In partnership with
Hello! 👋
It's Thursday, 11th June 2026. Hello and welcome back to Bold Efforts! The FIFA WC kicks off today and it is an event of massive scale.
For most of history, scale belonged to organizations. If you wanted to build something meaningful, you needed people, capital, equipment, distribution, and time. A factory needed machines. A media company needed writers and editors. A software company needed engineers, designers, product managers, QA, infrastructure, support, and all the coordination that comes with it. Small teams could be creative, but large organizations had the machinery to turn ideas into output.
AI is beginning to weaken that advantage. Not because it replaces every worker, and not because every task becomes automated overnight. The deeper shift is that individuals and small teams can now build systems around themselves. They can research, prototype, write, analyze, test, automate, and iterate with a level of leverage that used to require a department.
This is why the usual productivity debate feels too small. The question is not simply whether AI helps someone write an email faster, summarize a meeting, or produce a first draft. Those are useful improvements, but they are not the main story. The better question is what kind of company becomes possible when cognitive work becomes cheap, abundant, and programmable.
The most important change is that the worker is slowly becoming the designer of the factory. In the old model, most knowledge work was direct production. You wrote the memo. You built the spreadsheet. You created the presentation. You searched for the information. You wrote the code. You reviewed the document. You sent the follow-up. The person produced the output by doing the task.
In the new model, the best people will increasingly design the system that produces the output. A good engineer will not only build one feature. She will build the machinery that helps generate, test, and improve many features. A good marketer will not only write one campaign. He will build a system that studies customers, tests angles, repurposes ideas, measures response, and improves messaging over time. A good recruiter will not only search for candidates. She will design a workflow that finds people, filters relevance, checks freshness, drafts outreach, tracks replies, and learns from outcomes.
This is the move from task execution to factory design. It sounds abstract, but in practice it is simple. Every repeated workflow now deserves a question: can this become a script, agent, template, prompt, workflow, or internal tool? If the answer is yes, the person who builds that system gains leverage. The person who keeps repeating the task by hand may not feel the disadvantage immediately, but it will compound.
The reason this matters is that the cost of intelligence is collapsing. For decades, organizations treated human time as the scarce input. Good people were expensive. Expertise was hard to find. Analysis took time. Writing took time. Research took time. Software took time. AI compresses many of these costs. It does not make judgment cheap, but it makes the first draft of thinking cheap. Exploration becomes cheap. Iteration becomes cheap. Variations become cheap. Prototypes become cheap.
When something is expensive, people ration it. When something becomes cheap, people use more of it. When computing became cheap, software expanded everywhere. When distribution became cheap, media exploded. When coordination became cheap, platforms scaled. Now cognitive work is getting cheaper, which means the winners will not be the people who use the least AI. The winners will be the people who learn how to spend more machine intelligence to save human attention.
That requires a mental shift. Many people still treat AI like a careful assistant. Ask one question, get one answer, edit the answer, move on. That is underusing it. AI should often be used more like industrial energy. Run more attempts. Ask for alternatives. Stress-test assumptions. Generate counterarguments. Build rough versions. Check edge cases. Compare strategies. Rewrite from different angles. Simulate customer reactions. The point is not to make AI sound smart. The point is to use it until the human decision gets better.
This is also where the hard part begins, because if AI makes production easier, the new bottleneck becomes taste. Judgment becomes more valuable, not less. Direction becomes more valuable. Knowing what good looks like becomes more valuable. So does knowing which problem matters, when the output is wrong, what to ignore, and when to stop.
This is the uncomfortable part of the AI revolution. It gives people more leverage, but it also exposes weak thinking faster. If someone has poor judgment, AI gives them more ways to produce mediocre work. If someone has strong judgment, AI gives them more ways to explore, refine, and ship. The gap widens because leverage multiplies the quality of the person using it.
That is why the “AI will replace everyone” framing feels incomplete. AI will replace some tasks. It will pressure some roles. It will create new expectations. But in many fields, the bigger gap will be between people who can direct AI well and people who cannot. The future knowledge worker may look less like a pure producer and more like a director. The director does not act every role, design every set, edit every frame, and distribute the film alone. The director decides what belongs, what fails, what must be changed, and what the final thing should feel like.
Companies will face the same shift. For years, software was the obvious source of leverage. If you could build software, you could scale. A product could serve millions of users with a relatively small team. But AI makes basic software easier to create, which means software alone becomes less defensible. Software still matters, but the moat moves to the harder layers around it.
The durable advantage moves to proprietary data, distribution, trust, workflow ownership, domain expertise, speed of learning, and operating discipline. A company’s advantage will increasingly come from how well it combines AI with a specific market, a specific workflow, and a specific customer pain. This is why applied AI matters so much. The foundation models are powerful, but the business value sits inside messy real-world contexts: hiring, healthcare, construction, real estate, logistics, finance, education, legal operations, customer support, and government services.
These industries do not need prettier chatbots. They need better systems. They need clean data, reliable workflows, human review loops, compliance, integrations, accountability, and trust. The shallow version of AI says “We added AI”. The serious version says “We redesigned the workflow around AI”. That is the line to watch.
The small team also gets a new ceiling. A five-person company can now look much larger than it is. A solo founder can build prototypes that previously required a product team. A consultant can operate with the leverage of a boutique firm. A writer can run research, editing, distribution, and productization in parallel. This does not make success easy. It makes the attempt cheaper. That distinction matters because cheaper attempts mean more experiments, more competition, and more surprising companies from places the market is not watching.
For the last few years, the future-of-work conversation has been trapped in a narrow frame: remote versus office, hybrid versus flexible, productivity versus culture. These debates still matter, but they miss the larger shift. The future of work is about leverage. How much can one person create? How quickly can a small team learn? How much repetitive work can be turned into systems? How much judgment can be preserved while production becomes automated?
This is where the next generation of companies will separate themselves. The best companies will not simply give employees AI tools and hope productivity improves. They will redesign roles, workflows, metrics, and culture around this new leverage. They will reward people who improve systems, not only people who complete tasks. They will treat automation as an operating habit, not an IT project. They will build verification into every AI workflow because they will understand that the goal is not to remove humans from work. The goal is to move humans to the highest-value layer of work: direction, taste, trust, judgment, and accountability.
The industrial revolution gave us factories. The internet gave us networks. AI gives us personal factories. A person with strong judgment, clear taste, and the ability to build systems around themselves can now do far more than before.
That is exciting, but it is also demanding. The excuse of “I don’t have a team” becomes weaker. The excuse of “I don’t know how to start” becomes weaker. The excuse of “this would take too long” becomes weaker. AI does not remove the need for ambition, discipline, or taste. It simply gives those qualities more surface area.
The people who benefit most will not be the ones who ask AI to do their thinking. They will be the ones who use AI to multiply their thinking. The future belongs to people who can build the factory, direct the factory, and keep improving what the factory produces.
The factory is no longer only a building full of machines. Increasingly, the factory is a person with judgment, a laptop, and a system. Thank you for reading
Best,
Kartik
I write Bold Efforts every week to think clearly about where work and life are actually headed. If you want these essays in your inbox, you can subscribe here.