Steam, Steel, and Infinite Intelligence
TechFlow Selected TechFlow Selected
Steam, Steel, and Infinite Intelligence
We're still in the "water wheel stage" of AI, forcing chatbots into workflows designed for humans.
Navigating Web3 tides with focused insightsBy Ivan Zhao, CEO of Notion
Translated by AididiaoJP, Foresight News
Each era is shaped by its unique technological raw materials. Steel forged the Gilded Age, and semiconductors ushered in the digital age. Now artificial intelligence arrives in the form of infinite intelligence. History tells us: whoever masters the raw material defines the era.
Left: Young Andrew Carnegie and his brother. Right: Steel mills in Pittsburgh during the Gilded Age.
In the 1850s, Andrew Carnegie was still a telegraph operator running through the muddy streets of Pittsburgh, when six out of ten Americans were farmers. Just two generations later, Carnegie and his peers had forged the modern world—horses gave way to railways, candlelight to electric light, iron to steel.
Since then, work has shifted from factories to offices. Today I run a software company in San Francisco, building tools for millions of knowledge workers. In this tech town, everyone talks about Artificial General Intelligence (AGI), but most of the two billion office workers have yet to feel its presence. What will knowledge work look like soon? And what happens when tireless intelligence becomes embedded within organizational structures?
Early films often resembled stage plays, with a single camera filming directly at the stage.
The future is often hard to predict because it disguises itself as the past. Early phone calls were as brief as telegrams; early films were like recorded stage plays. As Marshall McLuhan said: "We drive into the future using only our rearview mirror."
Today’s most common AI still resembles past Google search. To quote McLuhan again: "We drive into the future using only our rearview mirror." Today we see AI chatbots mimicking the Google search box. We are deep in that uncomfortable transitional phase common to every technological shift.
I don’t have all the answers about what comes next. But I find a few historical metaphors helpful in thinking about how AI might operate across individuals, organizations, and even entire economies.
Individual: From Bicycle to Car
The earliest signs can be seen among the “high performers” of knowledge work—programmers.
My co-founder Simon used to be a “tenx programmer,” but lately he rarely writes code himself. Walking past his desk, you’ll see him orchestrating three or four AI programming assistants simultaneously. These assistants not only type faster—they think. This has turned him into an engineer who is 30 to 40 times more productive. He often queues up tasks before lunch or bedtime, letting the AI continue working while he’s away. He has become a manager of infinite intelligence.
A 1970s study on motion efficiency in Scientific American inspired Steve Jobs’ famous metaphor of the “bicycle for the mind.” Yet for decades since, we’ve been riding bicycles on information superhighways.
In the 1980s, Steve Jobs called the personal computer a “bicycle for the mind.” A decade later, we built an “information superhighway” called the internet. But today, most knowledge work still relies on human effort. It’s as if we’ve been riding bicycles on highways.
With AI assistants, people like Simon have already upgraded from bicycles to cars.
When will other types of knowledge workers get behind the wheel? Two problems must first be solved.
Why is AI-assisted knowledge work harder than programming assistance? Because knowledge work is more fragmented and harder to verify.
First is context fragmentation. In programming, tools and context are often centralized: integrated development environments, code repositories, terminals. But general knowledge work is scattered across dozens of tools. Imagine an AI assistant trying to draft a product brief—it needs to pull information from Slack threads, strategy documents, last quarter’s dashboard data, and organizational memory stored only in someone’s head. For now, humans act as glue, stitching everything together with copy-paste and browser tab switching. Until context is unified, AI assistants remain limited to narrow use cases.
The second missing element is verifiability. Code has a magical property: you can test it and check for errors. Model developers leverage this, training AI to code better via reinforcement learning and similar methods. But how do you verify whether a project is well-managed, or if a strategic memo is excellent? We haven’t yet found ways to improve general knowledge work models. Thus, humans must remain in the loop—to supervise, guide, and demonstrate what “good” looks like.
The 1865 Locomotive Act required vehicles to be preceded by a man carrying a red flag while driving on public roads (repealed in 1896).
This year’s experience with programming assistants shows that “humans in the loop” isn’t always ideal. It’s like having someone inspect every bolt on an assembly line, or walking ahead of a car waving a flag (see the 1865 Red Flag Act). We should place humans above the loop, not inside it. Once context is integrated and work becomes verifiable, billions of workers will shift from “riding bicycles” to “driving cars,” and eventually from “driving” to “autonomous driving.”
Organization: Steel and Steam
Companies are a modern invention, and they lose efficiency as they grow, eventually hitting limits.
An 1855 organizational chart of the New York and Erie Railroad Company. Modern corporations and their hierarchies evolved alongside railroads—the first businesses requiring coordination of thousands across vast distances.
Centuries ago, most companies were workshops of a dozen people. Today we have multinational corporations with hundreds of thousands of employees. Communication infrastructure, relying on meetings and human brains passing information, buckles under exponentially growing loads. We try to solve this with layers, processes, and documentation—but it’s like building skyscrapers with wood, using human-scale tools for industrial-scale problems.
Two historical metaphors illustrate how the future could look different when organizations gain access to new technological raw materials.
The miracle of steel: the Woolworth Building in New York, completed in 1913, was once the world’s tallest building.
The first is steel. Before steel, 19th-century buildings were limited to six or seven stories. Iron was strong but brittle and heavy—adding floors would cause structures to collapse under their own weight. Steel changed everything. Strong and flexible, it allowed lighter frames, thinner walls, and towers rising dozens of stories, enabling entirely new architectural forms.
AI is the “steel” of organizations. It promises continuity of context across workflows, surfacing decisions without noise. Human communication no longer needs to serve as load-bearing walls. Weekly two-hour alignment meetings may become five-minute asynchronous reviews; executive decisions requiring three layers of approval might be completed in minutes. Companies could truly scale without the performance decay we once considered inevitable.
A mill powered by a waterwheel. Water power was strong but unstable, constrained by location and season.
The second story is about the steam engine. In the early Industrial Revolution, textile factories were built along rivers, driven by waterwheels. When steam engines appeared, factory owners initially just replaced waterwheels with steam engines, keeping everything else unchanged—productivity gains were limited.
The real breakthrough came when owners realized they could break free from rivers entirely. They built larger factories near workers, ports, and raw materials, redesigning layouts around steam engines. Later, with electricity, they further decentralized by placing small motors on individual machines instead of relying on a central power shaft. Productivity exploded, marking the true rise of the Second Industrial Revolution.
Thomas Allom's 1835 engraving depicting a steam-powered textile mill in Lancashire, England.
We’re still in the “replacing waterwheels” phase. By forcing AI chatbots into workflows designed for humans, we haven’t yet reimagined what organizations could become when old constraints vanish—when companies can run on infinite intelligence that works while you sleep.
At my company Notion, we’ve been experimenting. Beyond our 1,000 employees, over 700 AI assistants now handle repetitive tasks: recording meetings, answering questions to consolidate team knowledge, processing IT requests, capturing customer feedback, helping new hires navigate benefits, writing weekly status reports to eliminate copy-pasting… This is just the beginning. The true potential is limited only by our imagination and inertia.
Economy: From Florence to Megacities
Steel and steam didn’t just transform buildings and factories—they reshaped cities.
Until a few centuries ago, cities were human-scaled. You could walk across Florence in forty minutes; life moved at the pace of footsteps and the range of the human voice.
Then steel frames made skyscrapers possible; steam-powered railways connected city centers with hinterlands; elevators, subways, and highways followed. Urban scale and density exploded—Tokyo, Chongqing, Dallas.
These aren’t just enlarged versions of Florence. They represent entirely new ways of living. Megacities are disorienting, anonymous, and hard to navigate. This “unrecognizability” is the price of scale. But they also offer more opportunities, greater freedom, and support more diverse combinations of activities involving more people—something Renaissance-era cities could never achieve.
I believe the knowledge economy is about to undergo a similar transformation.
Today, knowledge work accounts for nearly half of U.S. GDP, yet it largely operates at human scale: teams of dozens, workflows dictated by meeting and email rhythms, organizations that struggle beyond a hundred people… We’ve been building “Florence” with stone and wood.
When AI assistants are deployed at scale, we’ll build “Tokyo”—organizations composed of thousands of AI agents and humans; workflows that run continuously across time zones without waiting for someone to wake up; decisions synthesized with precisely calibrated human input.
It will be a different experience—faster, more leveraged, but initially more dizzying. The rhythms of weekly meetings, quarterly planning, and annual reviews may no longer apply. New rhythms will emerge. We’ll lose some clarity, but gain scale and speed.
Beyond the Waterwheel
Every technological material demands that people stop viewing the world through the rearview mirror and start imagining a new one. Carnegie looked at steel and saw city skylines; Lancashire factory owners looked at steam engines and envisioned factories far from rivers.
We’re still in AI’s “waterwheel phase,” jamming chatbots into human-designed workflows. We shouldn’t settle for AI as just a co-pilot. We must imagine what knowledge work becomes when human organizations are reinforced with steel, and mundane tasks are delegated to tireless intelligence.
Steel, steam, and infinite intelligence. The next skyline awaits, ready for us to build it.
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
@ivanhzhao
