What’s next in AI? Sonya Huang and Sarah Guo share what founders need to know

At Mercury Spheres, early-stage founders gathered for a day of hard-earned insights — from brand-building to hiring to navigating the AI era. One of the standout sessions featured Sonya Huang of Sequoia and Sarah Guo of Conviction, moderated by our CEO, Immad Akhund.

Both Huang and Guo are leading voices in AI investing — backing companies such as OpenAI, Runway, LangChain, Harvey, and others. But what made this session particularly valuable was their practical advice for founders trying to navigate today’s shifting AI landscape, whether or not their business is “AI-first.”

We’ve moved from copilots (auto-complete, code suggestions) to autopilots (agents that execute tasks), and now, we’re entering the era of air traffic control.

According to Huang, the most ambitious startups today are moving beyond building assistive tools. They're orchestrating networks of AI agents capable of running workflows end-to-end. This shift is visible in vertical applications, from legal to logistics.

What this means for founders

Your role is evolving. The scarce skill isn’t just engineering — it’s high-leverage creativity, coordination, and judgment across increasingly autonomous systems. Founders who embrace this new kind of leadership may find themselves well-positioned to outpace competitors still stuck in manual mode.

Two years ago, it seemed like the safest bet was building on top of foundation models. But now, Huang and Guo argue that the vertical path — i.e., deep expertise in specific industries — is where startups can truly differentiate.

Whether it’s legal AI (Harvey), medical research tools (Open Evidence), or logistics automation (Sola), startups that apply AI to domain-specific problems are seeing real traction, especially when paired with intuitive user experiences for nontechnical teams.

What this means for founders

Don’t AI-wash your company. If your value lies in operational insight or customer intimacy, lean into that. You don’t need a team of AI researchers: Many top startups today are building impressive systems by combining strong product instincts with public AI models and APIs.

Founders often feel pressure to add an “AI slide” to their pitch decks. But as Sarah Guo reminded the audience: VCs can see through it. If your product is solving a real problem, with happy customers and growing usage, that’s what matters most.

That said, the bar is undeniably higher. AI-native startups are reaching milestones — like $1M to $10M revenue runs — in record time. That acceleration shifts expectations, even for non-AI businesses.

What this means for founders

Investors will always care most about traction, differentiation, and clarity of vision. If you’re not building an AI product, you can still fundraise effectively, but you may need sharper storytelling and a clearer “why” now. And if you are building in AI, know that access to foundational models has lowered the technical bar, but raised the competitive one.

AI doesn’t just reshape products — it reshapes companies. Guo noted that in today’s landscape, small, agile teams can now compete at the scale of legacy enterprises. It’s no longer about how many people you hire, but how well your team uses AI tools to amplify output.

Huang echoed this, pointing out that startups operating “AI-native” from day one are already pulling ahead, and that traditional orgs may face a cultural reckoning if they don’t adapt quickly.

What this means for founders

Being small is a superpower, especially if you embed AI into the way your team works, from engineering to ops to customer support. The earlier you set those habits, the greater your long-term leverage.

One of the most unscripted and inspiring moments of the session came during audience Q&A, when a founder from a nonprofit asked how they could use AI to help low-income students in Southern California.

Huang and Guo lit up. Their advice: nonprofits can use AI in two ways — to scale their mission (like delivering tutoring through AI-native tools) and to streamline operations, freeing up resources for impact. AI, they argued, is a “great democratizer,” and organizations helping underserved groups adopt it will play a vital role in closing access gaps.

Want more investor perspectives? Watch our Expert Session with Sequoia (password: YNAE@D9y) for more candid thoughts on startup positioning best practices and the fundraising process.