Chapter 1
Five stages of getting results with AI
Understand what to prepare, in what order, to achieve results with AI
The CEO tells everyone to start working with AI. Weeks later, instructions, progress checks and meetings still fill the calendar, leaving the CEO exhausted late at night.
The same problem appeared in Chapter 0: the company adopted AI, but work did not become easier. In this chapter, you will identify why and learn what to put in place, in what order, to produce results.
Are you getting results with AI?
Ask people whether they use AI and you tend to hear two answers.
Some say, “Every day. I try the latest models and improve my prompts.” Others say, “Honestly, hardly at all. I feel behind, and it worries me.”
Ask the daily users a second question: What results has that use of AI produced? Higher revenue, lower costs, more orders: changes they can describe with numbers. Many struggle to answer.
Daily users and beginners often face the same problem. Knowing more about AI has not translated into better results. The order of learning matters. Many AI courses start with how to use AI and leave out how work succeeds. This course starts with how to organize work. Beginners can learn the approach directly, without first having to change established habits.
Does producing 1,000 documents mean AI is doing useful work?
Consider an extreme example. An AI produces all of the following overnight:
- 200 market research reports
- 1,000 proposals
- 5,000 social media posts
- 100 automation workflows
Has the AI done an exceptional job?
The CEO is impressed by the thousand documents AI produced overnight. Days later, the documents remain unused in a corner of the meeting room.
What if nobody uses the documents? No decisions change. No customer problem is solved. The company's numbers stay the same. Nothing has changed. What did AI actually produce?
It produced deliverables. It did not produce results.
- A deliverable is something you make: a document, report, post or workflow.
- A result is a change from the current state to a desired state.
A proposal is a deliverable. A customer deciding to buy after reading it is a result.
The distinction helps explain familiar claims: “We used AI.” “We got faster.” “We automated it.” “We built an impressive tool.” They sound like reports of results, but they describe means. When means are mistaken for results, people can get busier without accomplishing more.
The purpose of using AI is to achieve results with AI.
AI can work much faster than a person working unaided. People still need to decide what result to pursue. Learning how to operate AI without a clear objective can simply produce unnecessary work more quickly.
AI will respond to a request. Yet you may need to explain the full background every time, and a long explanation can still produce something different from what you intended. Unnecessary deliverables accumulate. Each employee's AI knows different things, and that knowledge is not shared. The missing information is consistent: what the company wants to achieve, its Goal, and how far the work has progressed, its current state. Those two facts were never placed where AI could read them whenever needed. Make them accessible, and repeated explanations and misunderstandings can decrease.
To decide which parts of work AI can do, first understand what work consists of. Chapter 2 explains the structure.
Five stages of achieving better results with AI
Keep the objective clear: use AI to achieve substantially better results. Reaching that state takes preparation. The course follows five stages.
- Understand how to work effectively (Chapter 2) — Study work itself, before introducing AI. Learn the structure that makes efficient work with AI possible.
- Delegate work to AI (Chapter 3) — Learn what AI needs to do a job, then begin delegating simple work.
- Create an environment where AI can work (Chapter 4) — Enable AI to gather information independently and produce the output you need.
- Automate work and let AI manage decisions (Chapter 5) — Progress from routine automation to autonomous work, including more complex activities such as writing articles.
- Build a shared workspace for people and AI (Chapter 6) — Help multiple people and AI agents achieve Goals efficiently, with information sharing, requests and feedback handled automatically through AI.
Review
Three ideas matter most:
- A result is a change toward a desired state. A deliverable is the thing you made.
- The purpose of using AI is to achieve results with AI.
- People decide what AI should work toward. Greater speed is useful only when the objective is clear.
What changes in practice
Our own company provides examples of the changes that follow these stages.
The course starts from two premises: AI can do the work that can be done on a computer, and asking AI can resolve most questions that arise. You can continue even if AI is unfamiliar.
Begin with the first stage
Start with the first of the five stages: understand how to work effectively.
The next chapter barely mentions AI. It focuses on work: what a Goal is, how to assess the current state, and how to choose the next action.
The reason is straightforward. Without a clear objective, faster execution can increase the amount of misdirected work. The next chapter helps you define that objective yourself.