Building 32: The AI Problems No one Has Solved Yet 🧠🚀
MIT CSAIL Alliances14 August 2026Watch on YouTube
Description
What are the biggest challenges in AI today? According to MIT CSAIL Associate Professor Vincent Sitzmann, it's not just making models smarter, it's teaching them to keep learning and to explore the world the way children do. On this episode of the Building 32 podcast, Sitzmann and host Karen Given discuss continual learning, AI-driven curiosity, and breakthroughs that could define the next generation of intelligent systems. Listen to the full episode of Building 32, meet the host, and more: csail.mit.edu/podcast #MIT #ArtificialIntelligence #AI #MachineLearning #CSAIL #FutureTech #Innovation #Robotics
What you'll learn
- Current AI models become smarter, but the real challenge is teaching them to continuously learn rather than relying only on training phases.
- AI-driven curiosity, where systems actively explore the world like children do, is essential for the next generation of intelligent systems.
- Vincent Sitzmann from MIT CSAIL identifies continual learning and exploration as unsolved core problems that will define future AI development.
Frequently asked questions
What are the biggest unsolved problems in AI according to Vincent Sitzmann?
Why is continual learning important for intelligent systems?
How does AI-driven curiosity differ from traditional machine learning?
Topics
In this video
Related reads
The speakers bringing HumanX to Amsterdam
HumanX Amsterdam opent op 22 september in de RAI met circa 200 sprekers, vijf tracks en drie dagen, waaronder oprichters van Legora, Lovable en Celonis naast enterprise-buyers van Diageo, ING en KLM.
Reaktor's first ecosystem company co-founded in the Netherlands
Reaktor Ecosystem heeft zijn eerste bedrijf in Nederland mede opgericht: GAIA Technology Systems BV, een AI-native software development consultancy gevestigd in Amsterdam, vier maanden na de opening van Reaktors Amsterdam-hub in mei 2026.