VCs are betting billions on AI’s next wave, so…
The $26 Million Question Nobody Expected
I still can’t wrap my head around what happened in Kentucky last week. An 82-year-old woman named Ida Huddleston turned down $26 million. Let me say that again — twenty-six million dollars. An AI company wanted to build a data center on her family’s 1,200-acre farm, and she looked them dead in the eye and said no.
“They call us old stupid farmers, you know, but we’re not,” she told reporters. “We know whenever our food is disappearing, our lands are disappearing, and we don’t have any water — and that poison. Well, we’ve had it.”
Her words hit me harder than any earnings report ever could. Here’s this woman who has probably never written a line of code, never pitched to a VC, never used ChatGPT — and she’s seeing something the rest of us are missing. While Silicon Valley celebrates every new billion-dollar funding round, Ida Huddleston is worried about water shortages and poisoned ground. And honestly? She’s not wrong.
The Money Keeps Flowing (But Where’s It Going?)
Let me give you the numbers because they’re honestly staggering. Kleiner Perkins just raised $3.5 billion. Not $350 million — $3.5 billion. That’s 75% more than they raised just two years ago. Thrive Capital locked down $10 billion. Founders Fund closed $6 billion. General Catalyst is reportedly chasing another $10 billion.
We’re talking about more money than the GDP of some countries being poured into AI startups. And here’s what gets me — these aren’t random bets. Kleiner Perkins has already scored big with Figma’s IPO and Google’s acqui-hire of Windsurf. They know what they’re doing.
I keep wondering what it’s like inside those pitch meetings. Do the founders walk in with a slide deck showing how they’ll shake up some industry? Do they demo a product that actually works, or just promise it’ll work someday? And how do partners at firms with five-decade track records decide which bets are worth making?
The thing is, Kleiner Perkins isn’t some newcomer chasing trends. They invested in Amazon when it was just an online bookstore. They backed Google when Yahoo was still king. They have a history of seeing around corners that other people miss.
But here’s the question that’s been keeping me up at night: What exactly are they betting on?
Because if you look at what’s actually happening in AI right now, the picture gets weird. Really weird.
OpenAI Just Killed Its Own App
Six months. That’s how long Sora lasted.
OpenAI launched Sora as this AI-first TikTok competitor back in September 2025. The idea was simple — let people generate AI videos and share them in a social feed. It had this feature called “cameos” where you could scan your face and create realistic deepfakes of yourself. The tech was genuinely impressive. The Sora 2 model could generate video and audio that looked almost real.
And now it’s dead.
OpenAI announced the shutdown on March 24, 2026. No real explanation. No timeline for when it officially disappears. Just — poof. Gone.
The numbers tell part of the story. Sora peaked at 3.3 million downloads in November 2025. By February 2026, that had crashed to 1.1 million. For context, ChatGPT has 900 million weekly active users. Sora was a rounding error.
But here’s what fascinates me: It wasn’t just that people didn’t want another TikTok. It’s that Sora became something nobody expected — a minefield of creepy deepfakes and legal nightmares.
When AI Gets Too Weird, Even for AI Companies
I remember when Sora first launched. TechCrunch writer Amanda Silberling posted a TikTok warning about the coming wave of deepfake tech. It got 300,000 views — way more than her usual content. People were genuinely scared.
And they should’ve been.
Within weeks, Sora was flooded with deepfakes of Sam Altman walking through slaughterhouses asking about “piggies enjoying their slop.” Martin Luther King Jr. videos started appearing. Robin Williams’ daughter had to go on Instagram begging people to stop making AI videos of her deceased father.
The guardrails that were supposed to prevent this? Useless. People found workarounds immediately.
Then came the copyrighted characters. Mario smoking weed. Naruto ordering Krabby Patties. Pikachu doing ASMR. Disney, a company famous for suing preschools over Mickey Mouse drawings, didn’t sue — they offered OpenAI a $1 billion investment and a licensing deal instead.
That deal died with Sora. No money actually changed hands.
The Real World Is Pushing Back
Here’s where Ida Huddleston’s story connects to everything else. That AI company that wanted her land? They’re still trying to rezone 2,000 acres nearby. They might build that data center after all, right next to her family’s farm.
But the resistance is spreading.
In Los Angeles, a jury just ordered Meta and Google to pay $6 million in damages for contributing to a young woman’s mental health struggles. The plaintiff, a 20-year-old named Kaley, convinced jurors that Instagram and YouTube’s addictive design caused her anxiety, depression, and body dysmorphia. Meta’s lawyers tried to blame her parents’ divorce. The jury didn’t buy it.
This wasn’t an isolated case. New Mexico handed Meta another defeat the day before. TikTok and Snap settled similar lawsuits rather than risk trials. We’re potentially looking at the “tobacco moment” for social media — the point where courts finally hold these companies accountable for the harms they cause.
What the VCs Actually See
So if consumer AI apps are flopping and the real world is pushing back, why are VCs throwing billions at the space?
I think the answer is simpler than it looks. They’re not betting on consumer apps. They’re betting on infrastructure.
Look at where the money is actually going. Drone delivery companies like Zipline just raised another $200 million. Lucid Bots got $20 million for window-washing drones. Brinc launched a drone that can supposedly replace police helicopters.
These aren’t toys. These are real businesses solving real problems.
The VCs see what I’m starting to see — the consumer AI hype cycle is crashing into reality, but enterprise AI and physical infrastructure are just getting started. Data centers, robotics, delivery drones, industrial automation — that’s where the next wave is actually building.
Kleiner Perkins’ portfolio tells the story. Together AI, Harvey, OpenEvidence — these are B2B tools that actually make money. Anthropic and SpaceX, both expected to IPO this year, aren’t consumer plays. They’re infrastructure.
The Disconnect Between Silicon Valley and Everyone Else
I’ve been thinking about this gap a lot lately. In San Francisco, people get excited about a new AI app that generates videos of cartoon characters. In Kentucky, an 82-year-old woman is worried about water pollution.
Both are reacting to the same technology. But they’re seeing completely different things.
Silicon Valley sees possibility — new ways to create, connect, and build. Ida Huddleston sees threat — land being consumed, resources being drained, communities being disrupted.
And here’s the uncomfortable truth: They’re both right.
The AI revolution is real. It’s already changing how we work, create, and communicate. But it’s also consuming enormous amounts of energy and water. It’s creating legal and ethical nightmares we’re not prepared for. It’s disrupting industries and communities that never asked for disruption.
Sora’s failure isn’t just a product flop. It’s a signal. The “move fast and break things” approach that worked for social media isn’t going to work for AI. The stakes are too high. The technology is too powerful. And the public is starting to notice.
What Happens Next
I’m not saying AI is overhyped. If anything, I think we’re still underestimating how transformative this technology will be. But I do think we’re in a transition period — a messy, confusing moment where the consumer AI gold rush is cooling off while the real infrastructure gets built.
OpenAI didn’t kill Sora because AI video generation is a bad idea. They killed it because Sora became a liability. The deepfakes, the lawsuits, the PR disasters — it wasn’t worth it for 1.1 million downloads and $2.1 million in revenue.
But the Sora 2 model isn’t dead. It’s just tucked behind the ChatGPT paywall now, where OpenAI can control who uses it and how. That’s the future of consumer AI, I think — not open social platforms where anyone can generate anything, but gated tools with real accountability.
What This Means for You
If you’re building in AI right now, here’s my honest take: Don’t build another consumer app that lets people generate content. That ship has sailed, and it’s not coming back.
Instead, look at what actually worked in this whole mess. Look at the drone companies finding traction while other robotics plays stalled. Look at the B2B tools that quietly make money without generating headlines. Look at the data center operators who don’t care about viral growth — they care about reliable power and water access.
The next wave of AI isn’t about giving everyone creative superpowers. It’s about solving real problems for real businesses. It’s about infrastructure that works quietly in the background, not apps that dominate your attention.
And if you’re just watching from the sidelines? Pay attention to the Ida Huddlestons of the world. When an 82-year-old farmer understands the risks better than most tech executives, something has shifted. The public is waking up to what AI actually means — not just the possibilities, but the costs.
That $26 million offer she turned down? It might be the most important number in tech this month. Not because of what it was, but because of what it represents — a line in the sand. A reminder that not everyone thinks the AI revolution is automatically good.
The VCs are betting billions because they see the infrastructure play. OpenAI killed Sora because the consumer play got too messy. And Ida Huddleston said no because she sees something the rest of us are still figuring out.
The AI wave isn’t stopping. But it’s definitely changing course. And honestly? I’m curious to see where it goes next.
📖 Related: Claude’s Paid User Surge: What It Means for You