AI has revolutionized how we interact with technology, with startups leveraging AI to create systems of intelligence that bridge multiple systems of record and provide intelligent recommendations and decisions based on data.
The rise of AI has led to the development of a new AI application stack, connecting systems of engagement with systems of record and enterprise data, presenting exciting opportunities for startups to leverage open source models and build niche applications.
Deep dives
The Power of AI in Building Systems of Intelligence
AI has evolved significantly in the past few years, transforming how we interact with technology. From Siri to sophisticated language models like GPT and BART, AI is being used to change how we engage with applications and build powerful problem-solving systems. It has become essential for businesses, permeating enterprise software, consumer experiences, and various industries like healthcare and self-driving cars. Startups are leveraging AI to create systems of intelligence that bridge multiple systems of record and provide intelligent recommendations and decisions based on data. The key component that has supercharged systems of intelligence is the development of large language models, offering innovative ways to solve complex problems.
The Emergence of a New AI Application Stack
With the rise of AI, a new AI application stack is being built to connect systems of engagement, such as chatbots and mobile apps, with systems of record and enterprise data. Startups are developing frameworks like LAMA Index to bridge personal and enterprise data with foundation models. These large language models are being used to interact with data, create intelligent workflows, and deliver information to users when and where they need it. The stack is still evolving, but it presents exciting opportunities for startups to leverage open source models, build AI infrastructure, and create applications that utilize the power of AI.
Navigating the Landscape and Future of AI
While there are risks and challenges for startups entering the AI space, such as rapidly evolving technologies and uncertainties around monetizing foundation models, there is still potential for startups to succeed. Open source AI has leveled the playing field, allowing startups to compete with big cloud providers and build niche applications. The question of where value accrues in the AI stack remains open, whether it's at the foundation model layer, infrastructure layer, or application layer. Ultimately, building a sustainable business model in AI requires focusing on fundamentals like scale, network effects, user experience, and market advantages, while AI becomes an ubiquitous ingredient in the technology landscape.
Greylock general partner Jerry Chen discusses key takeaways from his essay "The New New Moats: Why Systems of Intelligence are Still the Next Defensible Business Model." The essay is a fresh take on his "New Moats" blog from 2017, in which he postulated that startups would be able to build defensible moats using AI. With the explosion of AI activity in recent times, Chen revisited this analysis to see what holds true, what he got wrong, and what is still too early to tell.
You can read the 2023 essay here: https://greylock.com/greymatter/the-new-new-moats/
You can read the original 2017 essay here: https://greylock.com/greymatter/the-new-moats/