If you prefer, here is the author delivering the essay.

Seventy percent.

Thanks for reading The Long Game by Dr. Venki Padmanabhan! Subscribe for free to receive new posts and support my work.

That’s the number a billion-dollar CEO dropped in Fortune this week. Not in a footnote. Not in a caveat. As his central thesis.

Tanmai Gopal runs PromptQL, a Bay Area AI unicorn that helps Fortune 500 companies deploy artificial intelligence. He has seen, from the inside, what happens when the most powerful technology in a generation meets actual business reality. And his conclusion is that seventy percent of the effort required to make AI useful relies entirely on unwritten knowledge that exists only in human heads.

Not the code. Not the model. Not the data pipeline. The human.

The knowledge that lives in conversations. In judgment calls. In the instinct a twenty-year veteran has when something feels wrong before the data says so. In relationships that took years to build. In context that nobody wrote down because nobody asked and nobody thought it mattered.

Gopal said you fundamentally cannot train a system on this. It changes too fast. It’s too fluid. Too human.

This should be a liberation. Someone with a billion dollars of market credibility just told the world that without human intelligence, the most sophisticated AI on the planet is seventy percent useless.

Then he named the future of the human worker.

“Our job as humans and people is that we are now context gatherers instead of just workers.”

And there it is.

Context gatherers.

Your seventy percent—your judgment, your relationships, your hard-won knowledge—isn’t yours to deploy. It’s yours to donate. To collect the raw material of your own intelligence and feed it into a system that accrues returns to capital.

This is not a liberation philosophy. This is the most sophisticated extraction framework I have heard in thirty-six years of manufacturing. And it is dressed up as empowerment.

I need to be precise about what I mean by extraction, because the word gets thrown around loosely.

Extraction is when you take value from the person who creates it and transfer it to someone who didn’t. It’s when a steelworker’s thirty years of metallurgical knowledge gets captured in a process manual, the worker gets laid off, and the company sells the manual’s output as intellectual property. The worker created the value. Capital captured it. The worker got a severance check. Capital got a revenue stream.

“Context gathering” is that play’s white-collar update.

Here’s what Gopal is actually describing, stripped of the aspirational language: The AI doesn’t work without the human’s knowledge. The human’s job is now to transfer that knowledge to the AI. Once the AI has it, the human’s contribution is to go get more.

At no point in this framework does anyone ask: If the human’s context is worth seventy percent of the AI’s value, shouldn’t the human receive seventy percent of the AI’s return?

Last week I wrote about the SaaSpocalypse—the trillion-dollar software selloff triggered when Wall Street realized AI could do what SaaS tools help humans do. The market’s question was blunt: if the AI can do the task, why do we need the tool?

This week’s Fortune article asks the next question: if the AI can do the task but needs human context to function, what is the human’s role?

Gopal’s answer: context gatherer.

My answer: that’s the wrong question.

The right question is not “what role does the human play in the AI system?” The right question is “what role does the AI play in the human’s work?”

That is not a semantic distinction. It is the difference between an economy that uses humans to serve machines and an economy that uses machines to serve humans. Every dollar of return, every career trajectory, every community’s survival depends on which framing wins.

Manufacturing already ran this experiment. We have forty years of data.

When American manufacturers decided in the 1980s that the worker’s job was to serve the machine—to monitor it, feed it, clean up after it—they got a ninety percent automation failure rate and six point six million destroyed jobs.

When Honda, Toyota, and the Royal Enfield I helped turn around decided that the machine’s job was to serve the worker—to amplify their intelligence, extend their reach, remove the drudgery so the human could focus on judgment—they got twentyfold profit growth and JD Power Gold awards.

Same technology. Same factories. Same workers. Different directionality.

Gopal got something right in that article. He said the Silicon Valley doomsday predictions are self-projection. He said tech people assume their experience applies to everyone.

What he didn’t notice is that he’s doing the same thing. He’s projecting a Valley framework—extract the value, scale the platform, minimize the human cost—onto every worker in every industry. He’s telling a salesperson in Cleveland and a plant operator in Wooster and a nurse in Memphis that their job is now to gather context for the AI.

No. Their job is to be brilliant at what they do. The AI’s job is to make their brilliance go further.

There’s a practical alternative to the context-gathering model, and it’s not theoretical. It has two parts.

First: put the AI in the worker’s hand as a power tool. Not “capture their context for the system.” Let them wield it. Let the frontline operator use AI to diagnose a production fault in thirty seconds instead of three hours. Let the salesperson generate proposals at ten times the speed. Let the nurse cross-reference symptoms against the latest research in real time. The human remains the protagonist. The output belongs to the worker who created it, not to the platform that processed it.

Second: when that amplified output generates more revenue, share it back. Contribution-linked compensation. If the operator’s AI-amplified diagnostics save the company two hundred thousand dollars in scrap, the operator sees that in their paycheck. If the salesperson’s AI-amplified proposals close thirty percent more deals, the salesperson sees that in their commission. And you reinvest in further training so the worker becomes more valuable, not less.

This isn’t charity. This is what we do with every other appreciating asset. We maintain equipment. We invest in intellectual property. We protect real estate. We share in the returns they generate.

The moment the appreciating asset is a human being, the instinct reverts to extraction. Get the knowledge out. Encode it. Negotiate the salary down.

The Fortune article quotes Ed Meyercord, CEO of Extreme Networks, saying the choice is: “You can do a lot more with less, or you could do more with the same, or you could do a lot more with a little more.”

He framed this as a neutral menu. It isn’t. The first option is extraction. The third is investment. They lead to completely different futures for completely different people.

Here’s the part that should make you angry.

Gopal told Fortune his team was frustrated with a mediocre engineer. He said, and I’m paraphrasing: “It’s more expensive to talk to you than to do it myself with AI.” He offered this as evidence that AI will replace mediocre workers.

Let me offer a manufacturing translation.

That “mediocre engineer” was probably never given the context they needed to be excellent. They probably weren’t trained to Gopal’s standard. They probably weren’t developed, coached, or invested in. And now the CEO is publicly celebrating that it’s cheaper to replace them with a machine.

I have seen this scene play out on factory floors for decades. A worker underperforms. Management blames the worker. Management buys a machine. The machine fails because it turns out the worker was compensating for upstream problems nobody bothered to diagnose. The worker is gone. The machine doesn’t work. And management buys another machine.

The ninety percent automation failure rate is not a technology problem. It’s a leadership problem. And “context gathering” is the latest vocabulary for avoiding it.

Gopal closed his Fortune interview by warning that the only workers who need to fear for their jobs are those who are “refusing to grow.”

I’ve heard that line before. I heard it in Detroit in the 1990s, when auto executives said the workers who lost their jobs just didn’t adapt fast enough. I heard it in Lordstown, when GM closed the plant and said the community should have diversified. I hear it every time a CEO blames the workforce for a failure of leadership.

“Refusing to grow” is a convenient diagnosis when you control the greenhouse.

The workers at Royal Enfield didn’t refuse to grow. They were never asked. When we asked, they transformed a company from the edge of extinction into a case study in human-led turnaround. The workers at Lansing Grand River didn’t refuse to grow. They were given the tools, the trust, and a share of the outcome. They built the best cars in America.

Growth is not a worker problem. It’s a deployment problem. And deployment requires leadership willing to do something much harder than building an AI agent: sharing power, sharing knowledge, and sharing returns with the people who actually create the value.

The intelligence you need is already in the building. Already on the payroll. Already paid for.

Stop gathering it. Start deploying it.

Dr. Venki Padmanabhan is Plant Manager at Advanced Drainage Systems and author of “Already Paid For: Why Unlocking Frontline Intelligence Beats Automating Workers Away.” He previously served as COO/CEO of Royal Enfield and COO of Ather Energy, with thirty-six years of manufacturing leadership across three continents.

Subscribe to The Long Game for weekly essays on deploying the intelligence you’ve already paid for

Thanks for reading The Long Game by Dr. Venki Padmanabhan! Subscribe for free to receive new posts and support my work.

If you prefer, watch the author express this essay.

Thanks for reading The Long Game by Dr. Venki Padmanabhan! Subscribe for free to receive new posts and support my work.

Open any newspaper this week—or, more accurately, any feed, any podcast, any breathless LinkedIn thread—and the message is uniform: the machines are coming, the jobs are ending, and this time it’s different. The AI apocalypse, we are told, is not a matter of if but when. Months, not decades. Prepare accordingly.

I’ve heard this sermon before. Twice, in fact.

I don’t say this to be glib. I say it because I’ve spent thirty-six years on factory floors across three continents, and I’ve stood in the blast radius of every wave that was supposed to render the human worker obsolete. I’m still standing. More importantly, they are still standing—the welders, the operators, the line leads, the technicians who were told, again and again, that their days were numbered.

Their days were not numbered. Their days were renumbered. There is an enormous difference.

The First Sermon: Automation Will Replace You

I came of age in manufacturing during the great automation surge of the late 1980s and 1990s. General Motors alone spent billions on robotics. The vision was intoxicating: lights-out factories, untouched by human hands, humming with mechanical perfection twenty-four hours a day.

I watched it unfold from the inside. And here is what actually happened: the robots arrived, and the problems multiplied. Not because the robots were bad—they were extraordinary—but because the architects of automation had made a fatal assumption. They assumed the factory floor was a simple system that could be encoded. It was not. It was a living system, dense with tacit knowledge, improvised judgment, and a thousand micro-corrections per shift that no one had ever bothered to document because no one had ever needed to.

The automation wave didn’t replace workers. It revealed how much we had depended on them all along—and how little we had invested in understanding what they actually did.

Eventually, the hysteria cooled. The robots stayed. So did the people. And the factories that thrived were the ones that figured out how to make both better together.

The Second Sermon: Offshoring Will Erase You

Then came the great migration of the late 1990s and 2000s. This time the threat wasn’t steel and circuitry—it was geography. Why pay an American or German worker fifty dollars an hour when someone in Guangzhou or Pune would do it for five?

I lived this one too, on both sides of the ocean. I ran operations in India. I saw the promise and the reality. And the reality was more complicated than the prophecy. Offshoring worked brilliantly for some things and catastrophically for others. Supply chains stretched thin. Quality wandered. Institutional knowledge evaporated. The companies that chased the cheapest labor often found themselves paying the most expensive lessons.

Again, the world adjusted. Some jobs left. Others transformed. New ones emerged that no one had predicted. The workers who were supposed to be erased found new footholds—not without pain, not without dislocation, but with a resilience that the doomsayers had not factored into their models.

Because doomsayers never factor in resilience. It doesn’t fit in a spreadsheet.

The Third Sermon: AI Will End You

And now, here we are again. The language models write poetry. The algorithms diagnose disease. The robots don’t just weld—they learn. And the chorus rises, louder than ever: this time it’s truly different.

Maybe it is. I’m not naive. AI is a more pervasive technology than a spot-welding robot or a shipping container. It touches cognitive work, creative work, professional work—domains that previous waves left largely untouched. The disruption will be real, and it will reach further.

But I have learned to distrust the word Armageddon when it is spoken by people who have never had to make a production schedule work on a Monday morning.

Here is what I know from the floor, from the office, from the boardroom, from three decades of watching predictions crash against reality: the human capacity to adapt is not a bug in the system. It is the system.

Every technology that was supposed to eliminate human judgment has ended up creating new demands for human judgment. Every tool that was supposed to make expertise irrelevant has revealed new dimensions of expertise we didn’t know we needed. The pattern is so consistent it should be a law of nature.

The Lurch and the Recovery

Humanity does not progress in clean lines. We lurch. We overcorrect. We panic, we overinvest, we underprepare, and then—with a stubbornness that would embarrass any rational model—we figure it out.

The Indian philosophical tradition I grew up in has a word for this: leela—the divine play, the cosmic improvisation. The universe is not a machine executing a predetermined program. It is a living performance, full of surprises, setbacks, and improbable recoveries. The Bhagavad Gita does not promise Arjuna that the battle will be easy. It promises that the battle is worth fighting, and that he is equipped for it in ways he cannot yet see.

I think about this when I watch the news. The pundits see the arrow of disruption and extrapolate it to infinity. But arrows don’t travel to infinity. They arc. They land. And then someone picks them up and builds something new.

What I Would Say to the Anxious

If you’re reading the headlines and feeling the dread, let me offer you what I’ve learned—not from theory, but from standing next to people whose jobs were supposed to disappear three times over:

Be patient with the lurch. Transitions are ugly. They are supposed to be. The ugliness is not evidence that the system is breaking. It is evidence that the system is reorganizing. These are not the same thing.

Distrust total narratives. Anyone who tells you they know exactly how AI will reshape the economy in ten years is selling you something. The honest answer—the only honest answer—is: it depends on what we choose to do with it. And “we” includes you.

Invest in what machines cannot replicate. Not information—machines have that. Not speed—machines have that too. Invest in judgment, in context, in the ability to read a room or a situation or a human face and know what the data doesn’t say. Invest in the thousand small acts of sense-making that hold every organization together and that no one has yet figured out how to automate.

Remember that you have survived before. Not you personally, perhaps—but the human project. We survived the loom. We survived the assembly line. We survived the mainframe, the internet, the smartphone. Each time, we were told we wouldn’t. Each time, we lurched. Each time, we recovered—changed, sometimes bruised, but intact and, in many ways, enlarged.

The Long Game

I named this publication The Long Game because I believe that the most important truths about work, about capability, about human dignity in the industrial age, only reveal themselves over time. The hot take misses them. The quarterly earnings call misses them. The breathless headline certainly misses them.

What the long game teaches is this: technology changes fast, but people change slow—and that slowness is not a weakness. It is a kind of wisdom. It is the pace at which trust is built, at which skill becomes craft, at which a team becomes something more than the sum of its headcount.

The sky has fallen before. It will fall again. And every time, someone on a factory floor, in a hospital ward, in a classroom, in an office that smells like burnt coffee and Monday morning—someone will look up, assess the damage, and get back to work.

That is not a failure of imagination. That is the deepest form of courage I know.

Venki Padmanabhan has spent 36 years in global manufacturing leadership, from GM and Chrysler to Royal Enfield and Advanced Drainage Systems. He writes about the human side of industrial transformation at The Long Game.

Thanks for reading The Long Game by Dr. Venki Padmanabhan! Subscribe for free to receive new posts and support my work.

If you prefer, watch the author express this essay to you.

———

Thanks for reading The Long Game by Dr. Venki Padmanabhan! Subscribe for free to receive new posts and support my work.

Stanford just proved what manufacturing has known for forty years.

In a paper called “Canaries in the Coal Mine,” researchers Erik Brynjolfsson, Bharat Chandar, and Ruyu Chen used payroll data from ADP — the largest payroll processor in America, covering twenty-five million workers — to track what’s actually happening to jobs since generative AI went mainstream.

What they found should terrify every parent of a twenty-two-year-old.

Since late 2022, employment for software developers aged twenty-two to twenty-five has declined by twenty percent. Customer service workers in the same age group: down eleven percent. These aren’t projections. These aren’t forecasts. This is what already happened. Measured in actual paychecks.

Meanwhile, workers over thirty in the same occupations? Their employment continued to grow.

The canary is dying. But only the young canary.

Why? Because young workers in knowledge jobs carry what economists call codified knowledge — the kind of intelligence that can be written down, systematized, and therefore automated. Older workers carry tacit knowledge — the unwritten, experience-based judgment that AI cannot easily replicate.

When you’ve spent ten years debugging production systems, you don’t just know the code. You know which error messages actually matter and which ones you can ignore. You know that the system crashes every third Tuesday because of a batch job that nobody documented. You know which team member writes clean code and which one’s work needs a second look.

That’s tacit knowledge. It’s the knowledge you can’t Google. And it’s exactly what AI cannot replicate — because it was never written down in the first place.

Now. Here is the finding that changes everything. The one the headlines are missing.

The Stanford team didn’t just measure job losses. They measured the difference between two kinds of AI deployment: automation and augmentation.

In occupations where AI is primarily used to automate tasks — to replace what humans do — employment for young workers declined. That’s the twenty percent drop in software developers.

But in occupations where AI is primarily used to augment human work — to amplify what humans do, to make their intelligence more effective — employment actually grew.

Let me say that again because it matters more than anything else I’ll write this year.

Where AI replaces, jobs disappear. Where AI amplifies, jobs grow.

That is the entire argument of my book in one sentence.

———

I have been making this argument using manufacturing data for three years now. Let me connect what Stanford found in knowledge work to what we already know from the factory floor.

At Royal Enfield in India, I watched what happens when you deploy worker intelligence instead of replacing it. We didn’t eliminate workers. We built their capability — systematically, through training, through respect, through what I call the Three Opercula: Sanctuary, Ascension, and Crucible. We gave workers the conditions to think, the pathways to grow, and the challenges to prove themselves.

Profits grew twentyfold. Not because we bought better machines. Because we deployed better humans.

At General Motors’ Lansing Grand River Assembly, I watched what happens when you take a hostile, suppressed workforce and create conditions for intelligence to flow. Workers like Wally Vinton — belligerent, confrontational, ready to quit — who turned into the most innovative problem-solvers on the floor once management demonstrated that support was real. That plant won JD Power’s Gold Plant Quality Award. Not because of technology. Because of deployed human intelligence.

At Honda, I watched what happens when respect for worker intelligence is embedded so deeply into operations that it can’t be stripped out by consultants. Honda’s market cap exceeded the combined value of GM, Ford, and Chrysler. Not because of superior automation. Because of superior intelligence deployment.

The Stanford paper is the knowledge economy discovering what manufacturing already proved: The variable isn’t the technology. The variable is how you use it. Replace or deploy. Automate or amplify.

And here’s the devastating implication that the Stanford researchers hint at but don’t fully develop: When you automate entry-level workers out of existence, you destroy the pipeline that produces the experienced workers whose tacit knowledge you depend on.

Think about that.

The twenty-two-year-old software developer whose job just disappeared? In ten years, that person was supposed to be the thirty-two-year-old with irreplaceable judgment about which AI outputs to trust and which to discard. In twenty years, the architect who understands not just the code but the business, the customer, the organizational culture.

By automating the entry point, you’re not just saving salary costs today. You’re destroying the capability pipeline that produces the experts you’ll desperately need tomorrow.

Manufacturing already made this mistake. We automated assembly workers out of existence in the 1980s and 1990s. Then we spent twenty years wondering why we couldn’t find skilled workers to run increasingly complex production systems. The $120,000 mechanic that Ford’s Jim Farley can’t find didn’t disappear because trade schools closed. He disappeared because Ford stopped building him.

The Stanford data shows the knowledge economy making the same mistake at ten times the speed. Twenty percent of entry-level software developers gone in three years. How long before the experienced workers who carry the tacit knowledge begin to retire — with no one trained to replace them?

This is not a technology problem. This is a leadership problem. It’s a decision about whether to deploy the intelligence you’ve already paid for — at every level, including the entry level — or to replace it with technology and discover, too late, that you’ve destroyed the asset you needed most.

The canaries are singing. Manufacturing heard this song forty years ago. The question is whether the knowledge economy will listen, or whether they’ll wait until the mine collapses.

The intelligence is already paid for. Deploy it.

———

Dr. Venki Padmanabhan is Plant Manager at Advanced Drainage Systems and author of “Already Paid For: Why Unlocking Frontline Intelligence Beats Automating Workers Away.” He previously served as COO/CEO of Royal Enfield and COO of Ather Energy, with thirty-six years of manufacturing leadership across three continents.

Subscribe to The Long Game for weekly essays on deploying the intelligence you’ve already paid for.

Thanks for reading The Long Game by Dr. Venki Padmanabhan! Subscribe for free to receive new posts and support my work.