The rise of generative AI technology has led to a surge in startups, but as the initial excitement settles, two once-popular business models are facing challenges: LLM wrappers and AI aggregators.
According to Darren Mowry, who heads Google’s global startup division encompassing Cloud, DeepMind, and Alphabet, startups utilizing these models are showing signs of concern. LLM wrappers essentially involve startups wrapping existing large language models like Claude, GPT, or Gemini with a product or user experience layer to address specific issues. However, Mowry warns that relying solely on the backend model without significant differentiation may not be sustainable in the long run.
Mowry emphasizes the importance of building deep and wide moats to stand out in the market, whether through horizontal differentiation or catering to specific vertical markets. Examples of successful LLM wrapper startups with strong moats include Cursor, a coding assistant powered by GPT, and Harvey AI, a legal AI assistant.
In contrast, AI aggregators, which aggregate multiple LLMs into one interface or API layer to facilitate access to various models, are facing challenges. Mowry advises incoming startups to steer clear of the aggregator business, as users now seek intellectual property integrated into the solutions they use, rather than just access to multiple models.
Drawing parallels to the early days of cloud computing, Mowry notes how startups reselling AWS infrastructure faced difficulties once Amazon introduced its enterprise tools. Similarly, AI aggregators may face pressure as model providers expand into enterprise features, potentially marginalizing middlemen.
On a positive note, Mowry expresses optimism for vibe coding and developer platforms, highlighting the success of startups like Replit, Lovable, and Cursor. He also sees growth potential in direct-to-consumer tech, where companies provide powerful AI tools directly to customers, such as Google’s AI video generator Veo for film and TV students.
Looking beyond AI, Mowry identifies biotech and climate tech as burgeoning industries, attracting significant venture investment and leveraging vast amounts of data to create value in innovative ways. Overall, while the landscape for AI startups may be evolving, the potential for growth and innovation remains promising.
