01
The new knowledge worker
For a long time, scale meant adding people. More customers required more salespeople, more service staff, more operators, more managers, and more layers between a decision and the work.
Models and agents are changing that relationship. One capable person can engineer systems that handle work previously distributed across several roles. The person is not producing more by working longer. The computer is performing a greater share of the work.
A new kind of knowledge worker is taking shape inside this generation. This person combines professional expertise, systems thinking, and a love of working with software. They understand a field deeply enough to identify the work, structure it, and engineer systems that perform it. The work still depends on human judgment, reliability, and trust. Those qualities matter more when one person operates at greater scale.
People are moving toward this kind of work at different speeds.
- Access to models matters.
- Expectations at work matter.
- Interest in learning the technology matters.
Two people with the same title can now have very different productive capacity because one has learned to engineer with models and agents.
The difference is still difficult to see from outside the work. Many business leaders are experts in their industries, customers, and operations. Their days are not spent engineering systems with models and agents. Without that firsthand experience, the change can look smaller than it is.
02
Smaller companies
Companies will catch up at different times. Some already have. As more companies understand what one capable person can produce, teams will get smaller. Growing companies will add fewer people than comparable companies added in the past.
Smaller companies do not necessarily mean fewer opportunities to work. The work can spread across more small businesses, more entrepreneurs, more independent producers, and more tight groups of trusted people. A company that once needed several departments may become a collection of systems directed by a small number of people with strong judgment.
That makes trust more important, not less. One person can affect more customers, more data, and more of a company’s operations. Poor judgment reaches further through the systems. Capability matters, but capability without reliability becomes harder to absorb.
03
Trust and professional networks
Professional networks change with the same economics. Being widely known matters less when work happens inside smaller groups.
Being known by the right people for what you can produce matters more.
A smaller LinkedIn network can become more useful because the purpose is no longer broad attention. The purpose is recognition among people who understand one another’s work and trust one another to do it.
04
Venture capital
Venture capital follows the same direction.
Companies engineered using models and agents need fewer salaries and more compute. Compute is expensive, but in many forms of knowledge work it costs less than another salary and benefits. Building a company still requires capital. The amount required for each company changes when a small team can produce what once required a much larger one.
Venture dollars per company will fall. That does not mean venture capital disappears. The same pool of money can reach more entrepreneurs. Each entrepreneur can build more before raising, hire later, and operate longer with a smaller team.
Investors face the same gap in firsthand experience as business leaders. A person who has not engineered systems with models and agents sees the outputs without seeing the new productive capacity behind them. The economics become easier to understand once one person builds something that previously required several specialties.
05
Business-owned systems
Software companies also change in this direction. A company built around one narrow business problem will increasingly compete with systems built around the client’s business. The value moves from access to one function toward control of the data, the workflow, and the system performing the work.
Businesses are moving toward owning more of their data and more of the systems that run their operations. Over time, small, nontechnical businesses can own software built around how they work. Their operations no longer have to pass through a collection of separate services.
Outside expertise remains part of that arrangement. Engineers can design, build, and maintain the systems. The business owns the data, understands the work, and controls the system. A dozen disconnected subscriptions become one connected system built around the company using it.