Microsoft Research Podcast

Researcher Michel Galley explores how he and fellow researchers combined new and existing data to create MathVista, an open-source benchmark for measuring the mathematical reasoning capabilities of foundation models in scenarios that involve text and images.Read the paperGet the code & dataset

Rafah Hosn, partner group product manager for AI Frontiers at Microsoft Research, discusses the challenges and excitement of pushing AI boundaries. They talk about the importance of maintaining a culture of curiosity and skepticism, the need for rapid adaptation in AI research, and the significance of responsible innovation in AI development.

Tusher Chakraborty discusses Spectrumize paper addressing IoT connectivity challenges via satellite networks. Topics include leveraging small satellites for communication efficiency, democratizing global connectivity, satellite tone recognition using software, and challenges in satellite network research.

The new series “Ideas” debuts with guest Kalika Bali. The speech and language tech researcher talks sci-fi and its impact on her career, the design thinking philosophy behind her research, and the “outrageous idea” she had to work with low-resource languages. Learn more:The State and Fate of Linguistic Diversity and Inclusion in the NLP World | Publication, July 2020Project VeLLM | Project pageKahani: Visual Storytelling | Project pageKahani: Visual Storytelling through Culturally Nuanced Images | Microsoft Research Forum | Episode 1, January 2024Teachers in India help Microsoft Research design AI tool for creating great classroom content | Microsoft Research blog, October 2023Digital Labor: Project Karya | Project pageVillage by village, creating the building blocks for AI tools with work that also educates | Microsoft Source Asia blog, February 2024

AI scientist Ashley Llorens and Principal Researcher Ida Momennejad discuss the evolution of large-scale AI models and the potential for artificial general intelligence. They explore the concept of general intelligence in humans and machines, leveraging cognitive neuroscience to enhance AI capabilities. The conversation delves into creating prompts for evaluating language models, imitating human behavior with diffusion models, and the case for developing a 'prefrontal cortex' for AI.

Senior Researcher Chang Liu discusses M-OFDFT, a variation of orbital-free density functional theory (OFDFT) that leverages deep learning to help identify molecular properties in a way that minimizes the tradeoff between accuracy and efficiency. Read the paper

Can how we think about our thinking help us better incorporate generative AI in our lives & work? Explore metacognition’s potential to improve the tech’s usability on “Abstracts,” then sign up for Microsoft Research Forum for more on this & other AI work.Read the paper

Partner Research Manager and developer experience expert Nicole Forsgren talks about the future of software engineering with AI, why she loves tech, and her reliance on a spreadsheet and her gut when making career-changing decisions.Learn more:Nicole Forsgren at Microsoft ResearchNicole Forsgren websiteQuantifying the impact of developer experience | Microsoft Azure Blog, January 2024Yes, good DevEx increases productivity. Here is the data. | GitHub blog, January 2024Accelerate: The Science of Lean Software and DevOps: Building and Scaling High Performing Technology Organizations | Book, 2018

Partner Software Architect Ivan Tashev talks about applying his expertise in audio signal processing to the design and study of audio components for Microsoft products such as Kinect and shares how a focus on what he can control has fueled professional success.Learn more:Ivan Tashev at Microsoft ResearchDistributed Meetings: A Meeting Capture and Broadcasting System | Publication, December 2002Research Collection: The Unseen History of Audio and Acoustics Research at Microsoft | Microsoft Research blog, August 2020 

On “Abstracts,” Jordan Ash & Dipendra Misra discuss the parameter reduction method LASER. Tune in to learn how selective removal of stored data alone can boost LLM performance, then sign up for Microsoft Research Forum for more on LASER & related topics.Learn more:The Truth is in There: Improving Reasoning in Language Models with Layer-Selective Rank Reduction | Publication, December 2023LASER code on GitHub