📰 In 2024, YC has already invested in 630 companies: Winter 199, Spring 196, and Summer 235. What really stands out isn’t the number—it’s that more than 90% of them don’t do To C, and over half are focused on building AI pipelines.
🔥 The most crowded areas are Agent infrastructure and AI employees. The former is busy provisioning agents with email, phone numbers, 2FA, payments, memory, and debugging. The latter is more direct—taking over specific roles like accounting, legal affairs, government relations, and restaurant back-of-house operations. The product pages don’t really describe features; they list job titles.
💡 The direction of money is also shifting. From Spring to Summer, the share for AI infrastructure rose from 8% to 20%, while AI application products fell from 55% to 39%. To be honest, the change is pretty straightforward: once people realize models can actually do things, they all have to rebuild everything—identity, payments, interfaces, compute, and ongoing operations and maintenance.
👀 In the Summer batch, B2B makes up 52.3%, industrial 23%, and the share truly targeting individual consumers is only 5.5%. Education projects are just 1 company. But YC has also publicly solicited AI consumer products for the general public and AI tutors for children—so what people want to invest in and what actually gets invested in clearly don’t fully line up.
🤔 If you’re building an AI startup now, would you squeeze into Agent infrastructure and AI employees—or go for the areas with fewer players in consumer and education?
#YC #人工智能 #AI创业 #Agent