This Is How The Gen AI Crash Will Unfold
First, VCs will slow the rate of investment in Gen AI companies. This has already happened as Gen AI investment peaked in Q3 2023. Early stage venture funding is down 76% from the Q3 2023 peak.
Second, end users will not pay for premium Gen AI products and features at a level that justifies the hundreds of billions of investment Dollars that have been made in the Gen AI space, (chip manufacturing, LLM and multimodal model construction, and infrastructure and application software).
Google, Microsoft, AWS, Oracle and Meta will continue to include Gen AI services in their AI services portfolio. Google and Meta of course largely build their own LLMs while Microsoft leans heavily on OpenAI. Oracle uses Cohere’s LLMs while AWS leans heavily on Anthropic, although AWS will build more of its own LLMs.
OpenAI will continue to build LLMs, multimodal models and related services, but likely will see its ability to raise money scale back as valuations will not be as rich and as end-user demand wanes. Ditto for Anthropic, Mistral and the rest of the LLM cohort.
NVIDIA and other AI chip suppliers will see demand slow as fewer end-users will experiment with Gen AI. It is not as though end-users are inventing new use cases by the day. There are only so many chatbots / AI agents that one can build.