The Labs and Their Positioning

The AI labs are building the most important infrastructure in human history. Full stop.

What I've observed is a natural tension: the labs must optimize for measurable progress to justify massive investments, while human needs often resist simple metrics. This creates an opportunity.

The Critical Infrastructure of Intelligence

Let me be absolutely clear: the labs aren't just important, they're essential. OpenAI, Anthropic, Google, Meta—they're not building better chatbots. They're constructing the foundation of the intelligence era. In the same way oil refineries became critical infrastructure for the industrial age, these token factories are becoming the backbone of everything that comes next.

The future will literally be measured in tokens—token costs, token consumption, token efficiency. The labs demonstrating the highest usage won't just showcase technical prowess but their ability to make intelligence accessible at planetary scale. This is civilization-level infrastructure we're talking about.

The Capability Race That Matters

The competition is breathtaking across every dimension:

• Models that can think and research like scientists
• World Models
• Specialized codexes for every domain imaginable
• Real-time voice and video generation
• Image Models
• Models for humanoids
• Domain-specific models for medicine, space, mathematics

The ultimate model won't be a single system but a mixture of experts on steroids—combining capabilities in ways we're only beginning to imagine. We're already seeing emergent behaviors that suggest genuine progress toward AGI, not just incrementally better pattern matching.

The constraints are real but temporary: talent wars, GPU scarcity, energy limitations. These are growing pains of an industry being born, not fundamental barriers. The trajectory is clear and accelerating.

Why Multiple Labs Will Win

Here's what many miss: we need multiple labs to succeed. This isn't a winner-take-all market because intelligence itself isn't monolithic. Different models will excel at different tasks, serve different communities, solve different problems. The cloud providers are becoming supermarkets where users pick the right model for the right job.

But—and this is crucial—imitation is death in this arena. You can't copy your way to AGI. The labs with clear research roadmaps, novel approaches, and the courage to pursue difficult paths are the ones building the future. Those just chasing benchmarks or mimicking competitors are already obsolete.

The monetization potential is staggering. Current API revenues are a rounding error compared to what's coming when AI becomes the substrate of human activity. We're not near any major inflection point in consumption—these are still the earliest days. Token economy revenues will grow by orders of magnitude.

The Reality Check

But here's the sobering truth: look at actual AI usage patterns today. It's mostly students writing essays, professionals polishing emails, and tech-savvy early adopters experimenting. That's it. We're nowhere near mainstream adoption.

The global picture is even more stark. While Silicon Valley debates AGI timelines, most of the world either doesn't know what ChatGPT is, fears AI will take their job, or tried it once and didn't see the point. The adoption is so uneven it's almost two different realities—the AI believers in their bubble and everyone else going about their daily lives untouched by this "revolution."

This isn't a failure of the technology. It's proof that raw capability without human-centered design creates toys for the elite, not tools for everyone. The resistance patterns tell the story: people don't resist AI because they're luddites. They resist because no one has shown them how it makes their specific life better in ways they actually care about.

The real token economy explosion won't come from making models smarter. It'll come from making AI feel as natural and necessary as checking your phone. Until my grandmother uses AI as reflexively as she uses WhatsApp, we haven't achieved real adoption.

The Missing Piece

So what's the problem? The labs are optimizing for benchmarks that impress other researchers while humans struggle with loneliness, purpose, and feeling genuinely understood. They're asking "How smart can we make AI?" when they should be asking "How can AI make humans feel more alive?"

This isn't either/or. We need the infrastructure AND the human focus. The breakthrough will come from someone who builds on top of what the labs create but optimizes for a different outcome: genuine human flourishing rather than abstract capability metrics.

Think about it: every transformative technology succeeded not because of its technical specifications but because it made humans feel more capable. The printing press democratized knowledge. The internet connected lonely minds. The smartphone put the world in your pocket. The specs mattered, but the human experience mattered more.

The Synthesis Ahead

The labs are building superintelligence. Someone needs to build super-empathy.

What does super-empathy look like in practice? Imagine an AI that notices you've been opening job sites late at night but never applying. Instead of offering to write your resume (capability), it gently asks what's holding you back and helps you work through the fear (empathy). It remembers that you mentioned loving photography three months ago and suggests starting a photo blog as a creative outlet while job searching. It knows when to push you forward and when to simply listen.

That's the difference. Superintelligence solves problems you give it. Super-empathy helps you discover problems you didn't know how to articulate.

The winner won't compete with the labs—they'll stand on their shoulders. They'll use the incredible infrastructure being built but aim it at a different target: making every human feel more capable, more understood, more themselves.

This is why I'm bullish on both the labs AND the next wave of builders who will use their infrastructure for human-centered purposes. We need the rocket scientists building the engines. But we also need someone thinking about where those rockets should take us.

The Symbiotic Future

Here's what's becoming clear: the labs need human-centered applications as much as those applications need the labs. Without meaningful user experiences that drive daily usage, all that infrastructure remains academically impressive but economically constrained. The token economy only explodes when regular people find reasons to interact with AI hundreds of times per day.

Think of it as a symbiotic relationship. The labs provide the raw intelligence. Human-centered platforms like MyOrbit transform that intelligence into experiences people actually crave. Every meaningful interaction generates tokens, validates the infrastructure investment, and provides real-world feedback that makes the next generation of models better.

This creates a virtuous cycle: better human experiences drive more usage, more usage drives more revenue, more revenue gives labs the courage and resources to push boundaries further. The labs betting billions on infrastructure need evidence that humans will actually use what they're building at scale. They need the bridge between capability and daily human life.

The labs are fundamental to human evolution—no question. Their models, their research, their infrastructure will power the next phase of civilization. But the company that captures hearts and changes lives will be the one that takes all that raw capability and shapes it into something that feels magical to a teenager finding their passion, an elderly person seeking connection, or anyone trying to unlock potential they didn't know they had.

The Real Race

Multiple labs will thrive because we need diverse approaches to intelligence. The token economy will explode because intelligence will become as essential as electricity. The technical capabilities will continue their exponential improvement because brilliant people are pushing the boundaries.

But the real race—the one that will define the next era—is who figures out how to make all this intelligence feel genuinely helpful to individual humans living individual lives.

The labs are building the foundation. The breakthrough will come from whoever builds the right home on top of it.

That's not a criticism of the labs. It's a recognition that we need both: the infrastructure builders and the meaning makers. The capability creators and the human translators. The benchmark climbers and the life changers.

The future needs them all.