Google warned of risks for AI startups: some companies may not withstand the competition

Following the rapid increase in the number of new companies in the generative artificial intelligence market, some may face challenges. This was stated by Darren Mourey, who heads the global work with companies at Google (Cloud, DeepMind and Alphabet).

What is known

According to the vice president, there are two business models: companies that create products on top of ready large language models (LLM), as well as platforms that combine several models in one interface.

The first category consists of companies using models like GPT, Gemini, or Claude and adding their own interface or separate functionality. Mourey noted that if a startup relies mainly on the capabilities of the basic model without significant own technological component, it complicates further scaling.

The second type is platforms that combine several language models through a single API or interface. Such services route requests between different models and may offer monitoring or management tools. According to Mourey, the development of corporate solutions from model providers affects the positions of such companies.

He also drew a parallel with the initial stage of cloud services development in the late 2000s when companies appeared reselling AWS infrastructure. Later, after the expansion of tools by the providers themselves, some of these businesses lost market positions.

Source: TechCrunch