Why AI Alignment is Impossible: lessons from the Moral Machine Experiment about ethics' diversity
💡 About: Journey through the complex world of AI ethics, from humorous anecdotes to real-world moral dilemmas posed by self-driving cars. Understand why it is impossible to "align" AI with "human values", at least in the naive sense of aligning it with "the human values": cultural and ethical norms are too diverse to give AI one clear goal!
📚 Quick Lesson: AI's ethical programming is a reflection of our societal values and beliefs. As AI becomes more integrated into our lives, we must grapple with these diverse perspectives to guide its development responsibly.
See the full lecture here:
• DTSC: 10.3 AI alignment (part2): Who deter...
This snippet is taken from the 2023 version of the University of California online course called "Digital Technology & Social Change".
The lecture content can be found here on this channel: • Digital Technology & Social Change: online...
The full course for university credits can be taken here: https://uconline.edu/
The Open Online Version consists of a Specialization that grants an official Certificate (it also includes hands-on labs), and can be found here: https://www.coursera.org/instructor/m...
#EthicalAI
#MoralMachine
#MoralMachineExperiment
#AIEthicsDilemma
#CulturalValuesInAI
#AIDecisionMaking
#aialignment