Hiding Humans behind the AI Curtain

Posted on by Steve Sammartino

Why Big Tech is not as special as we imagine.

No drivers… in the seats that is.

Waymo’s recent admission in Washington — that its “driverless” robotaxis rely on real-time human guidance from offshore workers in the Philippines — shouldn’t surprise anyone. Sure, it made headlines and rattled lawmakers, but Big Tech has spent decades promoting frictionless automation while quietly relying on human labour to keep the magic alive. In the world of AI, humans still fill a zillion holes.

The list is quite long:

  • Amazon was leaning on more than a thousand human reviewers in India to make its “Just Walk Out” stores function.
  • Mechanical Turk was literally built as a marketplace for humans to perform tasks disguised as automation.
  • Facebook said AI would moderate and remove harmful content, but in reality it was (and still is) tens of thousands of human moderators dealing with humanity’s worst content in low-cost labour markets.
  • Self-driving car companies have quietly relied on tele-operators for years, stepping in whenever the algorithm becomes confused by a traffic cone or an unexpected patch of sunlight.
  • Siri / Alexa / Google Assistant all have humans listening and reviewing private audio recordings for quality training.
  • Dating apps have humans rating people on their looks so the ‘hot people’ can find each other.
  • Even ChatGPT is in on it. Its model depends on an enormous layer of humans training and curating the AI. People tag the data, label right and wrong answers, judge the outputs, provide reinforcement learning feedback, moderate the content.
  • We call all of this AI, but much of it is a carefully choreographed dance between machines and mass human labour.
The deeper truth is that Big Tech’s power doesn’t come from tech anymore. It did, in the first instance. What happened was their first technology offers built incredibly strong platforms with network effects. Think social networks which operate better if everyone is in the same place, a search engine which improves with every search, or an ecommerce platform that everyone sells on. These unlocks gave them access to untold profits from what essentially became ‘natural monopolies’. The result was that these ‘Big Tech firms’ initial offers turned into cash-generating firehoses which gave them the reputation and financial ability to fund all forms of technological advancement — from self-driving cars to now the generative AI boom.

The kicker is that when your company generates tens of billions in revenue every quarter, you can over-invest in moonshots, fill gaps with human workers, ship early, absorb regulatory heat, and fight lawsuits when stuff goes wrong. That’s the Chutzpah — the confidence that only insane cash flow can buy.

And for all the mythology around technical brilliance, in reality fewer than 10% of employees at major tech firms are software developers. The rest form the operational and logistical scaffolding that props up the story that tech is more autonomous than it actually is. Big Tech isn’t Magic, it’s the other ‘M’ – Money. This funds their ability to take risks and fake it (with humans) until the tech eventually catches up

Human Scaffolding: The Untold Accelerator

The real driver of tech adoption isn’t algorithms but human scaffolding. Every breakthrough arrives half-built, and people smooth the edges until machines can take over. Electricity needed operators. Early phones ran on human switchboards. Search engines relied on armies of taggers. Even early e-commerce depended on human pickers, not robots.

Today’s AI feels fully formed, but it’s the same pattern: human labour builds the trust, data and stability long before automation can truly scale.

Your company can do this too!

The biggest misconception in boardrooms is the belief that only Big Tech can pull off ambitious AI projects. But the truth is far more liberating: Big Tech isn’t ahead because its technology is flawless. It’s ahead because it has the 2 C’s – the Cash and the Courage to launch before everything works perfectly. They iterate fast, gloss over the messy parts, gap-fill with people, and fix on the fly.

Any firm can do the same, especially given many AI models are open source. The core ingredient is courage. Build prototypes, fill the gaps with humans, and automate those later. That’s how the giants are trying to maintain their dominance in this emergent AI era.

The Curtain Will Always Be There… so use it to your advantage.
Whether you’re a freelancer or a huge non-tech company remember this: You’re not competing with flawless machines created by god-tier engineers. You’re competing with companies that use humans, duct tape, and strong business models to manufacture the illusion of perfect automation.

That’s how Big Tech does it. And nothing is stopping you from doing the same

Keep Thinking,

Steve

Share

** Get me into do an AI keynote at your next event and as a blog reader – get 2 nights free at my luxury farmhouse on the Bellarine Peninsula.