Latest Interviews
Showing 46–60 of 74 interview transcripts.
Clear all filters- Lex Fridman1h 25m
Gary Marcus: Toward a Hybrid of Deep Learning and Symbolic AI | Lex Fridman Podcast #43
Gary Marcus argues that current artificial intelligence lacks the common sense and causal reasoning required for general intelligence, necessitating a shift from pure deep learning to a hybrid architecture that integrates symbolic logic. He contends that true "trustworthy AI" demands the explicit engineering of abstract ethical concepts and diverse testing frameworks, such as a "Turing Olympics," rather than relying on statistical correlations or black-box scaling. Ultimately, Marcus predicts a gradual evolutionary path where AI acquires physical and psychological understanding by mimicking human innate cognitive libraries, rather than through a singular disruptive breakthrough.
- Lex Fridman1h 3m
Peter Norvig: Artificial Intelligence: A Modern Approach | Lex Fridman Podcast #42
This discussion with Peter Norvig outlines the evolution of artificial intelligence from memory-constrained logic to modern neural networks, highlighting a philosophical shift toward defining utility functions and addressing ethical challenges like algorithmic fairness. The dialogue examines the limitations of deep learning in reasoning, the necessity of combining symbolic AI with neural approaches, and the societal risks of attention economies and weaponization rather than existential robot threats. Furthermore, Norvig reflects on the changing nature of programming expertise, the unique dynamics of online education, and future research directions focused on integrating common sense reasoning into code assistants and natural language systems.
- Lex Fridman57 min
Leonard Susskind: Quantum Mechanics, String Theory and Black Holes | Lex Fridman Podcast #41
Stanford Institute of Theoretical Physics founding director Leonard Susskind discusses the necessity of dual mindsets involving arrogance and humility to advance theoretical physics, while distinguishing true quantum computers from classical simulators capable of modeling complex quantum systems. He asserts that string theory fundamentally resolved the historical conflict between quantum mechanics and gravity and predicts that quantum computing's primary utility lies in simulating chemistry and material science rather than solving general algorithmic problems. Additionally, Susskind addresses the emergent nature of spatial dimensions, the thermodynamic origins of time's arrow, and the limitations of using scientific methods to answer questions regarding a purposeful intelligent agent or a simulation-based reality.
- Lex Fridman1h 17m
Regina Barzilay: Deep Learning for Cancer Diagnosis and Treatment | Lex Fridman Podcast #40
MIT professor Regina Barsley argues that the future of scientific breakthroughs in oncology and natural language processing depends on shifting from rigid mechanistic models to probabilistic approaches that address real-world suffering. She highlights critical barriers such as obsolete medical datasets and regulatory inertia, advocating for a patient-donation model to accelerate early cancer detection and generative drug design. Drawing from her personal breast cancer diagnosis, Barsley urges researchers to prioritize high-impact applications over incremental algorithmic improvements, emphasizing that successful adoption relies more on human advocacy than the intrinsic superiority of new ideas.
- Lex Fridman57 min
Vijay Kumar: Flying Robots | Lex Fridman Podcast #37
Vijay Kumar, Dean of Engineering at the University of Pennsylvania and a pioneer in multi-robot systems, discusses the evolution from rigid hexapods to agile aerial swarms inspired by ant behavior. He contrasts engineered swarms requiring global coordinates with natural swarms using local interactions, while highlighting critical challenges in battery density, energy-efficient perception, and the shift toward shared autonomy over human monitoring. Kumar concludes by emphasizing the urgent need for technology literacy to manage the societal implications of weaponized swarms and the difficulty of generalizing robotics to unstructured real-world environments.
- Lex Fridman1h 16m
Yann LeCun: Deep Learning, ConvNets, and Self-Supervised Learning | Lex Fridman Podcast #36
Yann LeCun argues against the inevitability of "evil" AI and the concept of general intelligence, positing instead that human-like capabilities emerge from specialized architectures equipped with working memory and world models trained via self-supervised learning. He contends that true autonomy requires grounding in physical reality through predictive simulations rather than pure reinforcement learning, explicitly rejecting the feasibility of solving complex tasks like autonomous driving without incorporating causal reasoning and continuous constraints. LeCun further warns against industry hype regarding current system capabilities, advocating for benchmarks that measure efficiency in reducing labeled data requirements and the ability to navigate interactive environments rather than static datasets.
- Lex Fridman1h 0m
Pamela McCorduck: Machines Who Think and the Early Days of AI | Lex Fridman Podcast #34
Pamela McCordick's 1979 book *Machines Who Think* chronicles the founding of artificial intelligence through interviews with key figures like John McCarthy and Marvin Minsky while navigating significant institutional resistance from the National Science Foundation. The work reframes AI history by challenging cultural fears of replacement, attributing the field's early "winters" to commercial hype rather than scientific stagnation, and critiquing the male-dominated anxieties surrounding the technology's future. McCordick argues that modern AI's shift toward deep learning risks embedding human biases, yet she remains cautiously optimistic that ethical programming and a focus on complex adaptive systems can ensure responsible development.
- Lex Fridman1h 13m
Keoki Jackson: Lockheed Martin | Lex Fridman Podcast #33
Lockheed Martin is advancing a dual strategy of deep space exploration, exemplified by the Orion spacecraft and Mars Base Camp concepts, while simultaneously integrating advanced AI like the Maya system to enhance human-machine teaming and robotic decision-making. In the defense sector, the company is modernizing strategic deterrence through hypersonic technologies and autonomous platforms such as the F-16 "loyal wingman," all operating under strict human control policies to address emerging geopolitical threats. This approach leverages digital twins, quantum computing, and commercial competition to drive innovation, aiming to establish sustainable infrastructure beyond low Earth orbit while maintaining global security through continuous technological reinvention.
- Lex Fridman58 min
Paola Arlotta: Brain Development from Stem Cell to Organoid | Lex Fridman Podcast #32
Harvard professor Paola Arlotta leads research utilizing patient-derived brain organoids to model the complex, time-dependent developmental processes of the human cerebral cortex, which differ fundamentally from murine models due to distinct cellular sequencing and mechanical forces. This work employs single-cell profiling to identify the molecular and structural origins of neurodevelopmental disorders like autism while simultaneously addressing ethical boundaries to ensure these systems remain disease models rather than attempts to engineer consciousness. Arlotta's findings highlight the brain's intrinsic plasticity and suggest that evolutionary adaptations, such as reduced myelination in newer neurons, enable the flexibility necessary for integrating with evolving technologies and artificial intelligence.
- Lex Fridman58 min
Kevin Scott: Microsoft CTO | Lex Fridman Podcast #30
Microsoft's Chief Technology Officer Kevin Scott outlines a strategic vision where artificial intelligence serves as a democratizing platform to empower individuals and organizations while addressing ethical challenges like data dignity, face recognition bias, and deep fakes. Through partnerships with researchers such as Glenn Weil and Jaron Lanier, the company is pioneering radical market mechanisms and cryptographic solutions to ensure transparent compensation for data contributions and verified content origins. This approach supports Microsoft's broader mission to deploy invisible AI infrastructure across productivity tools, mixed reality, and global problem-solving sectors, aiming to solve critical issues ranging from climate change to workforce management before regulatory frameworks fully adapt.
- Lex Fridman1h 26m
Kai-Fu Lee: AI Superpowers - China and Silicon Valley | Lex Fridman Podcast #27
Dr. Kai-Fu Lee contrasts the Chinese model's reliance on massive data volume and execution with the US focus on algorithmic innovation, predicting China's lead in Level 4 autonomous driving while the US advances toward Level 5 reasoning. He further outlines a shifting labor landscape where routine white-collar tasks face automation, necessitating a strategic pivot toward compassionate service roles and government-backed retraining initiatives. Ultimately, Lee advocates for global cooperation to prevent an AI arms race and emphasizes that enduring human value lies in the capacity for love and ethical judgment rather than technical prowess alone.
- Lex Fridman1h 0m
Rosalind Picard: Affective Computing, Emotion, Privacy, and Health | Lex Fridman Podcast #24
Rosalind Picard, the pioneer of affective computing, argues that while machines can now detect emotional states with over 80% accuracy using wearable sensors, the field remains limited to narrow contexts and poses significant ethical risks regarding privacy and authoritarian surveillance. She advocates for strict regulations that separate emotional analysis from commercial exploitation and criminalizes non-consensual emotion detection, emphasizing that current AI cannot replace genuine human connection or consciousness. Ultimately, Picard urges the industry to shift focus from generating wealth to solving critical health challenges, such as predicting fatal epilepsy seizures, thereby using technology to empower underserved populations rather than consolidating power.
- Lex Fridman1h 9m
Gavin Miller: Adobe Research | Lex Fridman Podcast #23
Adobe Research is advancing creative workflows by shifting from manual pixel manipulation to intent-based automation, exemplified by tools like Project Sky Replacement and Generative Fill that utilize deep learning to bridge the gap between human creativity and machine execution. The lab operates on a collaborative model where AI provides smart defaults and assistive suggestions, allowing human users to intervene for final quality assurance while leveraging vast data libraries to build context-aware educational features. Strategic initiatives including the Sensei platform and high-risk intern programs aim to centralize neural models and drive innovation, ultimately fostering a future where generative AI and immersive technologies enhance content velocity without compromising professional reliability or ethical standards.
- Lex Fridman1h 11m
Rajat Monga: TensorFlow | Lex Fridman Podcast #22
Launched by Google Brain in 2011, the TensorFlow ecosystem has evolved from a proprietary deep learning library into a globally adopted platform with over 41 million downloads and 1,800 contributors. The framework is currently transitioning to version 2.0, which defaults to eager execution and unifies its API around Keras to resolve developer confusion while preserving backward compatibility for enterprise systems. Driven by a distributed governance model and competition from alternatives like PyTorch, the project aims to democratize machine learning across diverse hardware, from mobile devices to cloud TPUs, by simplifying model development and addressing enterprise data organization challenges.
- Lex Fridman1h 13m
Chris Lattner: Compilers, LLVM, Swift, TPU, and ML Accelerators | Lex Fridman Podcast #21
Chris Lattner, the creator of LLVM and Swift and former lead of Tesla's Autopilot software, currently directs compiler infrastructure initiatives at Google including TensorFlow, TPU accelerators, and the emerging MLIR project. He details how the open-source LLVM community unites competing giants like Apple, NVIDIA, and Intel by sharing expensive optimization layers while pioneering techniques that apply machine learning to solve complex register allocation challenges. His career narrative highlights a strategic shift in the industry toward integrated automatic differentiation and dynamic compilation, fundamentally reshaping how diverse languages interact with modern hardware from mobile devices to neural network accelerators.