According to TechCrunch, investors at their Disrupt event made it clear that artificial intelligence is the overwhelming focus for venture capital next year. VCs like Nina Achadjian from Index Ventures, Jerry Chen from Greylock, and Peter Deng from Felicis all emphasized the sector’s breakneck pace and unprecedented growth. Achadjian warned of “false positives” in product-market fit, where enterprise demand for the “latest and greatest” AI can generate revenue without delivering real ROI for customers. She stressed that assessing a founder’s resilience is now more critical than ever. Deng, a former OpenAI employee, argued that startups must find a unique “data flywheel” to stand out. Chen noted that chat apps, coding apps, and AI in customer service are currently working, but massive change is coming to nearly every industry.
The AI gold rush is on, but the pickaxes are different
Look, it’s not exactly a shock that VCs are all-in on AI. But here’s the thing: the sheer unanimity is staggering. It’s not just “a” hot sector anymore; it’s basically the only sector getting serious oxygen in the room. So what does that mean for founders? Basically, you’re now pitching into a deafening roar of a thousand other companies with similar slide decks about large language models. The bar has moved. It’s no longer about having an AI idea. It’s about proving you can survive the brutal commoditization that’s already happening.
Resilience and data moats are the new metrics
The most interesting shift is what these investors are actually looking for beneath the AI surface. Achadjian’s point about resilience is key. When she says they spend an “enormous” amount of time assessing it, she’s talking about betting on the jockey, not just the horse. Why? Because the tech is moving so fast that today’s breakthrough is next month’s open-source project. Deng’s comment on unique data flywheels is the technical counterpart to that. Anyone can build a wrapper around GPT-4. But can you create a system where your product’s usage generates proprietary data that makes the product itself smarter and harder to replicate? That’s a real moat. Without it, you’re just a feature waiting to be absorbed by the foundational model makers—a risk Achadjian explicitly tells founders to have a hypothesis for.
So what’s next beyond chatbots?
It’s telling that the current “what’s working” list—chat, coding, customer service—feels almost… obvious. They’re the low-hanging fruit. The real bets are on what comes after. Deng’s excitement about AI-enabled marketplaces is fascinating. Think about it: an marketplace that doesn’t just connect buyers and sellers, but uses AI to dynamically manage inventory, pricing, and trust in ways humans can’t. Achadjian’s call on robotics makes sense, too. AI gives robots the brains they’ve always lacked. And Chen’s eye on SaaS and untouched markets is the classic VC play: find the wave before it crests. But even the “non-AI” opportunity they mentioned—digitizing pen and paper processes in blue-collar industries—gets a caveat: AI will probably automate it anyway. There’s no escape!
The takeaway for anyone building in AI
The message from top-tier VCs is brutally clear. Your AI demo is table stakes. Now, you need to answer harder questions. How will you pivot when the core model you rely on changes overnight? What’s your unique, accumulating data asset that competitors can’t access? And can you articulate why a giant won’t just build your “product” as a checkbox feature next quarter? It’s a shift from funding innovation to funding defensibility in an era where innovation is being democratized at lightning speed. The irony? In a field obsessed with intelligent machines, investors are doubling down on their assessment of human qualities: resilience, expertise, and strategic honesty. Go figure.
