The Innovation Disconnect
The Verge recently published a piece asking a question that should make every tech company uncomfortable: Has Silicon Valley forgotten how to build products normal people actually want?
The article traces a familiar pattern. NFTs promised to revolutionize ownership. The metaverse was going to replace physical reality. Now AI is being positioned as the solution to everything, whether people asked for it or not. Each wave follows the same script: breathless hype, massive investment, and then confusion when actual consumers shrug their shoulders.
Here's the uncomfortable truth: most people don't want to live in a virtual world, own digital apes, or have AI rewrite their text messages. They want products that solve real problems without requiring a PhD to understand.
The AI Exception That Proves The Rule
But here's where it gets interesting. AI isn't like NFTs or the metaverse. The difference? AI is already solving actual problems for actual people, right now.
The key is asking the right question upfront: how can AI solve this specific problem? Not "how can we shoehorn AI into this product?" but "what genuine pain point can AI address better than any alternative?"
Customer service is the perfect example. Every business faces the same fundamental challenge: customers need help 24/7, across multiple channels, in multiple languages. The traditional solution is hiring more people, which is expensive, slow to scale, and operationally complex. The AI solution? An AI workforce that handles conversations across chat, email, and phone without the overhead.
This isn't innovation for innovation's sake. It's solving a real problem that every growing business faces.
What Normal People Actually Want
Let's get specific about what "normal people" means in a business context.
A customer emailing your support team at 11 PM doesn't want to wait until morning. They want an answer now. They don't care whether it comes from a human or an AI agent. They care that it's accurate, helpful, and fast.
A business owner doesn't want to spend weeks recruiting, training, and managing a support team just to handle basic inquiries. They want to focus on building their product and growing their business. The mechanics of how customer conversations get handled? That should just work.
A support manager drowning in tickets doesn't want another dashboard or analytics tool. They want fewer tickets in the queue and more time to handle complex issues that actually require human judgment.
These aren't science fiction problems. They're the daily reality for thousands of businesses. And they don't require revolutionary new paradigms or asking people to change their behavior. They just require AI that works.
The Real Innovation Is Invisible
The best technology disappears into the background. Your customers shouldn't need to know they're talking to an AI agent. They should just get their problem solved.
This is why so much of Silicon Valley's recent innovation has failed. NFTs required people to understand blockchain, crypto wallets, and digital scarcity. The metaverse required VR headsets, new social behaviors, and a willingness to conduct life in a cartoon world. Both asked normal people to completely change how they live and think.
Effective AI does the opposite. It adapts to how people already behave. Someone sends an email. They get a helpful response. Someone starts a chat. Their question gets answered. Someone calls. The conversation happens naturally. The AI is invisible.
The Surface-Level AI Trap
Here's where most companies go wrong: they implement AI at the surface level without understanding what's actually happening underneath.
They add a chatbot to their website that can't handle anything beyond FAQ lookups. They implement "AI" that's really just keyword matching with extra steps. They build systems that work perfectly in demos but fall apart when real customers ask real questions.
This surface-level approach is exactly what the article criticizes. It's AI for AI's sake, not AI to solve problems. And customers see through it immediately.
The companies that succeed with AI are the ones that dive deep into the details. They understand not just that AI can handle customer conversations, but how it handles edge cases, context switching, multi-turn dialogues, and the thousands of small details that separate a working system from a frustrating one.
What Businesses Actually Need
Strip away the hype and the fundamental need is simple: businesses need to handle customer conversations at scale without exponentially increasing costs or complexity.
Every channel a business adds multiplies the operational burden. Chat, email, phone, social media—each requires staffing, training, quality control, and management. The traditional model doesn't scale without massive headcount increases.
An AI workforce changes the equation. Instead of scaling headcount, you scale intelligence. The same AI agents that handle chat can handle email and phone. They work 24/7 across every channel. They learn from every conversation. They get better over time instead of burning out.
This isn't asking businesses to change how they operate or customers to change how they communicate. It's just removing the bottleneck that's always existed between customer demand and business capacity.
Building For Real People
The Verge article asks if Silicon Valley remembers how to build consumer products. The answer isn't going back to pre-AI approaches. It's building AI products that actually solve problems people have.
That means starting with the problem, not the technology. It means testing with real customers in real scenarios, not just optimizing demo performance. It means obsessing over the experience, not the underlying tech stack.
When someone's customer service experience improves because they get instant, accurate help instead of waiting in a queue, that's successful AI. When a business scales support without tripling their team, that's successful AI. When the technology disappears and only the results remain, that's successful AI.
The Path Forward
Silicon Valley hasn't forgotten how to build products normal people want. But it has developed a bad habit of building products it thinks people should want, then being surprised when they don't.
The companies that succeed with AI will be the ones that stay grounded in real problems, real users, and real results. Not the ones chasing the next hype cycle or trying to create artificial demand for solutions without problems.
The AI revolution in customer service isn't coming. It's already here. But it's not the flashy, headline-grabbing kind. It's the kind where customers get better service, businesses operate more efficiently, and everyone wonders why it wasn't always this way.
That's what normal people actually want. And that's exactly what AI should deliver.