Building AI Boston

The Ethical AI Puzzle with guest Cansu Canca on Building AI Boston

38 min

Cansu Canca, Director of Responsible AI Practice at Northeastern University, discusses the critical role of academia in navigating the ethical complexities of AI. She emphasizes that ethical AI is about designing systems that are beneficial, fair, and respect human autonomy, arguing that responsible innovation leads to better technology and business outcomes. Canca highlights the need for practical, iterative approaches to AI ethics, like her PIE model, to keep pace with rapid technological advancements.

Summarized by Podsumo

🎧 Listen 🎙️ Ask about this episode

✨ Key Takeaways

1

Cansu Canca's extensive background: A philosopher by training, she applies her expertise in high-stakes ethics from medical and law fields to AI, working with global organizations like the UN and Interpol.
2

Defining Ethical AI: It's not about AI having morals, but about humans ensuring AI systems are beneficial, align with societal values, respect autonomy, reduce harm, and are fair, often leading to *better* technology.
3

Academia's unique role: Universities, particularly Northeastern, provide an impartial, interdisciplinary environment to develop structured philosophical thinking and practical solutions for AI ethics, free from profit motives.
4

The PIE Model for practical ethics: Canca's "Puzzle-Solving in Ethics" model advocates for an iterative, real-time approach to ethical decision-making in AI development and deployment, rather than slow, committee-based policy.
5

The urgency of ethical design: With AI evolving rapidly, Canca stresses that if we don't actively decide on ethical frameworks, others will, and responsible innovation from the outset is crucial for preventing malfunctioning or biased systems.

💬 Notable Quotes

"These should questions are really philosophical questions... we use the theoretical and the structural thinking that philosophy brings in... to figure out our path forward."

— Cansu Canca
"Ultimately what we are trying to say is the following. We want to create AI systems... that are beneficial for the society, that are in alignment with our general society expectations and our legal frameworks."

— Cansu Canca
"If you are not deciding, somebody else is certainly deciding."

— Cansu Canca