Stanford CS 329M Machine Programming Guest Lecturers

Abstract:

Humanity's review systems are being overwhelmed (from code review to judicial review). As AI systems get more capable, this problem of review only gets worse. Frontier labs look to "alignment" as a solvable problem that sidesteps the review problem, but brings political and ethical problems. The alternative to alignment is scaling human review by generalizing principles from Formal Verification. This talk will briefly expand on the problem of review, describe specification-based AI as a possible solution for addressing code review, generalize spec-based AI to other generative AI domains, and contextualize this work as part of a broader research agenda.

Speaker Bio:

Evan leads Atlas Computing, a nonprofit mapping and prototyping ways to scale human review and provable safety of advanced AI. He previously built and led a venture studio designing and deploying new coordination systems for humanity, as well as the building the research grants and metascience team at Protocol Labs (the company that initially created IPFS and Filecoin). He completed a PhD in Applied Physics at Caltech, developing hardware for a secure quantum internet, and a BS in Materials Engineering from Stanford.

Abstract:

Humans describe and engineer the world in terms of "governing equations", manifest in software in the form of physics solvers for computational fluid dynamics et al. But these descriptions and computations of reality are largely incomplete: they often fail to describe common real-world physics situations, let alone coordinate multiple physics that may vary over spatial & temporal scales. Now consider differentiable programming, arguably the most powerful concept in deep learning: parameterized software modules that can be trained with some form of gradient-based optimization. What if we could build physics programs that are end-to-end differentiable and thus machine-learnable? And what if we could build such a software system with abstractions for domain experts without ML/AI knowledge? These two questions are explored in this talk, with demonstrations from Pasteur Labs software that is designed to become "the IDE for reality."

Speaker Bio:

Alexander Lavin (lavin.io) is a leading expert in AI-for-science and probabilistic computing. He's Founder & CEO of Pasteur Labs (and non-profit "sister" Institute for Simulation Intelligence; simulation.science), reshaping R&D with a new class of AI-native simulators, commercializing in energy security, aerospace, materials & manufacturing sectors (@simai4science). For the last dozen years, Lavin has focused on artificial general intelligence (AGI) research with top startups in neuroscience and robotics (Vicarious, Numenta), and sold his prior ML-simulation startup Latent Sciences to undisclosed pharmaco in neurodegeneration R&D. Lavin also serves as AI Advisor for NASA, overseeing physics-ML efforts for the NASA & ESA "Digital Twin Earth" projects. Previously, Lavin was a spacecraft engineer with NASA and Blue Origin, and won several international awards for work in rocket science and space robotics (including Google Lunar XPrize during graduate studies at Carnegie Mellon). Lavin was named Forbes 30 Under 30 in Science, and a Patrick J. McGovern Tech for Humanity Changemaker.

Abstract:

What if we've been solving the wrong equation? While everyone's busy implementing ML/AI, let's talk about how we use it to transform the open source ecosystem. Drawing from real-world examples across the CNCF landscape, we'll decode how ML is revolutionizing our code and our entire approach to open source collaboration.

Pattern Recognition:

• How ML helps us understand contribution patterns and community health

• Where AI assists in code review, documentation, and issue triage

• Why semantic understanding matters more than synthetic output

Critical Variables:

• The human factors that no model can replace

• Why intention and critical thinking remain our most valuable constants

• Real workflows where AI amplifies (rather than replaces) human insight

Practical Functions:

• Tools and approaches that work (and why they work)

• Patterns for integrating AI into open-source workflows

• Methods for maintaining human judgment in automated processes

Join me for an honest exploration of how ML/AI serves the open source ecosystem - not the other way around. Perfect for anyone interested in the real-world intersection of ML and open source, where success depends more on asking the right questions than having all the answers.

Speaker Bio:

Taylor Dolezal is the Head of Ecosystem at the Cloud Native Computing Foundation, where he steers initiatives and collaboration across the cloud native community. Based in Los Angeles, Taylor combines a love of tech and psychology with a keen focus on fostering innovation in the open source landscape. A long-time advocate of cloud native technologies, Taylor brings both strategic insight and practical knowledge to his work, helping drive the community forward while making it more accessible to all.

Abstract:

This Ask Me Anything session provides an opportunity to discuss Corporate Venture Capital. We will share ideas, brainstorm, and, most importantly, serve as an open forum to ask questions. This AMA will be interactive to help students understand the VC strategies and market trends.

Speaker Bio:

Emily Fontaine is the IBM Global Head of Venture Capital. As a business leader, she is a thought pioneer, value creator, and trusted advisor.

Over a long and distinguished career at IBM, Emily has led bold transformations and driven exceptional client delivery. She has extensive experience in operations and management. Emily’s success is a testament to her leadership and strategic acumen. Prior to taking on the Global Head of Venture Capital role, she served in the office of IBM's Chairman and CEO, following her role as AI Federal Leader for IBM Consulting.

As Global Head of Venture Capital, Emily leads IBM Ventures – including IBM’s $500 million Enterprise AI venture capital fund. IBM Ventures invests in a range of AI companies - from early-stage to hyper-growth startups - focused on accelerating generative AI technology and research for the enterprise. Since inception, IBM has invested in multiple companies across AI, data, cybersecurity, quantum computing, and sustainability – including Hugging Face and HiddenLayer. 

Emily holds a bachelor’s degree from Mount Holyoke College and an MBA in Business Administration from George Washington University. Emily is based in New York City.