Earlier this month, I attended the HumanX AI conference in Las Vegas, joining about 3,000 other attendees. I also had the privilege of moderating a panel with early-stage investors from three well-known venture capital firms that are heavily investing in artificial intelligence: Kleiner Perkins, Bain Capital Ventures, and Primary Venture Partners.

For these firms, AI-related investments made up an impressive 80% to 90% of their new portfolio deals in 2024. The discussion revolved around how AI is reshaping early-stage investments and the unique risks and rewards of backing startups in this growing space. Below are key takeaways from my conversation with Leigh Marie Braswell of Kleiner Perkins, Aaref Hilaly of Bain Capital Ventures, and Cassie Young of Primary Venture Partners.

The AI Revolution
AI’s impact has been “less predictable and straightforward” than previous technological shifts, according to Braswell from Kleiner Perkins. However, the demand for AI solutions across various industries and the real proof of a significant shift in technology is “unprecedented.”

Braswell highlighted several startups benefiting from AI, including Codeium, an AI coding startup, Glean, an AI-powered corporate search engine, and Ambience Healthcare, an AI assistant for healthcare professionals. These companies are tapping into AI’s potential in areas like code-generation, customer support, sales, and medical documentation.

Finding Non-Obvious Opportunities
Hilaly from Bain Capital Ventures emphasized that there are still many untapped AI applications. As AI models advance, particularly in reasoning and automation, new possibilities continue to emerge. Bain Capital Ventures has backed companies like MagicSchool AI, Decagon, and Moveworks, all of which have leveraged AI to unlock fresh opportunities.

The Importance of Speed and Market Fit
In today’s competitive startup environment, Braswell often asks founders, “What is your head start?” Startups need to stay ahead of the competition, even as it’s assumed that their competitors are equally capable. Founders must be able to ask themselves, “How far can I get before competitors can catch up?”

Hilaly added that there are no lasting barriers to entry at the early stage. “There are no moats, no defensibility,” he said. In such a competitive market, the key to success is bringing products to market quickly and consistently.

Skepticism Around Early Traction
While early traction can be a good indicator, Young from Primary Venture Partners cautioned that it can sometimes be misleading. She cited a recent study showing that up to 60% of generative AI purchases are driven by innovation and experimental budgets, making it harder to predict sustainable success.

Young described a term she uses for AI founders: “next-level missionaries.” These founders are incredibly customer-focused and determined to succeed in crowded markets.

AI Startups and Capital Efficiency
According to Hilaly, AI startups often require less capital because they can achieve more with smaller teams. However, this also makes the investment landscape more challenging, as many companies are able to grow revenue more quickly early on.

Are AI Agents Taking Over Jobs?
Throughout the conference, the theme of AI agents and automation was frequently discussed. While there’s hype around AI agents replacing jobs, Braswell emphasized that agents aren’t designed to automate every single task. AI tools excel in specific areas, such as building workflows or streamlining sales processes.

Eventually, AI models will become sophisticated enough to predict and optimize workflows so well that users may not need to switch between different applications. According to Hilaly, when AI models reach this level of proficiency, software applications themselves may function more like agents, performing tasks without user input.