Machine Learning Street Talk (MLST)
Machine Learning Street Talk (MLST)
Welcome! We engage in fascinating discussions with pre-eminent figures in the AI field. Our flagship show covers current affairs in AI, cognitive science, neuroscience and philosophy of mind with in-depth analysis. Our approach is unrivalled in terms of scope and rigour – we believe in intellectual diversity in AI, and we touch on all of the main ideas in the field with the hype surgically removed. MLST is run by Tim Scarfe, Ph.D (https://www.linkedin.com/in/ecsquizor/) and features regular appearances from MIT Doctor of Philosophy Keith Duggar (https://www.linkedin.com/in/dr-keith-duggar/).
Episodes
Mentioned books
164 snips
Mar 24, 2025 • 54min
ARC Prize v2 Launch! (Francois Chollet and Mike Knoop)
Francois Chollet, an AI researcher known for Keras and the ARC challenge, joins Mike Knoop, collaborator on the ARC challenge, to launch the new version of the ARC prize. They discuss how ARC v2 integrates human calibration and adversarial selection, ensuring that even top LLMs struggle against it. The conversation highlights the evolution from ARC v1 to v2, the complexities of AI task design, and the urgent need for rigorous testing methods to bridge the gap between human and AI intelligence in the quest for artificial general intelligence.
183 snips
Mar 22, 2025 • 1h 4min
Test-Time Adaptation: the key to reasoning with DL (Mohamed Osman)
Mohamed Osman, an AI researcher at Tufa Labs in Zurich, discusses the groundbreaking strategies behind his team’s success in the ARC challenge 2024. He highlights the concept of test-time fine-tuning, emphasizing its role in enhancing model performance. The conversation dives into the balance of flexibility and correctness in neural networks, as well as innovative techniques like synthetic data and novel voting mechanisms. Osman also critiques current compute strategies and explores the need for adaptability in AI models, shedding light on the future of machine learning.
140 snips
Mar 19, 2025 • 1h 11min
GSMSymbolic paper - Iman Mirzadeh (Apple)
Iman Mirzadeh, an AI researcher at Apple, presents fresh insights from his GSM-Symbolic paper. He distinguishes between intelligence and achievement in AI, emphasizing that current methodologies fall short. The conversation explores the limitations of Large Language Models in genuine reasoning and the impact of integrating tools for improved AI performance. Mirzadeh advocates for rethinking benchmarks to capture true intelligence and discusses the importance of active engagement in learning processes, suggesting a paradigm shift is essential for future advancements.
375 snips
Mar 18, 2025 • 1h 23min
Reasoning, Robustness, and Human Feedback in AI - Max Bartolo (Cohere)
Max Bartolo, a researcher at Cohere, dives into the world of machine learning, focusing on model reasoning and robustness. He highlights the DynaBench platform's role in dynamic benchmarking and the complex challenges of evaluating AI performance. The conversation reveals the limitations of human feedback in training AI and the surprising reliance on distributed knowledge. Bartolo discusses the impact of adversarial examples on model reliability and emphasizes the need for tailored approaches to enhance AI systems, ensuring they align with human values.
11 snips
Mar 12, 2025 • 1h 41min
Tau Language: The Software Synthesis Future (sponsored)
Mathematician Ohad Asor, a software developer specializing in AI, introduces the innovative Tau language. He highlights the limitations of machine learning in guaranteeing correctness and discusses how Tau provides a logical framework for software development. Asor reveals its potential applications in enhancing blockchain systems and decentralized governance. The conversation touches on program synthesis, user autonomy in software control, and the role of language in AI, advocating for a future where technology aligns more closely with human intent.
11 snips
Mar 10, 2025 • 55min
John Palazza - Vice President of Global Sales @ CentML ( sponsored)
Join John Palazza, Vice President of Global Sales at CentML, as he delves into the vital role of infrastructure optimization for AI and machine learning. He highlights the shift from innovation to production in enterprises, emphasizing efficient GPU utilization and cost management. The conversation touches on the open-source versus proprietary debate, the rise of AI agents, and the importance of avoiding vendor lock-in. Palazza also discusses strategic partnerships with industry giants like NVIDIA that shape business strategies in a competitive cloud landscape.
72 snips
Mar 8, 2025 • 1h 1min
Transformers Need Glasses! - Federico Barbero
Federico Barbero, a lead author at DeepMind/Oxford, dives into the quirks of transformers and why large language models falter at tasks like counting. He reveals fascinating architectural bottlenecks that affect their performance. By drawing parallels with graph neural networks, he sheds light on the softmax function's role in limiting decision-making clarity. But not all hope is lost! Federico shares innovative 'glasses' to enhance transformer performance, including input tweaks and structural modifications to boost their clarity and efficiency.
102 snips
Mar 1, 2025 • 1h 38min
Sakana AI - Chris Lu, Robert Tjarko Lange, Cong Lu
Chris Lu, a recent Oxford DPhil graduate specializing in meta-learning, and Robert Tjarko Lange, a TU Berlin PhD candidate focused on evolutionary algorithms, join forces to discuss innovative approaches to AI. They explore how language models can automate algorithm discovery and enhance training processes. The conversation dives into the interplay of human creativity and AI, addressing challenges like infinite regress in loss functions and the implications of evolutionary optimization. Together, they envision a future where AI systems co-create alongside researchers.
42 snips
Feb 19, 2025 • 51min
Clement Bonnet - Can Latent Program Networks Solve Abstract Reasoning?
Clement Bonnet, a researcher specializing in abstract reasoning, shares his cutting-edge approach to the ARC challenge using latent program networks. He contrasts his method of embedding programs in latent spaces with traditional neural networks, highlighting their struggles with tasks requiring genuine understanding. The discussion dives into the importance of induction versus transduction in machine learning, explores innovative training techniques, and examines the creative limitations of large language models, advocating for a balance between human cognition and AI capabilities.
98 snips
Feb 18, 2025 • 54min
Prof. Jakob Foerster - ImageNet Moment for Reinforcement Learning?
Jakob Foerster, a prominent AI researcher at Oxford University and Meta, joins to discuss the future of AI. He emphasizes the shift from mimicking human behavior to developing intelligent agents that can learn independently. The conversation delves into the importance of open-source AI for responsible innovation and addresses challenges such as AI scaling and goal misalignment. They also explore advancements in deep reinforcement learning, the significance of creativity, and the need for democratization in AI to foster collaboration and mitigate risks.
