2025 - the year of AI Agents - and investing in Agentic AI startups

Y Combinator is going all-in on AI agents, making up nearly 50% of the latest batch.

Y Combinator's Spring 2025 batch saw a significant focus on AI agent companies. Out of 144 startups in the batch, 67, or 46%, were identified as AI agent companies.

To myself and most who follow YC, this was not surprising and, in fact, expected.

YC has been, by far, the hottest startup accelerator for many years now, and they receive thousands of applications for each batch, allowing them to be highly selective.

Their MO is to go all-in on the hottest new trends by investing in as many of the promising new founding teams in those fields all at once through the accelerator batch programs, which this year increased to 4 batches annually, once for each season, vs. just 2.

They even go a step further and specifically, go out of their way to publish a "Request for Startups" for each batch, and for the spring 2025 batch specifically posted: "2025 is indeed shaping up to be the year of AI agents. We've put together a list of AI Agent startup ideas that we think are especially promising. Some of these draw attention to trends that are already in full swing, and some of them are where we think things are going next."

(see https://www.ycombinator.com/rfs for the full post by YC)

These AI agent startups are exploring diverse applications across various industries, including cybersecurity, consumer fintech, and healthcare.

As of this writing, I am personally excited about the coming era of AI agents for every aspect of business and our personal lives, spending a lot of my reflection time contemplating what life will look like once these become ubiquitous and integrated into every aspect of our lives.

I genuinely see the next generation growing up with AI agents woven into every aspect of life, and they would not be able to imagine life before or without them. This is definitely a major paradigm shift.

YC, with its batch model, can invest in a broad base of the most promising startup teams all at once, betting that a few will wind up being the winners who have breakout success and go on to be the next Google or Amazon of the coming generation of tech giants.

For VCs, such as myself, managing https://emerging.vc, we don't have that luxury and, therefore, face a dilemma similar to VCs investing in Internet startups in the late '90s.

For that reason, personally, I remain very curious and am carefully watching the progress of these startups and their products, yet I am mostly holding off on investing, as there are way too many, and it appears to be too early. The technology is evolving so rapidly, and none seem to have a particularly strong moat, so I cannot confidently pick winners; therefore, I will abstain from the category for the short term.