Ethical AI - Heart Vs Data Convictions ?

The way Artificial Intelligence (AI) is rapidly penetrating in almost every industry, we may overlook another perspective related to AI and that is, the ethical AI. Technologists nullify the existence of something called ethical AI, believing that the ethical dilemmas related to AI can be overcome with code. The proponents of ethical AI state that AI should treat people in a non-discriminatory way, empower them equally, perform in a reliable and safe way, be secure & understandable and maintain privacy with algorithm accountability. Is this possible to achieve? Or AI is still itself automating the ethics?

My 2 cents. Thanks to the speakers in AI Summit Bangalore for igniting this thought process.

AI to be aligned with current human values and fair with user data rights, as well as be socially beneficial and under human control.

The Giants Behind “Ethical AI”

Some giant and mainstream players like IBM, Microsoft and DeepMind, have already started to establish their ethical principles for the formation and implementation of AI. There is a wide-spread belief that AI codes are the basis of ethical automation and that once a board consensus is achieved around such codes, the dilemma of seeing an ethically positive future for computer code will have started to be solved. But, is this view, right? We believe it’s not!

The Global Dilemmas & Underlying Ethical Beliefs

Recently came across this beautiful article, Nature Machine Intelligence, while we can see a global junction rising around these ethical principle; transparency, justice and fairness, responsibility, and privacy, what exactly these principles mean, is quite another stuff. There is still some practical divergence related to the way these principles are interpreted, why they are important, what domain, issue or actors they are related to, and the way these should be applied.  

Ethical views vary from person to person. Maybe the most effective myth of our time is related to machine super intelligence, the capacity of which concurrently turns A.I. ethics into a deal with current threats and a design process geared at evicting banish human unreason, at subcontracting society’s highest dilemmas to presumed superhuman entities in the shape of A.I.

Blending Community & Collaboration – The Basis of Ethical AI, A community is developed when its members have mutual goals, values and interests. In terms of A, it means bringing together the right people to solve the mutual problem. Taking into an account a huge effect that most of today’s AI solutions put on society, it would be damaging to devise solutions in isolation from the social circumstances of people.

The Rise of Collaborative AI

The idea of collaborative AI blends the community and collaboration by uniting experts, organisations, and students to develop solutions that are trusted, ethical and value-oriented; and most importantly, are beneficial to people and society. It brings the following opportunities:

• Providing students an opportunity to acquire AI skills by working on real-world projects.
• Letting organisations build trust, develop value-added products, and harness crowd wisdom, diversity, and inclusion.
• Easy access to huge volume of high-quality data via project communities
• Democratising AI by connecting global talent and real-world AI projects

Humanising AI – Harnessing the Human Culture

We have already started to give AI human-like names, physical features and mannerisms. This is mainly to grow human acceptance, as there are a lot of examples when humans were cruel to robots. People are willing to endure and use AI when it more thoroughly resembles the human form, and so the confidence has been to produce personal assistants (Siri and Alexa) and social robots, like those introduced in healthcare centers and nursing homes, which feel like human and inspire fellow-feeling in their users.

Humanising AI could also help integrate with human culture as well as society. Perhaps it means that a slight humanising is okay, or at least worth a try, to give robots the potential for such emotions. But it would be mandatory for the creators to be more hands-on with the learning process. To engage with the growth of higher functions to have more context and to help with understanding. This may lessen previous issues where the learning and knowledge has gone awry, leading to unexpected and bizarre results.

Concluding thoughts

There is no definite set of ethical principles related to the AI; though be realizing the significance of collaborative AI, combining community and collaboration, AI has the potential to be one of the greatest and most beneficial technologies of future for the social and community welfare and development – an idea that surely forms the basis of ethical AI.