Your Strange New Employee
The article below describes the two things your organization’s AI deployment needs before it can deliver real results: deep context on your business and the initiative to act on that context without being asked.
If you could use help setting up AI that works this way rather than tossing away tokens and hoping for the best, we built ChatCEO for you. CEOs who lead a company of 50+ employees are eligible to become founding users and get one year of ChatCEO free.
Imagine you just recruited a most unusual new employee.
He graduated top of his class, with separate degrees in computer science, finance, marketing, operations, HR, law, accounting, and a dozen other disciplines. He’s a true prodigy. He can write code, build financial models, draft legal briefs, run a marketing campaign, and analyze your supply chain. And he can do it all in the same afternoon.
You’re thrilled to have him on the team, aren’t you?
Well, before you give him a raise, realize that he also has a couple of drawbacks. See if these change your mind about him.
First, he struggles to retain the basic details of your business. Those he does learn, he forgets quickly. Despite his vast skill set and oceanic knowledge base, he needs to be reminded frequently what his role is. And he doesn’t hang around the watercooler, so he doesn’t pick up the company-specific cultural and institutional intelligence your other employees do.
This unusual employee has another weak point: He has zero initiative. If you ask him to do something specific, he’ll knock it out of the park. But you have to poke him every time you need him to act. You can bounce ideas off him, but he’ll never come to you with one of his own.
How do you feel about this employee now?
As you may have guessed, this employee’s name is AI, and he’s already at work in your company. Given those two big weaknesses, is it any wonder that lots of leaders are still using AI as little more than a glorified search engine?
If you want to turn your strange-but-brilliant new hire into an actually productive contributor to the business, you must ensure it has the context and the initiative you would expect from a human employee.
Providing the Context
How does any new employee gain enough context to be successful in your organization?
Typically, you put them to work and let the organization do the teaching. Over weeks and months, the employee absorbs things that were never written down anywhere:
How decisions actually get made (vs. the power structure reflected in the org chart)
Which customers are the most critical (vs. merely being the largest or loudest)
What the CEO really cares about (vs. what they say in all-hands meetings)
What behaviors are rewarded by the company culture (and which aren’t)
This context-building occurs mainly through osmosis. The employee watches and learns. Everyone they interact with in the organization teaches them something new. The accumulated context of your specific business is what transforms general competence into organizational effectiveness.
Unfortunately, AI can’t absorb context by osmosis. It can’t overhear the hallway conversation, pick up on body language in a meeting, or intuit from a CEO’s reaction that a certain topic is off-limits.
So, your first step is to ensure that AI at your company understands the context of your business. That means providing it with foundational documents that clarify:
The company’s mission, vision, values, strategy, and founding purpose
Its competitive positioning and primary rivals/threats
Its cultural priorities and non-negotiables
Its brand and all internal lingo and definitions (what does a “good deal” mean, for example?)
Personality types and behavioral styles of all employees it works with
This means writing down many things that usually wouldn’t be written down, the stuff you would expect a human employee to pick up naturally. You might just find that this practice clarifies some things in your mind, too. Think of it as onboarding for one of the most consequential hires you’ll ever make.
After onboarding, this context must become a living intelligence layer. As brilliant as your AI system is, it’s not going to be great at picking up the nuances that matter. You have to be explicit and intentionally steer it with the latest context.
Sparking Initiative
The second drawback in your strange new employee is passivity. AI is anything but a self-starter. Your best employees do more than complete the assignments you hand them. Ideally, they see broken things and fix them. They think about your business throughout the day, and when they have an idea, they bring it to you. Or they simply implement it on their own.
AI, by contrast, waits to be asked. It executes, but it doesn’t originate, at least not without thoughtful, systematic design.
Like the context problem, this one is solvable. Instead of treating AI as a faster answer machine, you can architect workflows that prompt the AI to act on its own. For example, you can configure it to:
Scan the competitive environment on a regular basis, reporting when it finds something novel
Proactively surface at-risk goals inside the company
Spot consequential engagements on your calendar on its own, and present a brief and agenda 24 hours before it happens
As the models’ memories grow (as with the newly released Fable 5), the capacity to mimic initiative will grow. Your AI system must above all understand the operating rhythm of your business and the types of insights you value - that’s a part of giving it context. Only then can it do more than run off and complete the individual tasks you give it.
Will you ever train AI to have the drive of a superstar employee? Not likely. But when you combine AI’s inbuilt proficiency with a rhythm of proactive operations, you work around its inherent laziness.
The Operating Loop
Once these proactive workflows are in place, the system surfaces an insight, you act on it, your action generates new data, and the system gets smarter as a result. The value compounds week over week. The AI you’re working with six months from now will understand your business in a way no freshly installed tool ever could.
This is also why context and initiative, which look like two separate configuration projects, turn out to flow from the same source. When your operating rhythm is structured well, both emerge from it naturally. You run your week, and the system learns from the rhythm itself.
The companies getting real value from AI today have one thing in common: They built the operating rhythm to put those tools to work. Ultimately, the question worth asking is less “What can AI do?” and more “What do I need to tell AI, and how do I need to structure its role, for it to add real strategic value?”
When you get this right, AI is no longer the omni-talented, highly credentialed employee whose results never live up to the hype. Instead, you get a tool that meaningfully contributes to the highest goals of the business.