The three rules of responsible AI: From the lab to the boardroom | David Pereira | TEDxEsade Salon

Quick Overview

David Pereira outlines three essential rules for implementing responsible Artificial Intelligence (AI) in corporations: Rule 1 dictates that new technology necessitates new responsibilities, Rule 2 warns that the race for AI power can lead to negative outcomes if unchecked, and Rule 3 emphasizes that a lack of coordination between internal and external stakeholders results in tragedy, advocating for proactive ethical integration rather than reactive fixes.

Key Points: Rule 1 establishes that new technology like AI brings inherent new responsibilities, citing that companies previously failed to consider privacy rights before the internet's widespread adoption. Rule 2 warns against the 'race' for AI power, noting that companies focus on short-term efficiency gains (e.g., 15% efficiency) while risking significant unanticipated side effects (e.g., 1/3 of use cases). The speaker cites the example of IBM's image recognition technology used for mass surveillance as a costly, high-risk decision where ethical considerations were secondary to capability. Rule 3 warns that 'No Coordination leads to Tragedy,' stressing the need for both internal coordination (e.g., between AI/ethics teams) and external conversations with regulators and civil society before deploying AI. The speaker highlights companies like Unilever, Santander, J&J, and Patagonia as examples of organizations actively mapping responsibilities and using AI to ensure fairness and sustainability, contrasting them with those who only address ethics reactively. The core message is that ethics must be integrated proactively from 'The Lab to the Boardroom' to avoid ethical failures that cost trust and money, as summarized by the quote: 'It's not a principle until it costs you money' (Bill Bernbach).

Context: David Pereira presents at a TEDxEsade Salon event, focusing on the ethical implementation of Artificial Intelligence within corporate structures. The talk is titled "The three rules of responsible AI: From the lab to the boardroom," positioning AI not just as a technical challenge but as a governance and leadership imperative. Pereira draws on examples from major corporations to illustrate the necessity of embedding ethical considerations into the entire AI lifecycle, from development to deployment.