The Intersection of Generative AI and the Power of Reading & Writing

The Intersection of Generative AI and the Power of Reading & Writing

Introduction:
In the rapidly evolving field of generative AI, various players are vying for dominance. From infrastructure vendors to application companies and model providers, each entity has its own challenges and opportunities. At the same time, the importance of reading and writing in fostering a prepared mind cannot be underestimated. This article explores the commonalities between the generative AI platform and the power of reading and writing, shedding light on potential strategies for success.

The Growth of Generative AI Applications:
Generative AI applications have witnessed remarkable growth, driven by novelty and diverse use cases. Notably, image generation, copywriting, and code writing have already surpassed $100 million in annualized revenue. However, application companies often face difficulties in retention, product differentiation, and gross margins. To build long-term customer value, B2B and B2C apps leverage network effects, data ownership, and intricate workflows. While selling end-user apps is a common approach, it may not be the sole path to establishing a sustainable generative AI business.

The Role of Model Providers:
Model providers, crucial for the existence of the generative AI market, have yet to achieve significant commercial scale. However, the commercialization of model providers is closely linked to hosting. The demand for proprietary APIs, such as those offered by OpenAI, is rapidly increasing. Hosting services for open-source models, like Hugging Face and Replicate, facilitate seamless sharing and integration of models. This highlights the importance of hosting as a key factor in the success of model providers.

The Public Good and Value Capture:
Given the immense potential and potential harm of generative AI, many model providers have embraced the concept of public benefit corporations (B Corps) and incorporated social and environmental impact into their mission. Interestingly, this approach has not hindered their fundraising efforts. However, a pertinent discussion revolves around whether most model providers aim to capture value and if they should prioritize it.

The Dominance of Infrastructure Companies:
Infrastructure companies play a significant role in the generative AI market, capturing a substantial share of the revenue flowing through the stack. On average, app companies spend a significant portion of their revenue on inference and fine-tuning, either directly to cloud providers or third-party model providers. This suggests that a considerable portion (10-20%) of total generative AI revenue currently goes to infrastructure companies. Notably, Nvidia stands out as a major winner, with its data center GPU revenue reaching $3.8 billion.

Defensibility and Moats:
Within the generative AI landscape, infrastructure companies possess lucrative and durable positions. Their scale, supply-chain, ecosystem, algorithmic, distribution, and data pipeline moats contribute to their defensibility. However, the long-term sustainability of these moats remains uncertain. The presence of strong, direct network effects across different layers of the stack is yet to be determined.

The Future Landscape:
The generative AI market is unlikely to exhibit a winner-take-all dynamic. Both horizontal and vertical companies are poised to succeed, with the best approach determined by end-markets and end-users. Verticalization, where user-facing apps are tightly coupled with home-grown models, may prevail when the AI itself is the primary differentiator. On the other hand, horizontalization may be more prominent when AI is part of a broader feature set.

The Power of Reading and Writing:
Reading and writing have always been integral to human existence. While Warren Buffett's reading habits are often highlighted, networking, building relationships, and traveling have equally contributed to his success. The industrial revolution has significantly expanded literacy rates worldwide, altering the volume and mediums of information. Reading serves as the foundation for developing a prepared mind, enabling individuals to engage in more informed conversations.

The Importance of Writing:
Writing complements reading, allowing individuals to articulate their thoughts and bring ideas to life. Writing becomes a medium through which one's perspectives are shared and refined. By putting thoughts into words, writers gain a deeper understanding of their own thinking process. As Flannery O'Connor famously said, "I write because I don't know what I think until I read what I say."

Actionable Advice:

1. Embrace a holistic approach: Building a successful generative AI business requires considering various factors beyond just selling end-user apps. Explore different avenues, such as hosting services and proprietary APIs, to capture value and foster growth.

2. Focus on differentiation: In an industry with limited technical differentiation, vertical integration can provide a competitive edge. By tightly coupling user-facing apps with home-grown models, companies can create unique offerings that stand out in the market.

3. Prioritize reading and writing: Invest time in reading to build a prepared mind, enabling you to enter conversations with confidence and expertise. Additionally, embrace writing as a tool to articulate ideas and gain a deeper understanding of your own thoughts.

Conclusion:
The generative AI platform offers immense possibilities, with infrastructure companies currently reaping significant rewards. However, the landscape remains dynamic, and various players have the potential to succeed. By embracing a holistic approach, focusing on differentiation, and harnessing the power of reading and writing, businesses can navigate this evolving field and position themselves for long-term success.