THE SEARCH FOR ETHICAL AI
People and machines should be nurtured to behave ethically – Sanjeewaka Kulathunga
Since the beginning of civilisation, ethics and leadership have been linked to each other like two sides of a coin. There was no possibility for sustainable leadership to exist in the corporate world by the 21st century without a properly crafted code of ethics.
However, with the increased use of AI enabled solutions, ethics and leadership are moving into uncharted territory.
In broader terms, AI is the sort of technology that’s able to imitate intelligent human behaviour with vast capabilities. Machine learning, deep learning, natural language processing and computer vision are the four major categories of contemporary AI systems. And they’re statistically enabled to improve and develop their performance without following externally programmed instructions.
By the dawn of the Fourth Industrial Revolution, managing AI that’s similar to or more advanced than humanistic ethics has come under the spotlight in the industrial and business worlds. The ethics of AI is specific to robots and other artificially intelligent creations; typically, it’s bifurcated into roboethics and machine ethics.
In roboethics, the ethical behaviour of humans is concerned with the whole process of design, construction and use of artificially intelligent creations. In contrast, machine ethics is concerned with the ethical features and performance of artificial moral agents.
The performance of machines has already outpaced the capacity of the human brain in most fields such as accountancy, medicine, law and so on. Ultimately, whatever code of ethics is algorithmically installed in a machine can only be created relative to the ethical perception of the human mind.
‘Robot rights’ is a unique and novel concept, which is based on the notion that people should have moral obligations to their machines that are similar to human or animal rights, as such robots are also endowed with the right to exist and perform their own missions.
For instance, the humanoid Sophia was granted honorary citizenship in Saudi Arabia in October 2017. Further, a contemporary philosophy called ‘sentientism’ has also greatly influenced extending moral obligations not only to animals but also robots.
AI developers will have to act as the representatives of future humanity with an ethical obligation to be transparent in their efforts. Meanwhile, nonprofit research company Open AI was established by Elon Musk to develop open source artificial intelligence domains for the benefit of future human civilisation. However, open source for AI is not yet transparent and ethically reliable.
The Institute of Electrical and Electronics Engineers (IEEE) has been attempting to establish a standardisation code to make AI transparent and ethical for the wellbeing of humanity. It has identified multiple scales of transparency for different users in the AI industry.
In addition to this, there’s concern about bringing the full capacity of artificial intelligence into the global market due to its harmful effects on employment opportunities for people around the world.
Global tech giant Microsoft is also hesitant to allow universal access to its face recognition software because of ethical considerations even though it could sell this software for a hefty price in the global market.
A code of ethics for AI is collectively updated by AlgorithmWatch and shared online for global consumption. The UN and EU are presently working hard to craft strategies for regulating AI to adopt an ethical approach that benefits every global citizen.
For example, the High-Level Expert Group on Artificial Intelligence (AI HLEG) related to the EU has taken the initiative to publish policy and investment recommendations, for ethical and trustworthy AI.
With the advent of artificial intelligence, machines are acquiring the autonomous ability to repair themselves and grow without the intervention of human beings. This can be a potential threat to the existence of human civilisation – though AI is acting as a problem solving tool with greater promise, cyber-attacks and social manipulation are proving to be its dark side.
Corporate leaders need to streamline their organisational approach to the AI life cycle from an ethical perspective. They should develop and enforce corporate strategies, policies, procedures and standards for AI based solutions while aligning them with corporate vision, mission, objectives and so on.
Further, machines and employees should be developed to behave ethically without conflicts of interest. Protocols should be established to incentivise employee ethical behaviour while stimulating ethical decision making along the AI life cycle. This is an important step towards achieving sustainable ethical AI leadership.