Signals and Threads

John Crepezzi is a software engineer who has built developer tools at Tumblr, Square, GitHub, and Jane Street. He discusses how ChatGPT redirected an OCaml search project into a firmwide AI platform, the security challenges of agentic coding, and why code review must evolve as prototypes become abundant. They also explore sandboxes, human code understanding, and the cultural future of programming.

Aaron Bauer, a Jane Street software engineer and developer educator, explores teaching software engineering in the age of AI. He discusses pandemic classrooms, live lectures, active learning, and cognitive load. The conversation also examines editors as engineers’ home base, interactive AI workflows, and how Jane Street redesigned onboarding and internships around design, testing, code review, and judgment.

Eric Mannes, a Jane Street commodities trader and technology leader, explores the messy path from external information to tradable data. He revisits the oil future that settled below zero, commodities risk managed in one giant Excel spreadsheet, and the challenge of defining a company. They also discuss satellite imagery, credit-card data, data provenance, and why stronger ML makes rigorous engineering even more crucial.

Jacob Baskin, a software engineer who built Jane Street’s Superstore and worked on Google’s Ad Exchange, discusses building a distributed columnar warehouse from scratch. He talks about real-time writes with analytic queries, columnar storage trade-offs, metadata and sequencing decisions, and treating cluster scheduling as an auction-like problem. Short, technical, and full of system design stories.

Nate Foster, a professor blending programming languages with networking, discusses treating networks like programs. He explores NetKAT, P4 and programmable switch pipelines. They cover in-network computing tradeoffs, why multicast faded yet still matters in trading, and how high-level policy languages like Butane compile to BGP for safer, testable routing.

Will Wilson, founder and CEO of Antithesis and former mathematician, explains deterministic simulation and property-based testing. He discusses fuzzing, chaos testing, making systems deterministic, and turning observability into testable properties. The conversation covers exploring state spaces, replaying bugs, and why blending many techniques finds the subtle, show-stopping failures.

Chris Lattner, creator of LLVM and a key figure in developing Swift at Apple, dives into the exciting world of Mojo, a new programming language focused on enhancing GPU performance. He discusses the challenges of existing programming languages like C++ and how Mojo aims to revolutionize AI coding with metaprogramming and a unified type system. Lattner emphasizes the importance of making programming manageable while aiding developers in engaging with hardware complexities. His vision seeks to democratize AI compute and break the cycle of vendor control in technology.

Daniel Pontecorvo leads the physical engineering team at Jane Street, fusing architecture and engineering to create functional spaces. He dives into the challenges of cooling densely packed data centers and how it parallels keeping traders comfortable. They explore innovative cooling technologies, including immersion cooling, and the implications of machine learning on infrastructure. Pontecorvo also emphasizes the importance of robust monitoring systems and discusses the benefits of a modular workspace design for collaboration in trading environments.

Ian Henry, a software engineer at Jane Street with a background at Trello and Warby Parker, dives into the unique world of trading tools. He shares insights on the challenges of transitioning from consumer apps to specialized trading software. Ian discusses the innovative use of functional programming, notably OCaml, and explores whimsical projects like Bauble for rendering 3D shapes. He also addresses the complexities of alert systems, strategies for balancing trading configurations, and the advantages of using Lisp variants like Janet in programming, all while emphasizing creativity and user experience.

In Young Cho, Co-founder and CEO of DeepMind, discusses her unexpected transition from medicine to finance through a trading internship. She delves into the integration of trading, research, and software engineering. The conversation highlights the challenges of machine learning in chaotic environments and the evolution from basic models to deep neural networks. Cho also addresses the importance of robust data practices and explores the complexities of using Python notebooks in financial research.