What are the ethical concerns associated with AI? How can ethics help in resolving them?
Abhi Jain, AIR 34, 2024
[UNESCO's AI declaration] pointed out the ethical issues involved in use of Artificial Intelligence.
Ethical Issues
1. Loss of values: as machine taking decisions for humans Eg Youtube's AI recommendations 2. Super Intelligence syndrome: Machines becoming capable to decide & overpower humans (NASA report) 3. Erosion of ethics: like London's AI-FRT recognition: picks out colored people as criminals 4. Loss of employment & livelihood of 30 million people by 2030 (World Bank report) 5. Privacy violation: like tracking & tracing data & sharing without consent
How can ethics help?
1. Principle of Proportionality: usage of AI should be proportional to stated objective Eg data governance architecture 2. Principle of No-Harm (UNESCO): to cause no harm to human sentiments Eg US' Black people recognized as criminals in [FBI-directory] 3. Transparency & privacy protection 4. [Ethical code of AI] - should be created, as per UNESCO's principles 5. Only assigning complex-tasks to AI 6. Pseudo-AI should be weeded-out
AI-based governance is essential but not at the cost of losing out the value of "human-ness" (Technology without character is a sin).