Episode
Melanie Mitchell: Concepts, Analogies, Common Sense & Future of AI
1:53:07
Published: Sat Dec 28 2019
Description
Melanie Mitchell is a professor of computer science at Portland State University and an external professor at Santa Fe Institute. She has worked on and written about artificial intelligence from fascinating perspectives including adaptive complex systems, genetic algorithms, and the Copycat cognitive architecture which places the process of analogy making at the core of human cognition. From her doctoral work with her advisors Douglas Hofstadter and John Holland to today, she has contributed a lot of important ideas to the field of AI, including her recent book, simply called Artificial Intelligence: A Guide for Thinking Humans. This conversation is part of the Artificial Intelligence podcast. If you would like to get more information about this podcast go to https://lexfridman.com/ai or connect with @lexfridman on Twitter, LinkedIn, Facebook, Medium, or YouTube where you can watch the video versions of these conversations. If you enjoy the podcast, please rate it 5 stars on Apple Podcasts, follow on Spotify, or support it on Patreon. This episode is presented by Cash App. Download it (App Store, Google Play), use code "LexPodcast". Episode Links: AI: A Guide for Thinking Humans (book) Here's the outline of the episode. On some podcast players you should be able to click the timestamp to jump to that time. 00:00 - Introduction 02:33 - The term "artificial intelligence" 06:30 - Line between weak and strong AI 12:46 - Why have people dreamed of creating AI? 15:24 - Complex systems and intelligence 18:38 - Why are we bad at predicting the future with regard to AI? 22:05 - Are fundamental breakthroughs in AI needed? 25:13 - Different AI communities 31:28 - Copycat cognitive architecture 36:51 - Concepts and analogies 55:33 - Deep learning and the formation of concepts 1:09:07 - Autonomous vehicles 1:20:21 - Embodied AI and emotion 1:25:01 - Fear of superintelligent AI 1:36:14 - Good test for intelligence 1:38:09 - What is complexity? 1:43:09 - Santa Fe Institute 1:47:34 - Douglas Hofstadter 1:49:42 - Proudest moment
Chapters
Use code LEXPodcast and download Cash App from the App Store or Google Play to get $10 and donate $10 to First, an organization inspiring girls and boys to dream of engineering a better world.
00:00 - 02:57 (02:57)
Summary
Use code LEXPodcast and download Cash App from the App Store or Google Play to get $10 and donate $10 to First, an organization inspiring girls and boys to dream of engineering a better world. Timestamps are provided for the start of the conversation, but listening to the ad helps support the podcast.
EpisodeMelanie Mitchell: Concepts, Analogies, Common Sense & Future of AI
PodcastLex Fridman Podcast
The concept of intelligence extends beyond just human beings or animals and can be observed in various complex systems such as insect colonies, cellular processes, immune systems, and societal processes.
02:57 - 18:18 (15:21)
Summary
The concept of intelligence extends beyond just human beings or animals and can be observed in various complex systems such as insect colonies, cellular processes, immune systems, and societal processes.
EpisodeMelanie Mitchell: Concepts, Analogies, Common Sense & Future of AI
PodcastLex Fridman Podcast
Some experts believe that scaling deep learning will lead us to human-level AI, while others argue that understanding the fundamentals of human cognitive systems is essential to achieving true AI.
18:18 - 26:47 (08:29)
Summary
Some experts believe that scaling deep learning will lead us to human-level AI, while others argue that understanding the fundamentals of human cognitive systems is essential to achieving true AI. There is significant divergence among AI experts on opinions about the current state and future of the field.
EpisodeMelanie Mitchell: Concepts, Analogies, Common Sense & Future of AI
PodcastLex Fridman Podcast
This podcast discusses different views on the role of deep learning in the larger framework of AI, including the opposing views of putting it to the side versus combining it with symbolic approaches.
26:47 - 34:01 (07:13)
Summary
This podcast discusses different views on the role of deep learning in the larger framework of AI, including the opposing views of putting it to the side versus combining it with symbolic approaches. Jan Lekun agrees with the developmental psychology, intuition, and physics aspects but still sees a need for learning.
EpisodeMelanie Mitchell: Concepts, Analogies, Common Sense & Future of AI
PodcastLex Fridman Podcast
Copycat is a program that simulates an agent-based system where agents pick things to look at, influence other agents and decide whether something is worth looking at more, creating new concepts by combining existing ones.
34:01 - 39:17 (05:15)
Summary
Copycat is a program that simulates an agent-based system where agents pick things to look at, influence other agents and decide whether something is worth looking at more, creating new concepts by combining existing ones.
EpisodeMelanie Mitchell: Concepts, Analogies, Common Sense & Future of AI
PodcastLex Fridman Podcast
Mental models are essential as they allow individuals to do mental simulations of various situations allowing for predictions and expectations to be developed.
39:17 - 45:53 (06:35)
Summary
Mental models are essential as they allow individuals to do mental simulations of various situations allowing for predictions and expectations to be developed. Concepts and situations encountered in the world require the use of mental simulations to predict what is to happen.
EpisodeMelanie Mitchell: Concepts, Analogies, Common Sense & Future of AI
PodcastLex Fridman Podcast
The development of common sense in AI has been a challenging task due to the need to encode all of the invisible knowledge in a logical representation.
45:53 - 50:19 (04:25)
Summary
The development of common sense in AI has been a challenging task due to the need to encode all of the invisible knowledge in a logical representation. Despite this challenge, people have been working on this issue for a long time.
EpisodeMelanie Mitchell: Concepts, Analogies, Common Sense & Future of AI
PodcastLex Fridman Podcast
The current deep learning approaches used to program artificial intelligence systems are capable of building hierarchies of concepts from raw sensory information.
50:19 - 59:04 (08:45)
Summary
The current deep learning approaches used to program artificial intelligence systems are capable of building hierarchies of concepts from raw sensory information. However, it is unlikely that these approaches can scale up to meet the demands of more complex AI systems in the future.
EpisodeMelanie Mitchell: Concepts, Analogies, Common Sense & Future of AI
PodcastLex Fridman Podcast
Neural networks are limited in representing the world and constructing mental models and analogies, as well as achieving what humans can do.
59:04 - 1:06:29 (07:25)
Summary
Neural networks are limited in representing the world and constructing mental models and analogies, as well as achieving what humans can do. This poses a challenge in creating more advanced AI technology in the future.
EpisodeMelanie Mitchell: Concepts, Analogies, Common Sense & Future of AI
PodcastLex Fridman Podcast
The challenges of autonomous vehicles lie in the many different types of situations the car could encounter with the long tail problem being a significant issue.
1:06:30 - 1:15:47 (09:17)
Summary
The challenges of autonomous vehicles lie in the many different types of situations the car could encounter with the long tail problem being a significant issue. Multitask learning is an emerging approach that could help break these problems into many smaller, more manageable learning problems.
EpisodeMelanie Mitchell: Concepts, Analogies, Common Sense & Future of AI
PodcastLex Fridman Podcast
Defining self-driving or autonomous cars can be tricky because it might require full human intelligence.
1:15:47 - 1:25:33 (09:46)
Summary
Defining self-driving or autonomous cars can be tricky because it might require full human intelligence. However, due to legal and other reasons, the definition of self-driving or autonomous cars will change in the future, making them safer and more accessible.
EpisodeMelanie Mitchell: Concepts, Analogies, Common Sense & Future of AI
PodcastLex Fridman Podcast
Rob Reich and Sam Harris discuss the false dichotomy of intelligence and the complexity of separating rationality, values, and emotions from each other in building superintelligence.
1:25:36 - 1:35:07 (09:30)
Summary
Rob Reich and Sam Harris discuss the false dichotomy of intelligence and the complexity of separating rationality, values, and emotions from each other in building superintelligence. They also explore the idea that any system approaching greater intelligence will have a holistic nature, which naturally emerges to prevent any one dimension from running away.
EpisodeMelanie Mitchell: Concepts, Analogies, Common Sense & Future of AI
PodcastLex Fridman Podcast
In this episode, the host delves into the Turing Test and its implications for AI in the future, as well as discussing the concept of reductionism and whether it is sufficient to help us understand complex systems.
1:35:07 - 1:41:33 (06:25)
Summary
In this episode, the host delves into the Turing Test and its implications for AI in the future, as well as discussing the concept of reductionism and whether it is sufficient to help us understand complex systems.
EpisodeMelanie Mitchell: Concepts, Analogies, Common Sense & Future of AI
PodcastLex Fridman Podcast
Experts suggest that people wanting to explore and implement quantum computing in their work should start by studying the basics and then branching out to more complex problems.
1:41:33 - 1:51:47 (10:13)
Summary
Experts suggest that people wanting to explore and implement quantum computing in their work should start by studying the basics and then branching out to more complex problems. They opine that the integration of quantum computing with different disciplines can accelerate progress in solving real-world problems.
EpisodeMelanie Mitchell: Concepts, Analogies, Common Sense & Future of AI
PodcastLex Fridman Podcast
Lex Fridman invites Melanie Mitchell to talk about the importance of concepts in AI and the open problem faced by the community in forming and fluidly using them.
1:51:47 - 1:52:46 (00:59)
Summary
Lex Fridman invites Melanie Mitchell to talk about the importance of concepts in AI and the open problem faced by the community in forming and fluidly using them.