# How to build a mind: Exploring insect-inspired AI for autonomous robots Source: https://www.youtube.com/watch?v=BJ8ief4M0V8 Recap page: https://rapidrecap.app/video/BJ8ief4M0V8 Generated: 2026-01-15T09:32:51.065+00:00 --- ## Quick Overview The presentation advocates for a brain-inspired approach to AI autonomy, contrasting it with current, highly resource-intensive methods like SLAM and deep reinforcement learning, by highlighting biological systems (insects, sea squirts, birds) that solve complex navigation problems with far greater efficiency and robustness, suggesting that future AI should mimic this sparse, efficient processing. **Key Points:** - Humanity has long dreamed of creating artificial intelligences that coexist with us, but current AI approaches like SLAM and deep reinforcement learning are computationally intensive, requiring massive resources (e.g., the Apollo program cost $288 billion in inflation-adjusted dollars). - Biological systems offer inspiration for efficient autonomy; for example, fruit fly brains solve navigation problems using only about 100,000 neurons in a cubic millimeter, contrasting sharply with contemporary AI resource demands. - Tunicates (sea squirts) solve movement first, then perception, digesting their own brain once they settle down to free up resources, illustrating efficiency through selective resource management. - Kestrels stabilize their visual perception by locking their heads in three dimensions while hovering, solving the inverse kinematics problem simply to keep their head stable while looking for prey. - The speaker proposes that AI should move away from massive data and compute reliance, focusing instead on sparse, biologically inspired circuits that solve the specific problems of spatial intelligence and navigation robustly, as demonstrated by the fly brain's navigation circuit. - The company Optran offers a vision-only alternative to SLAM, aiming to solve localization, mapping, and navigation in challenging environments (scene change, featureless spaces, reflections, glass, dynamic lighting) while drastically cutting compute and sensor costs. - The speaker concludes by stating that the goal is not Artificial General Intelligence (AGI) but robust, simple, and efficient algorithms inspired by nature to solve real-world autonomy challenges. ![Screenshot at 00:09: Professor James Marshall begins his presentation on building efficient AI minds by showing the title slide, 'How to Build a Mind,' referencing his affiliation with the University of Sheffield's Centre for Machine Intelligence.](https://ss.rapidrecap.app/screens/BJ8ief4M0V8/00-00-09.jpg) **Context:** Professor James Marshall delivers a presentation titled "How to build a mind: Exploring insect-inspired AI for autonomous robots" at the UN AI for Good Global Summit in Geneva, Switzerland (July 8-11, 2025). He contrasts the massive computational cost of current AI solutions, such as those used in autonomous driving and large-scale simulations, with the highly efficient, low-resource autonomy demonstrated by various biological systems, ranging from insects to vertebrates. ## Detailed Analysis Professor James Marshall argues that current approaches to Artificial General Intelligence (AGI) and autonomous systems are overly complex and resource-intensive, evidenced by the comparison between the cost of the Apollo Moon landing ($288 billion adjusted) and the efficiency of biological brains. He posits that humanity should look to nature for solutions to autonomy. He showcases the fruit fly brain, which solves navigation with only 100,000 neurons in a cubic millimeter, vastly more efficiently than modern AI. He cites three lessons from evolution: first, the tunicate's ability to digest its own brain after settling down to conserve resources; second, the kestrel's ability to lock its head to stabilize visual input while hovering, solving inverse kinematics; and third, the simple, robust navigation circuits in insect brains that contrast with complex, brute-force AI algorithms. Marshall introduces his company, Optran, which develops a vision-only alternative to SLAM (Simultaneous Localization and Mapping) designed to work in challenging real-world conditions (reflections, dynamic lighting, featureless spaces) while drastically cutting compute and sensor costs. He emphasizes that the goal is not AGI, but achieving robust, simple, and scalable autonomy by modeling these bio-inspired solutions. ### Introduction & Critique of Current AI - Humanity has dreamed of AI coexistence, but current approaches like SLAM and deep reinforcement learning are prohibitively expensive and compute-intensive - The Apollo Moon program cost $288 billion adjusted, highlighting the scale of resources used for complex tasks. ### Biological Inspiration for Efficiency - Fruit fly brains solve navigation with only 100,000 neurons in a cubic millimeter, showing efficient autonomy - Tunicates digest their own brains after settling down to free up resources, demonstrating resource optimization. ### Evolutionary Lessons for Autonomy - Kestrels stabilize perception by locking their heads to solve inverse kinematics while hunting - Brains evolved over 600 million years to solve movement first, then perception, offering a robust solution compared to current AI. ### Optran's Vision-Only Solution - Optran offers a vision-only alternative to SLAM, designed for localization, mapping, and navigation in challenging environments (scene change, featureless spaces, reflections, dynamic lighting) - This approach slashes compute and sensor costs by an order of magnitude compared to existing methods. ### Conclusion and Future Direction - The goal is not AGI but developing simple, robust, brain-inspired algorithms that can be deployed on edge hardware, contrasting with the massive data-hungry approaches currently dominating AI research. ![Screenshot at 00:09: The opening title slide for Professor James Marshall's presentation, 'How to Build a Mind,' hosted at the UN AI for Good Global Summit in Geneva, July 2025.](https://ss.rapidrecap.app/screens/BJ8ief4M0V8/00-00-09.jpg) ![Screenshot at 00:30: A comparison between a stylized humanoid robot demonstration and the presenter, illustrating the complexity of current AI tasks versus the potential of bio-inspired systems.](https://ss.rapidrecap.app/screens/BJ8ief4M0V8/00-00-30.jpg) ![Screenshot at 01:26: The presenter compares the immense cost of the Apollo Moon Missions to the efficiency of nature, showing an image of the moon's surface next to the lunar lander.](https://ss.rapidrecap.app/screens/BJ8ief4M0V8/00-01-26.jpg) ![Screenshot at 02:58: A colorful, complex visualization of the fruit fly brain's neural structure, highlighting the dense but small scale of biological computation.](https://ss.rapidrecap.app/screens/BJ8ief4M0V8/00-02-58.jpg) ![Screenshot at 11:06: A futuristic-looking humanoid robot in a shopping mall, representing the complex, messy real-world environments that current AI struggles to handle robustly.](https://ss.rapidrecap.app/screens/BJ8ief4M0V8/00-11-06.jpg)