6 Big Ethical Questions about the Future of AI

For this assignment, I selected the Ted Talk video, “6 big ethical questions about the future of AI”, by Genevieve Bell (Bell, 2020). The talk focuses on integrating Artificial Intelligence (AI) into society and the responsibility that occupies those decisions. The scaling of technology is predictable and assumed to be stable; however, AI scaling must be done safely. The 3Ai Institute at the Australian National University, started in 2017, has a single mission; to establish a new branch of engineering that takes AI safely, sustainably, and responsibly to scale. The institution believes the asking and framing of questions are as important as problem-solving. Questioning reveals new possibilities and challenges that often lack consideration during traditional problem-solving.

In the video, Professor Bell elaborates on six key ethical questions for taking AI safely, sustainably, and responsibly to scale.

1)  Autonomy. Is the system autonomous?  Autonomous systems can take action without being told to do so.

2)  Agency. Does the system have agency?  Agency refers to controls and limitations that prevent specific actions under certain conditions.

3)  Assurance. How do we think about assurance?  Assurance includes all the pieces that comprise a system, including safety, security, trust, risk, liability, manageability, ethics, law, and regulations.
 
4)  Indicators. What will the indicators be that measure success?  The two most common measures of a system are productivity and efficiency. AI-driven systems must also be sustainable, safe, and responsible.
 
5)  Interfaces. What will the interfaces be?  As AI drives new technology and systems, the methods we interact with these systems will change.
 
6)  Intentionality. What is the system’s intent?  Intentionality is the most critical question and addresses the purpose and reasoning behind the system. We must consider what world the system is creating and its relationship to the world we live in today.

In considering an AI-Driven system, the six questions are not simple. When framing the questions correctly, they contribute to designing, building, regulating, and even decommissioning legacy systems. Further, the questions are not independent but thoughtfully addressed as a collaborative effort. The questions enable both scalability and boundaries, which are especially important for safety. The institute anchors the six questions on previous work in cybernetics, as early as the 1950s when researchers addressed concerns around control and communication of technology.

Two forces to elaborate on in consideration of the six questions include intentional and ethical. While the future often appears optimistic in the AI domain, good stewards must consider both intended and unintended (but predictable) consequences that might surface from the solution. Poorly implemented AI systems may yield ambiguous models, bias, or liabilities that enable an adversary. The six questions help address intentionality and mitigate risks (Zhen et al., 2019).

Ethics is another force that is critical to enabling or bounding AI solutions. For example, in the healthcare field, privacy, ownership, and efficacy are essential elements to both the provider and patient, requiring careful protection and stewardship of information. While the goal is to provide better care and reduced costs through system efficiency, failure to address ethical sensitivities jeopardizes public trust in the entire system (Morley et al., 2020).

References 

Bell, G. (2020). 6 big ethical questions about the future of AI https://www.ted.com/talks/genevieve_bell_6_big_ethical_questions_about_the_future_of_ai#t-426100

 

Morley, J., Machado, C. C. V., Burr, C., Cowls, J., Joshi, I., Taddeo, M., & Floridi, L. (2020). The ethics of AI in health care: A mapping review. Social Science & Medicine, 260, 113172. https://doi.org/10.1016/j.socscimed.2020.113172

 

Zhen, J., Beam, A., Saria, S., & Mendonça, E. (2019). Potential trade‐offs and unintended consequences of AI. Artificial Intelligence in Health Care: The Hope, the Hype, the Promise, the Peril, 89. https://ehealthresearch.no/files/documents/Rapporter/Andre/2019-12-AI-in-Health-Care.pdf#page=113