An in-depth exploration of the evolving landscape of generative AI
This in-depth examination offers a sophisticated grasp of the generative AI domain, offering a vivid picture of a developing world set for massive change.
The rise of generative artificial intelligence (AI) heralds a technological paradigm shift, with a rush of businesses racing to define the underlying stack, establish foundation models, and craft AI-native products. Unlike many exaggerated phenomena, the generative AI surge isn't just a fad—it's delivering meaningful gains in real businesses, exhibiting incredible user growth, and producing significant profits in a short period of time.
Nonetheless, a critical question arises in the midst of this surge: Where does the genuine value lay inside this booming market?
A year of immersion with founders and operators in this industry reveals fascinating dynamics. Surprisingly, while application firms enjoy strong revenue growth, infrastructure vendors appear to be the principal winners, taking a sizable share of the revenue flow. Meanwhile, despite being critical to the market's existence, the companies developing the generative AI models have not garnered equivalent commercial-scale rewards.
The pursuit of value marks a significant schism. While the early adopters of generative AI models aren't reaping a great deal of the benefit, anticipating the next evolutionary phase remains challenging. However, beyond traditional incumbents' moats, distinctiveness and defensibility across numerous stack layers will fundamentally affect market structure, margins, and retention—the basis of long-term value.
Here's a look of how the generative AI stack is progressing:
High-level tech stack: infrastructure, models, and apps
Applications:
A flood of generative AI apps on the market faces retention and differentiation challenges.
Profitable growth is dependent not just on outstanding user acquisition but also on sustained high gross margins and long-term customer retention, which challenges traditional thinking about developing long-term software enterprises.
Models:
Despite their pivotal importance, entities developing generative AI models have failed to scale up commercially.
Commoditization, the risk of clients switching to in-house AI, and the balance between chasing value and prioritizing public good in model provider missions are all issues that will be addressed in the future trajectory.
Infrastructure providers:
Infrastructure is a cornerstone, sucking in a sizable portion of generative AI revenue.
Leading cloud providers such as AWS, GCP, and Azure, as well as hardware giants such as Nvidia, dominate this space, however real competition is slowly emerging.
Where Does Value Accrue?
There are currently no systemic moats in the search of value in generative AI. Applications, models, cloud providers, and even hardware manufacturers face difficulties in creating long-term distinction, resulting in an environment devoid of apparent winner-take-all dynamics.
This ambiguity, however, holds promise. With a market potential covering all software and human pursuits, a slew of players and fierce competition across stack tiers is expected. The success of horizontal and vertical businesses will most likely be determined by end-market demands. The function of AI inside the end-product will determine whether vertical integration or horizontal diversification wins.
The rules are changing in real time in this fluid landscape. Generative AI has the potential to unlock significant value, altering the technology environment and ushering in a dynamic period of innovation and transformation.
Source:
Who Owns the Generative AI Platform?
Matt Bornstein, Guido Apprenzeller, and Martin Casado