Michael I. Jordan is a professor at Berkeley, and one of the most influential people in the history of machine learning, statistics, and artificial intelligence. He has been cited over 170,000 times and has mentored many of the world-class researchers defining the field of AI today, including Andrew Ng, Zoubin Ghahramani, Ben Taskar, and Yoshua Bengio. EPISODE LINKS: (Blog post) Artificial Intelligence—The Revolution Hasn’t Happened Yet 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". Here's the outline of the episode. On some podcast players you should be able to click the timestamp to jump to that time. OUTLINE: 00:00 - Introduction 03:02 - How far are we in development of AI? 08:25 - Neuralink and brain-computer interfaces 14:49 - The term "artificial intelligence" 19:00 - Does science progress by ideas or personalities? 19:55 - Disagreement with Yann LeCun 23:53 - Recommender systems and distributed decision-making at scale 43:34 - Facebook, privacy, and trust 1:01:11 - Are human beings fundamentally good? 1:02:32 - Can a human life and society be modeled as an optimization problem? 1:04:27 - Is the world deterministic? 1:04:59 - Role of optimization in multi-agent systems 1:09:52 - Optimization of neural networks 1:16:08 - Beautiful idea in optimization: Nesterov acceleration 1:19:02 - What is statistics? 1:29:21 - What is intelligence? 1:37:01 - Advice for students 1:39:57 - Which language is more beautiful: English or French?
In this podcast episode, Michael I. Jordan, a professor at Berkeley and one of the most influential people in the history of machine learning, discusses the connection between AI and empowering human beings, as well as the use of AI in the stock market with the Cash App.
The field of artificial intelligence (AI), specifically machine learning, is still in its early stages and breakthroughs are not on the immediate horizon, despite some people's claims, due to the complex and theoretical aspects required to truly understand and apply human data and decisions towards it.
In this podcast, experts discuss the possibility of breakthroughs in engineering versus scientific fundamental principles, using examples from previous centuries such as electrical and chemical engineering.
The discussion focuses on the principles of artificial intelligence and how it is applied in large collections of decisions. The guest emphasizes that this field is not limited to a single agent or human decision-making process, but rather involves a range of AI systems and representation methods.
The speaker questions whether personalities or fundamental principles are what drive progress in science and mentions their hesitation to trust the predictions of neural nets for their personal medical decisions.
The difference between pattern recognition and decision-making tools is a constrained lab dataset versus consequential decisions in the messiness of the real world that touches human beings with market forces impacting decisions. Companies like Amazon use both types of tools with half their workforce working on decision-making tools and the other half on pattern recognition in the context of strategic goals for data collection and analysis.
The speaker discusses their personal experience with Amazon and how they initially disliked the company for putting local booksellers out of business. They then go on to discuss the effectiveness of good recommender systems and their limitations.
The trade-off for privacy in exchange for personalized recommendations comes with a range of benefits such as the creation of tailored content and a better user experience. However, in areas such as politics, recommendation systems can lead to controversial and messy conversations.
There is growing skepticism within Amazon about the company's push to upsell customers on additional products, especially with the use of subtle tactics that may not be obvious to the consumer. Despite the practice being relatively inexpensive, many within the company are still questioning its effectiveness.
The annoyance caused by advertising has surpassed the information it provides, and society's changing behavior might lead to control being taken away from people, thereby reducing automated advertising. Transparency in consumer-creator relationships should be the primary goal of advertising.
In this podcast, J Isaacson discusses the flaws in the current music licensing model and his perspective on the future of the industry. He also talks about his company, Jukin Media, and their success in licensing music for various entertainment platforms.
Can AI have conversations to determine music preferences? The speaker explores the idea of using natural language processing to provide personalized music recommendations.
The use of AI technology can empower individuals to have more control over their privacy and daily interactions, ultimately leading to increased human happiness.
The speaker discusses the importance of building trust and locality within communities to facilitate productive discussions about topics such as privacy.
The speaker talks about anticipating opponent's moves in simultaneity games and the optimization method of looking for saddle points.
The widespread use of TensorFlow has led many young people entering the field of machine learning to believe that there is no need for further algorithm development. However, as the field becomes more stochastic and decentralized, there is a need for new algorithms to be developed to solve complex problems.
The properties of stochasticity are appealing in high dimensions for the law of large number reasons. Stochasticity can save us from some of the particular features of surfaces that are discontinuous in a first derivative, which creates issues with only using gradients.
This episode discusses the origins of statistics and its evolution throughout history through analyzing data for the state and the development of probability to explain gambling situations.
The frequentist approach is to view data as random and average over the distribution, while the Bayesian approach is to update prior beliefs based on observed data.
The concept of intelligence extends beyond just human or animal abilities to include systems such as markets and their principles. True intelligence encompasses a larger scope of understanding.
The ability to communicate and have empathy towards others is crucial in building relationships and understanding different perspectives. Focusing too much on fantasies like science fiction may hinder personal growth and development.
Lex Fridman speaks with Michael I. Jordan, a computer scientist and professor at the University of California, Berkeley, about the ideas in his book "Machine Learning: A Probabilistic Perspective." They discuss the applications and limitations of machine learning, the role of probability in AI, the unavoidable ambiguity in language, the necessity of broad thinking, and more.