Summary
Highlights
Flynn Coleman, an international human rights lawyer, outlines the historical emergence of human rights during the Renaissance. She highlights the overwhelming statistics of human rights abuses today, such as extreme poverty, lack of electricity, modern slavery, child soldiers, and displaced people. Coleman argues that while statistics can be numbing, individual stories of bravery and resilience resonate deeply with people, demonstrating the power of narrative.
Coleman reflects on her early work using satellite imagery for genocide prevention, which could only verify atrocities after they occurred. She notes that throughout history, technology has democratized knowledge and amplified voices, from Gutenberg's printing press to the internet and social media. However, she stresses that technology hasn't yet provided tools to prevent humanity's darkest impulses.
The speaker argues that the coming AI revolution will dramatically disrupt human rights, and the future depends on how we choose to develop it. She emphasizes the critical need to teach values and ethical concepts to machines, as AI will soon exceed human intellectual capacity exponentially. Coleman cites futurists like Ray Kurzweil who predict AI surpassing human intelligence within decades and explains that while this is daunting, AI could offer solutions to persistent human rights challenges if programmed with our best values.
Coleman acknowledges the theories and science fiction about AI posing existential threats, referencing warnings from figures like Stephen Hawking. She points out that AI excels at identifying patterns but lacks common sense, making it crucial to deliberately choose which patterns and values we instill. She differentiates values from rules, asserting that values provide a deeper moral compass. The recent example of a Twitter chatbot quickly learning racist rants underscores that the danger lies not in technology itself, but in the people who influence it.
Philosopher Stuart Russell believes instilling values in AI is vital for human survival, and Coleman suggests that human stories are the most natural way to achieve this. By uploading literature, poetry, myths, and fables, AI can learn what it means to be a moral human. However, she warns that simply uploading data isn't enough, as human biases and perspectives can cloud understanding. Coleman's experience with conflict resolution has shown her that the illusion of separation is the root of most conflict, making the concept of equality paramount for machines to understand.
Given humanity's history of human rights abuses, Coleman questions whether we are capable of providing AI with the right answers. She articulates her torment over the trauma and injustice she has witnessed and emphasizes the urgent need to address these questions. Coleman envisions a future where machines embody the bravery and ideals of figures like Malala and Mandela, and even help humans learn these values. She warns that if we fail to impart our highest ideals, we risk becoming like our most fearful idea of robots, echoing Sydney J. Harris's caution that the real danger is people thinking like computers.
Coleman concludes by reflecting on how our fictional portrayals of loyal and generous robots, like R2-D2 and Wall-E, demonstrate our aspirations for humanity. She states that stories are 'data with a soul' and that for AI to truly advance, it needs to understand the rich tapestry of human culture, history, and literature. This process of building ethical AI must be a continuous cycle of striving to do better than we have in the past. She poses a final question: Can we create machines that embody our highest values and help us improve ours, and what future do we want to see when we look through the looking glass?