The GenAI Application Layer

The GenAI Application Layer

September 18, 2024

As the Head of Generative AI Product at Canva, I feel incredibly privileged to have a front-row seat to the AI revolution. Every day, I witness how AI is reshaping the creative landscape, empowering millions of users to design with unprecedented ease and sophistication. This experience has not only shaped my perspective on the future of AI but has also informed my point of view on the biggest opportunities for early stage startup founders in 2024 and beyond.

The AI Ecosystem: A Layered Approach

To understand where the most exciting opportunities lie, let's break down the AI ecosystem into its key layers:

  1. Semiconductors: The foundation, where companies like NVIDIA build the GPUs and specialized chips that fuel AI computations. Generally sold on a unit basis.
  2. Infrastructure: Giants like AWS, Google Cloud, and Microsoft Azure that provide the computational muscle needed for AI at scale. Generally sold on a usage basis.
  3. Modeling Layer: Where organizations create advanced foundation models, the "brains" that power AI applications. Includes OpenAI, Anthropic as well as Google’s Gemini and AWS’ Bedrock. Generally sold on a usage basis (usually with integrated compute)
  4. Application Layer: Where AI technology is woven into specific solutions and products. This is where Canva operates, integrating AI into our design platform to create magical experiences for our users. Priced based on the value provided to end users.
image

Today, the lion’s share of revenues and profits have accrued at the Semiconductor and Infrastructure levels as the blog post The Economics of Generative AI does a great job at outlining the value chain of AI today and how he sees this evolving in the future.

However, if history is indication over the next decade the vast majority of economic opportunity lies at the application level.

image

AI-Native Applications: The New Frontier

The real game-changers will be AI-native applications—products built from the ground up with AI at their core. Let me share a few examples that have caught my eye:

  • GitHub Copilot: This AI coding assistant is responsible for over 40% of Microsoft's incremental revenue growth, generating hundreds of millions in revenue.
  • Jasper: This AI writing platform reached a $1.5 billion valuation in just 18 months. Jasper uses advanced language models to help users generate high-quality, original content across various formats, learning and improving from each interaction.
  • Harvey: Making waves in the legal industry, Harvey is an AI-powered legal research and analysis tool. By focusing on a specific industry and tailoring its AI capabilities to the unique needs of legal professionals, Harvey exemplifies how AI applications can create significant value in niche markets.

Why I'm Betting on the Application Layer

As an investor and advisor, I'm particularly excited about companies operating in the application layer of generative AI. Here's why:

  1. Model Agnosticism: They can leverage improvements across various model providers, ensuring their value proposition evolves with advancements in underlying technology.
  2. Value Beyond the Model: These aren't just AI interfaces; they're comprehensive solutions solving real-world problems.
  3. Data Advantage: Every interaction generates valuable data, creating a virtuous cycle of improvement, user growth and model quality improvements.
  4. Specialized Solutions: While general-purpose AI is powerful, the real value lies in tailored solutions for specific industries or use cases.

Having seen the transformative power of AI applications firsthand at Canva, I'm excited to support the next wave of innovators in this space. I'm particularly interested in startups that are building AI-native applications that solve specific, high-value problems and transforming traditional industries through innovative AI applications.

The Road Ahead

We're at the beginning of a new era in technology. The companies that can effectively harness AI to solve specific, high-value problems will be the ones that shape our future. They'll not only generate substantial returns but also push the boundaries of what's possible with AI.

The question isn't whether this shift will happen, but who will lead it. Will it be established tech giants pivoting to embrace AI, or scrappy startups unencumbered by legacy systems? Will it be in fields we expect, like software development and creative work, or will it revolutionize industries we haven't even considered?

One thing is certain: the AI revolution is here, and its impact will be felt far beyond the tech world. As we stand on this frontier, the opportunities are boundless.