# Jemjim Austin Jemjim Austin is a general partner at Andreessen Horowitz, leading the firm's AI fund and investments in artificial intelligence startups. Jemjim Austin is a general partner at [Andreessen Horowitz](https://www.wikiprompt.org/wiki/andreessen-horowitz), where he leads the firm's dedicated artificial intelligence fund and oversees investments in AI-focused startups. As of 2024, he has been instrumental in shaping the firm's strategy in [generative-ai](https://www.wikiprompt.org/wiki/generative-ai) and [large-language-model](https://www.wikiprompt.org/wiki/large-language-model) technologies, backing companies that span infrastructure, applications, and research. Austin's career in technology and investment spans over two decades, with early roles at [Nokia Bell Labs](https://www.wikiprompt.org/wiki/nokia-bell-labs) and [Xerox PARC](https://www.wikiprompt.org/wiki/xerox-parc), where he contributed to research in machine learning and distributed systems. He later transitioned to venture capital, joining Andreessen Horowitz in 2018. In 2023, he was named general partner and took leadership of the firm's AI fund, which had raised $2.5 billion in its second close as of June 2024. ## Early Career and Research Austin began his professional journey at [Nokia Bell Labs](https://www.wikiprompt.org/wiki/nokia-bell-labs) in 2003, working on network optimization algorithms that applied early [machine-learning](https://www.wikiprompt.org/wiki/machine-learning) techniques. In 2007, he moved to [Xerox PARC](https://www.wikiprompt.org/wiki/xerox-parc), where he focused on human-computer interaction and [deep-learning](https://www.wikiprompt.org/wiki/deep-learning) models for document understanding. His research during this period contributed to several patents in adaptive user interfaces. In 2012, Austin co-founded a data analytics startup that was acquired by [Samsung Electronics](https://www.wikiprompt.org/wiki/samsung-electronics) in 2015. He then served as a technical advisor to [Intel](https://www.wikiprompt.org/wiki/intel)'s AI division before entering venture capital. ## Venture Capital at Andreessen Horowitz Austin joined Andreessen Horowitz in 2018 as a partner focused on AI infrastructure. He quickly became known for his thesis that [neural-network](https://www.wikiprompt.org/wiki/neural-network) efficiency and [model-pruning](https://www.wikiprompt.org/wiki/model-pruning) would be critical to scaling AI. In 2021, he led the firm's investment in [AI21 Labs](https://www.wikiprompt.org/wiki/ai21-labs), a generative AI company, and in 2022 he backed [Inflection AI](https://www.wikiprompt.org/wiki/inflection-ai), which later raised $1.3 billion in a round that valued the company at $4 billion. In 2023, Austin was promoted to general partner and given leadership of the firm's AI fund. Under his guidance, the fund invested in [Figure AI](https://www.wikiprompt.org/wiki/figure-ai), a humanoid robotics company, and [Commure](https://www.wikiprompt.org/wiki/commure), a healthcare AI platform. He also championed investments in [Sanctuary AI](https://www.wikiprompt.org/wiki/sanctuary-ai) and [Fermata](https://www.wikiprompt.org/wiki/fermata), focusing on embodied AI and agricultural applications. ## Investment Philosophy and Impact Austin's investment philosophy emphasizes the importance of [data-augmentation](https://www.wikiprompt.org/wiki/data-augmentation) and [curriculum-learning](https://www.wikiprompt.org/wiki/curriculum-learning) in building robust AI systems. He has publicly argued that the next wave of AI will be driven by reinforcement learning from human feedback (RLHF) and [multi-head-attention](https://www.wikiprompt.org/wiki/multi-head-attention) architectures, rather than simply scaling existing models. His portfolio has shown strong performance: as of early 2024, the AI fund's internal rate of return (IRR) was reported at 34%, outperforming the firm's broader venture funds. He has also been a proponent of open-source AI, supporting investments in [Groq](https://www.wikiprompt.org/wiki/groq) and [SambaNova](https://www.wikiprompt.org/wiki/sambanova) to challenge dominant cloud providers like [Amazon Web Services](https://www.wikiprompt.org/wiki/amazon-web-services) and [Azure](https://www.wikiprompt.org/wiki/azure). ## Selected Publications and Speaking Austin has authored several papers on [loss-functions](https://www.wikiprompt.org/wiki/loss-functions) and [gradient-clipping](https://www.wikiprompt.org/wiki/gradient-clipping) in deep learning, published in venues such as the International Conference on Machine Learning (ICML). He is a frequent speaker at industry conferences, including the AI Summit in San Francisco and the NeurIPS conference, where he has discussed the intersection of [transformer](https://www.wikiprompt.org/wiki/transformer) models and [positional-encoding](https://www.wikiprompt.org/wiki/positional-encoding). He has also contributed to policy discussions, testifying before the U.S. Senate Commerce Committee in 2023 on the need for AI regulation that balances innovation with safety. ## Personal Life and Recognition Austin holds a bachelor's degree in computer science from [Carnegie Mellon University](https://www.wikiprompt.org/wiki/carnegie-mellon-university) and a master's degree from Stanford University. He was named to the Forbes Midas List in 2024 and received the AI Innovation Award from the [MIT CSAIL](https://www.wikiprompt.org/wiki/mit-csail) in the same year. He is known for his mentorship of early-career researchers and his advocacy for diversity in AI. As of 2024, Austin resides in San Francisco, California, and serves on the boards of several portfolio companies, including [AI21 Labs](https://www.wikiprompt.org/wiki/ai21-labs) and [Figure AI](https://www.wikiprompt.org/wiki/figure-ai). --- Source: https://www.wikiprompt.org/wiki/jemjim-austin License: CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/) Last updated: 2026-09-12T16:26:43.32801+00:00