Stuart Russell, a leading Professor of Computer Science at UC Berkeley and author of 'Human Compatible,' delves into the intricate challenges of AI control. He discusses the unsettling consequences of superhuman AI and the manipulation of user behavior by social media algorithms. The conversation highlights the urgent moral implications of AI technology, the necessity for human oversight, and a reevaluation of AI design to align with human values. Stuart emphasizes the importance of understanding the risks of misaligned objectives, drawing parallels with historical figures like Alan Turing.
01:49:02
forum Ask episode
web_stories AI Snips
view_agenda Chapters
menu_book Books
auto_awesome Transcript
info_circle Episode notes
question_answer ANECDOTE
King Midas Analogy
King Midas' golden touch illustrates the dangers of poorly defined objectives in AI.
Getting exactly what you wish for can be disastrous, like AI fulfilling a flawed objective.
insights INSIGHT
Deep Learning's Knowledge Gap
Current deep learning systems lack true knowledge and struggle to accumulate it over time.
They learn input-output functions, not fundamental principles like Newton's laws, hindering their ability to generalize.
insights INSIGHT
Language Models and Ptolemaic Astronomy
Language models like GPT-3 predict the next word without understanding the meaning or context.
They resemble Ptolemaic astronomy, accurately predicting patterns without grasping underlying causal mechanisms.
Get the Snipd Podcast app to discover more snips from this episode
Derek Parfit's "Reasons and Persons" is a landmark work in contemporary philosophy, profoundly impacting discussions on personal identity, ethics, and rationality. Parfit challenges traditional notions of the self, arguing that our sense of personal identity is less coherent than we assume. He explores the implications of this for our moral obligations, particularly concerning future generations. The book delves into the complexities of decision-making under uncertainty, examining how we should weigh our present interests against the potential consequences of our actions for the future. Parfit's rigorous analysis and thought-provoking arguments have had a lasting influence on various fields, including ethics, political philosophy, and decision theory. His work continues to stimulate debate and inspire new research.
Human Compatible
Artificial Intelligence and the Problem of Control
Stuart J. Russell
In this book, Stuart Russell explores the concept of intelligence in humans and machines, outlining the near-term benefits and potential risks of AI. He discusses the misuse of AI, from lethal autonomous weapons to viral sabotage, and proposes a novel solution by rebuilding AI on a new foundation where machines are inherently uncertain about human preferences. This approach aims to create machines that are humble, altruistic, and committed to pursuing human objectives, ensuring they remain provably deferential and beneficial to humans.
The Machine Stops
E. M. Forster
Published in 1909, 'The Machine Stops' is a prophetic tale by E.M. Forster that depicts a future where humans live underground in isolated rooms, relying on 'The Machine' for all their needs. The story follows Vashti and her son Kuno, who live on opposite sides of the world and communicate through the Machine. Kuno's desire to experience the surface world and his warnings about the Machine's impending failure are met with resistance from Vashti, who is deeply entrenched in the Machine's culture. The narrative explores themes of technological dependence, social isolation, and the erosion of human values, culminating in a catastrophic event when the Machine stops functioning, leading to the collapse of the society it supports.
Stuart Russell is a Professor of Computer Science at the University of California and an author.
Programming machines to do what we want them to is a challenge. The consequences of getting this wrong become very grave if that machine is superintelligent with essentially limitless resources and no regard for humanity's wellbeing. Stuart literally wrote the textbook on Artificial Intelligence which is now used in hundreds of countries, so hopefully he's got an answer to perhaps the most important question of this century.
Expect to learn how AI systems have already manipulated your preferences to make you more predictable, why social media companies genuinely don't know what their own algorithms are doing, why our reliance on machines can be a weakness, Stuart's better solution for giving machines goals and much more...
Modern Wisdom 
