analysis of finance relate data and providing financial advice.
• Legal Sector: Analysing cases and reducing legal workload.
• Agricultural sector: AI can be used to predict advisories for sowing, pest control, input control etc., and bring increased income and stability to farmers.
• Manufacturing sector: Use of Robotics to ease, fasten and detail manufacturing processes. Use of additive manufacturing (3D Printing).
• Industry 4.0: The fourth industrial revolution is based on adoption of AI for progressive automation of production processes. It is the digital transformation of industry.
• Intelligent Robots: Can perform tasks given by humans. They better detect data from physical world, such as light, heat, temperature, movement, sound, bump, and pressure.
  • They have high memory power, learn from their errors and adapt easily.
• Digital assistants: Digital assistants and smart speakers like Siri, Alexa, Cortana, and Google Home use AI to personalise and automatise mundane household functions.
• Strategic games: In May 1997 for examples, an IBM super-computer called Deep Blue defeated world chess champion Gary Kasparov in a chess match.
• Military: Smart border surveillance and monitoring can be applied to enhance security infrastructure.
  • With use of robotic army in counter insurgency and patrolling operations, there can be less damage to human personnel.
• Social welfare: Targeted delivery of services, schemes, and subsidy. Automate government processes can maximize transparency and accountability.
• Road safety: Machine learning can be applied to monitor and control traffic.
• Weather forecasting: Weather forecasting models can become more advanced, and weather related mishaps can well be prepared for in advance.
  • Disaster management can be faster and more accessible with the help of robots and intelligent machines.

Ethical and other issues around AI

• Job loss: Decrease in demand for human labour due to machines and intelligent robots taking over jobs in manufacturing and services sectors.
• Lethal Automated Weapons (LAWs): There is a global arms race for LAWs. Israel Aerospace Industrie's Harpy, for instance crashes into the source of enemy radar signal, destroying the target and itself.
  • Algorithms used in LAWs internalize prejudices, but do not account for human suffering and, therefore, could cause extensive violence.
• Terrorist/cyber threat: Incidents show that even the most advanced security systems are susceptible to In the hands of unruly elements of society, AI can unleash great damage.
  • Terrorists or rogue states could use such weapons/technology on civilians.
• Breaking social codes: It may lead to moral degradation in society due to decreased human to human interactions. It can also alter social and cultural codes.
• Costly and inequitable: Much of the sophisticated AI related technology is unavailable to people on an equitable basis because its costly. For example, majority of people cannot afford technologies like Google Home, or have the requisite support infrastructure needed for it.
  • Similarly, the less developed nations fall behind in development of AI related technologies and infrastructure.
• Privacy concerns: Increasing availability of facial-recognition technology has given rise to concerns regarding privacy, security, and civil liberties.
  • Real-time tracking and facial recognition technology can be misused by the possessors of private information.
• Can it run without human monitoring: While robots and AI technology are deemed to be logical, devoid of emotions, and boasted of to make rational decisions, examples where self-driven cars have caused accidents and led to fatal casualties, raise concern about capability of AI to function without human monitoring.
• Can it cause intentional harm: When the humanoid robot 'Sophie' was made to chat with another robot, after a point, the conversation became serious and even incomprehensible to human faculty. This raises question whether AI technology can go out of hand of its own creator.
• Who should be held responsible for the unintended actions: There is an accountability question as to who should be held responsible for the unintended damages caused by AI related technologies like LAWs or self-driven cars. Should it be the developer, customers, regulatory agencies who allowed it, or just no one.
• Do they deserve certain right: With such fast paced development of AI and their race to achieve mind, consciousness and emotions same human beings, the question of granting them certain rights has arisen; and in many cases citizenship
• Existential risks: Stephen Hawkins warned of fully developed AI taking over human race.

Conclusion

Despite its many threats and challenges, AI is here to stay. AI is a highly collaborative domain, and any framework aimed at promoting AI needs to be aligned accordingly. A multi-pronged approach, involving various stakeholders is required for promoting development and use of AI in different fields. It is the global responsibility of all nations to make AI-ecosystem more accurate, just, and one that is withdrawn of the many risks that it can pose.