The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)

In this conversation, Alexandre Pesant, AI lead at Lovable, delves into vibe coding, a transformative practice in software development. He explains how this approach shifts focus from typing code to expressing intent through higher-level abstractions. Alex shares insights on coding agents and the critical role of context engineering. He discusses Lovable’s evolution from complex agent systems to simpler workflows and back as technologies improved. Listeners also gain perspectives on AI's impact on engineering roles and the importance of robust evaluations in creating successful AI-native development tools.

Kunle Olukotun, a Stanford professor and chief technologist at SambaNova Systems, dives into reconfigurable dataflow architectures for AI. He explains how this innovative approach enhances AI inference by dynamically matching hardware to model dataflows, leading to reduced bandwidth bottlenecks. Kunle also highlights the benefits of fast model switching and efficient multi-model serving, crucial for low-latency applications. Plus, he explores future possibilities of using AI to create compilers for evolving hardware setups, offering insights into significant performance improvements.

Jacob Buckman, co-founder and CEO of Manifest AI, dives deep into the world of long-context transformers. He discusses innovative techniques like windowed attention and the revolutionary Power Retention approach, which melds attention and recurrence for astonishing training speeds. Buckman also shares insights on Manifest AI's open-source tools, Vidrial and PowerCoder, and explores the significance of metrics in measuring context utility. Learn about the balance between state and weight for optimal compute architectures and the future potential of context length in AI.

Illia Polosukhin, co-author of the groundbreaking 'Attention Is All You Need' paper and co-founder of Near AI, dives into the intersection of decentralization and private AI. He shares his journey from Google to revolutionizing blockchain with the NEAR Protocol. Topics include confidential computing for user data protection and the risks of AI centralization. Illia emphasizes a trust-building approach through open model training and formal verification. He also explores tokenized incentives for data contributions and the future of privacy in AI deployment.

Oliver Wang, a Principal Scientist at Google DeepMind, shares insights on the transformative capabilities of the Gemini 2.5 Flash Image, codenamed 'Nano Banana.' He explores the evolution from specialized image generators to integrated multimodal agents, highlighting how Nano Banana generates and edits images while preserving consistency. Oliver discusses the balance between aesthetics and accuracy, unexpected creative applications, and the future of AI models that could ‘think’ in images. He also warns about the risks associated with training on synthetic data.

In this discussion, Aditi Raghunathan, an assistant professor at Carnegie Mellon University, tackles the limitations of large language models (LLMs). She presents insights from her award-winning paper on enhancing creativity beyond next-token prediction. Aditi introduces the innovative 'Roll the dice' method to foster randomness and 'Look before you leap' for deeper thought processes. The conversation also covers the paradox of 'catastrophic overtraining' and her pursuit of more controllable models through concepts like 'memorization sinks.' Her research aims to reshape our understanding of AI adaptability.

Join Animesh Koratana, founder and CEO of PlayerZero, as he delves into the exciting world of AI-assisted coding tools. Discover how rapid advancements in AI have created a gap between code generation speed and effective maintenance processes. Animesh discusses the innovative use of code simulations to build a memory bank of past bugs and predictive models to enhance software reliability. He also explores the future of the software development lifecycle, emphasizing the need to adapt organizational workflows for an AI-driven landscape.

Christian Szegedy, Chief Scientist at Morph Labs and a pioneer of the Inception architecture, discusses the future of AI through autoformalization. He explains how translating mathematical concepts into formal logic can enhance AI safety and reliability. The conversation highlights the contrast between informal reasoning in current models and the provable correctness of formal systems. Szegedy envisions AI surpassing human scientists in specialized fields while serving as a tool for humanity's self-understanding.

Prince Canuma, an ML engineer and open-source developer known for his contributions to Apple's MLX ecosystem, discusses his journey in optimizing AI for Apple Silicon. He shares insights on adapting models, the trade-offs between GPU and Neural Engine, and innovative techniques like pruning and quantization for enhanced performance. Prince introduces 'Fusion,' a unique approach to model behavior without retraining, and presents Marvis, a real-time voice agent. His vision for future AI focuses on multimodal models that adapt seamlessly across various media.

In this engaging discussion, Jack Parker-Holder and Shlomi Fruchter, both researchers at Google DeepMind, dive into Genie 3, a groundbreaking model that creates playable virtual worlds. They explore the evolution of world models in AI, emphasizing their importance for decision-making and planning. The duo sheds light on Genie 3’s real-time interactivity, visual memory capabilities, and the challenges faced in its development. They also touch on the innovative concept of promptability, showcasing how the model can dynamically manipulate virtual environments, paving the way for exciting applications.