Sarah Wang

Sarah Wang is a general partner at Andreessen Horowitz (a16z), where she leads growth-stage investments across AI, enterprise applications, and infrastructure. Her documented work includes joining Character.AI’s board in 2023, developing a joint World Labs investment rationale with Martin Casado, and announcing a16z’s investment in Mind Robotics in March 2026.[3][5][7][8]

Wang works with Casado across the boundary between venture and growth investing. In a Latent Space interview published February 19, 2026, she explained that frontier-model startups remain early bets on technical founders while requiring growth-scale resources from their inception. Compute agreements can involve equity, strategic partnerships, and go-to-market support—negotiations that young companies previously would not have needed at that scale.[1]

Her analysis describes a compute-to-product financing cycle: raise money for compute, achieve a capability breakthrough, put that breakthrough into a vertically integrated application, gain users, and raise again at peak momentum. Wang explicitly conditions the connection between funding and capability on scaling laws continuing to hold and capabilities advancing. She also identifies a competitive tension: model companies supply APIs to application developers while competing with them through their own applications.[1]

Wang’s case for backing multiple model companies rests on founders’ demonstrated prior work, the teams they assemble, and the possibility of valuable specialization. She argues that the market need not be a zero-sum contest won by one general model. In the February-published interview, she also remained optimistic about Thinking Machines Lab after being asked about its founder split, saying, “we’re more excited than ever about them.”[1]

Wang suggests that enterprise companies may find room to build where a model already performs a task well enough that further capability gains add little practical value. In those cases, she proposes, services and implementation can make the technology useful to customers. This is a tentative investment view about where value might accrue, rather than a claim that enterprise tasks have broadly stopped benefiting from better models.[1]

Her own example was a late-night Claude Cowork trial with two colleagues. Wang recalled giving it a raw file for data analysis and checking the resulting numbers successfully. She described the result as an “aha moment” for growth-investor work such as analyzing customer cohorts and retention—tasks she had previously done by hand. The reported accuracy concerned that trial.[1]

In the February-published interview, Wang expressed caution about robotics funding: she had not yet seen a ChatGPT-like breakthrough on the hardware side, while investment appeared to assume one. On March 11, 2026, she announced a16z’s investment in Mind Robotics, led by RJ Scaringe. Her rationale emphasized the industrial requirements of embodied AI: “Embodied AI is not just a modeling challenge, but an industrial one.”[1][8]

Wang argued for designing models trained on real-world data, hardware capable of fine manipulation, and deployment infrastructure together. She highlighted Mind Robotics’ partnership with Rivian as a manufacturing proving ground: deployment generates data, data improves models and control systems, and improved performance can support wider deployment. She credited Jamie Sullivan, David George, Raghu Raghuram, Erin Price-Wright, Jacob Zietek, and Guido Appenzeller with instrumental contributions to the deal.[8]

Wang’s earlier investments provide context for these later views on integrated AI systems. Her 2023 Character.AI rationale focused on owning both the model and customer relationship; her 2024 work with Casado examined the technical requirements of spatial intelligence.[6][7]

On March 23, 2023, Character.AI announced that a16z was leading its Series A and Wang was joining its board. In her same-day investment account, Wang recalled that her first trial of the product “completely hijacked my husband’s birthday dinner”: twelve friends shifted from conversation to exploring celebrity characters and creating their own. She described the investment as a shared conviction with Marc Andreessen.[5][6][3]

Wang favored Character.AI’s end-to-end structure, arguing that controlling its models and customer relationship allowed user engagement to feed back into model improvement. Noam Shazeer, Daniel De Freitas, and their team built the models and application. Looking back in the February-published 2026 interview, Wang used Character.AI to illustrate a resource dilemma: scarce GPUs must be divided among serving products, near-term research, and long-term AGI research, even as product revenue helps finance further compute.[6][5][1]

Wang and Casado coauthored a World Labs investment announcement on September 13, 2024. They framed generating interactive 3D worlds as a spatial-reasoning problem requiring combined expertise in data, graphics, and AI. Their rationale credited founders Fei-Fei Li, Justin Johnson, Ben Mildenhall, and Christoph Lassner with the relevant technical experience. The announcement specifically attributed hands-on help balancing research and product to Casado.[7]

Before these AI investments, Wang joined a16z in 2019 from TA Associates, where she was a vice president focused on growth-stage investing and served on the ZoomInfo/DiscoverOrg board. She also gained venture experience at DCM Ventures and Radicle Impact, and began her career at Boston Consulting Group and Morgan Stanley. David George announced her promotion to general partner on January 10, 2022, crediting her with leading a16z’s investments in Sourcegraph and Crossbeam.[3][4]

Wang holds an MBA from Stanford Graduate School of Business and an undergraduate degree from Harvard University.[3]