13 Tommaso Poggio

Tomaso Poggio

Researcher

Tomaso Armando Poggio, is the Eugene McDermott professor in the Department of Brain and Cognitive Sciences, an investigator at the McGovern Institute for Brain Research, a member of the MIT Computer ...

Source: Wikipedia

  • Place of birth: Genoa, Italy
  • Books: Visual Cortex and Deep Networks: Learning Invariant Representations
  • Notable students: Demis Hassabis; Anya Hurlbert; Christof Koch; Partha Niyogi; Amnon Shashua
  • Awards: AAAI Fellow, Max Planck Research Award, and Azriel Rosenfeld Award
  • Affiliation: Massachusetts Institute of Technology
  • Research interests: Machine Learning, Learning Theory, AI, and more

The Main Arguments

  • Understanding Visual Intelligence: Poggio emphasizes the importance of visual intelligence, particularly how humans perceive and navigate their environment. He argues that understanding the visual cortex and how we interpret our surroundings is crucial for advancing AI systems that can replicate these capabilities. This point underscores the need for AI to achieve a level of situational awareness similar to humans.

  • The Role of Self-Awareness in Perception: Poggio discusses an experiment conducted at Google X, where subjects experienced a sense of self-awareness while viewing the world through a robot's perspective. This highlights the connection between scene understanding and self-awareness, suggesting that a deeper understanding of consciousness may be necessary for developing advanced AI.

  • Curiosity and Collaboration in Science: Poggio shares insights on what it takes to succeed in science and engineering, emphasizing curiosity and the joy of collaboration. He believes that working with other curious minds enhances the discovery process, suggesting that a supportive and fun environment is essential for innovation.

  • The Nature of Intelligence: Poggio reiterates that intelligence—both human and artificial—is one of the greatest challenges in science. He posits that solving the mysteries of intelligence could lead to breakthroughs across various fields, reinforcing the interconnectedness of scientific disciplines.

  • Ethics and the Hard Problem of Consciousness: The conversation touches on the ethical implications of AI and the philosophical challenges surrounding consciousness. Poggio suggests that while AI can be intelligent, it may not need consciousness to function effectively, raising questions about the moral responsibilities of creating intelligent systems.

Any Notable Quotes

  • "The problem of intelligence is the greatest problem in science."

  • This quote encapsulates Poggio's view on the significance of understanding intelligence, framing it as a foundational issue that could lead to breakthroughs across various fields.

  • "We can learn from a small number of examples, while machines need vast amounts of labeled data."

  • This highlights the inefficiency of current AI training methods and underscores the need for advancements in unsupervised learning.

  • "Deep networks are much more powerful than shallow networks when the function has a compositional structure."

  • This statement emphasizes the importance of understanding the architecture of neural networks in relation to their capabilities.

  • "I think ethics is learnable."

  • Poggio's assertion that ethics can be understood and taught to machines suggests a pathway for developing responsible AI systems.

  • "Consciousness is a difficult problem to define."

  • This quote reflects the complexity of the consciousness debate and its implications for AI development.

Relevant Topics or Themes

  • Visual Intelligence: The episode delves into the intricacies of visual intelligence, exploring how humans perceive their environment and the implications for AI. Poggio's discussion of the Google X experiment illustrates the potential for AI to achieve a form of situational awareness.

  • Collaboration in Science: Poggio emphasizes the importance of collaboration and a supportive environment in scientific discovery. He shares personal anecdotes about his experiences with other researchers, highlighting how curiosity and fun can drive innovation.

  • Artificial General Intelligence (AGI): The quest for AGI is a recurring theme, with discussions on the challenges and timelines associated with creating machines that can think and learn like humans. Poggio's insights suggest that understanding human intelligence is key to achieving AGI.

  • Ethics in AI: The ethical considerations surrounding AI development are a significant focus, with discussions on how to instill ethical frameworks in intelligent systems. Poggio's belief that ethics can be taught to machines opens up avenues for responsible AI development.

  • Consciousness and Self-Awareness: The philosophical implications of consciousness in AI are explored, questioning whether machines can or should possess self-awareness. Poggio's insights into the hard problem of consciousness highlight the complexities involved in replicating human-like intelligence.

Overall, the episode presents a rich tapestry of ideas that intertwine the fields of neuroscience, artificial intelligence, ethics, and philosophy, providing listeners with a comprehensive understanding of the current landscape and future directions in AI research. The engaging dialogue between Fridman and Poggio fosters a thought-provoking exploration of these critical topics.