The report by Sarah Wang and Shangda Xu from Andreessen Horowitz highlights the rapidly evolving landscape of generative AI (genAI) in the enterprise sector. In 2023, generative AI made significant inroads into the consumer market, but it’s in 2024 that the real potential for revenue multiplication exists within enterprises. Enterprises are moving beyond the initial narrow scope of use cases and “GPT-wrapper” products, significantly increasing their budgets for genAI applications, which has resulted in a shift towards more varied and production-level use cases utilizing smaller, open-source models. The report emphasizes a dramatic increase in enterprise budgets for genAI, with companies planning to invest 2x to 5x more in 2024 compared to 2023, pushing the average spend to around $7 million. This surge in investment is motivated by the promising results from early genAI experiments and a strategic shift towards incorporating these technologies into core business functions.
Furthermore, the landscape is transitioning towards a multi-model, open-source world, indicating a preference for diverse solutions that avoid vendor lock-in and provide flexibility. Open-source models, in particular, are gaining traction for their cost-effectiveness and the control they offer over data security and customization. Enterprises are optimizing their applications to switch between models seamlessly, demonstrating a preference for building rather than buying applications to maintain strategic control and adapt quickly to new advancements. Overall, the enterprise genAI sector is poised for significant growth, with total spending on model APIs and fine-tuning expected to reach a $5 billion run-rate by the end of 2024, driven by increased budgets, diversified use cases, and a strategic pivot towards genAI across a range of industries.