skip to main content
leadership & mindset

The 11 Universal Laws in the Age of AI

By Kalika Yap · October 22, 2026 · 6 min read

There’s an idea, often credited to futurist Roy Amara of the Institute for the Future, that every founder should hear right now:

“We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.”

That’s AI in one sentence.

In the short run, we expect a new tool to save us by Friday. In the long run, it changes how our whole business works.

Last week, we walked through 11 universal laws, from Parkinson’s Law to Brooks’ Law. The natural next question is this: how do they work now that AI is everywhere?

So this week is Part 2. Same 11 laws, now with AI in the room.


What I’m Learning From Building My Own AI Team

Right now, I’m building what I call an AI workforce across all four of my businesses.

It’s a tiered system of AI agents, each with a clear job. Some of the ideas came from a workshop with Allie K. Miller and Mark Cuban, and I’ve been adapting them ever since.

And one of my biggest goals right now is using AI to help build 1 million websites for small business owners.

Every small business owner deserves a website that works as hard as they do. AI makes it possible to help them at a scale that once felt out of reach.

A goal that big makes every one of these laws matter. Chesterton’s fence reminds us to understand each step a good designer takes before we hand it to AI. Goodhart’s Law reminds us to count happy business owners, along with websites launched.

Here’s what surprised me most. The tools are brand new, and the human patterns are as old as ever.

Projects still ran long. Too many tool choices still slowed us down. And the most useful question was still the oldest one: why do we do it this way?


Where I See Founders Get Stuck With AI

Many of the founders I coach fall into one of two camps.

The first camp signs up for every new tool, tries each one for a week, and ends up with a dozen logins and no real change in how the business runs.

The second camp feels so overwhelmed that they wait. I’ll figure out AI once things settle down.

Both camps are smart, capable people. What helps both camps is a simple way to decide. That’s exactly what the 11 laws give you.


Fast Tools, Same Laws

Here’s the phrase I’d love you to carry into this week. Fast tools, same laws.

AI can write the draft, summarize the meeting, and sort the inbox in seconds.

People, teams, and projects still behave the way they always have. When you plan with the laws, the speed actually helps.


The 11 Laws, With AI in the Room

1. Parkinson’s Law: reset your deadlines

If AI drafts a proposal in an hour and the deadline is still a week, the saved time fills up with extra revisions.

Try this: When a tool speeds up a task, shrink the deadline to match.

2. Hofstadter’s Law: AI projects run long too

The tool itself is fast. Cleaning up your data, writing clear instructions, training your team, and sorting out permissions take real time.

Try this: Start with one small pilot, and pad the timeline.

3. Hanlon’s Razor: meet hesitation with kindness

When a team member hesitates about AI, it’s usually unfamiliarity or worry about their role. And when AI makes a mistake, it’s a mistake to catch and correct.

Try this: Respond with training, reassurance, and a simple review step.

4. The Pareto Principle: automate your vital 20% first

Look for the repetitive tasks that eat the most hours. Follow-up emails, meeting notes, first drafts, and weekly reports are common places to start.

Try this: Start there, and leave everything else alone for now.

5. The Peter Principle: roles are changing

Many people were promoted for doing tasks AI can now help with. The skills that matter more each month are judgment, client care, and checking AI’s work.

Try this: Train your team for the job each role is becoming.

6. Hick’s Law: choose a small AI stack

There’s a new AI tool every week, and trying them all slows every decision you make.

Try this: Choose two or three tools that cover most of your needs. When you prompt, ask for three options instead of twenty.

7. Goodhart’s Law: measure results

If you track “posts made with AI” or “how many people use the tool,” you’ll get more of both and possibly nothing else.

Try this: Track the outcomes you care about, like hours saved, leads, revenue, and client happiness.

8. The Dunning-Kruger Effect: keep an expert in the loop

A few good prompts can make anyone feel like an expert, and AI sounds confident even when it’s wrong. You need real knowledge to spot the errors.

Try this: Have someone who knows the subject review anything legal, financial, or client-facing.

9. Occam’s Razor: try the simple fix first

Before you buy a big AI platform, see whether a template, a checklist, or one well-written prompt solves the problem.

Try this: Test the simplest version for a week before you spend more.

10. Chesterton’s Fence: understand the step before you automate it

That approval step that feels slow might be the one that catches pricing errors or legal problems.

Try this: Ask why each step exists before you hand it to AI. Then automate with confidence.

11. Brooks’ Law: bring AI in early

Adding new AI tools to a late project works a lot like adding new people. Someone has to set them up, test them, and coordinate them.

Try this: Introduce AI when there’s time to learn it, well before the crunch.


Your Step-by-Step Guide: Bring AI Into Your Business Wisely This Week

Step 1. List your repeat tasks

Spend 20 minutes writing down every task you or your team repeat each week. Next to each one, estimate the hours it takes.

Circle the three that take the most time.

Your first AI win is usually hiding in your most repetitive work.

Step 2. Walk the fence

For each circled task, write down why it exists and what could go wrong if a step disappeared.

Keep the steps that protect your clients, your money, or your reputation. Those stay human, or get a human review.

Know why a step is there before you hand it to AI.

Step 3. Choose one tool and one pilot

Pick one task and one tool. If you already pay for a tool that can do it, start there.

Give the pilot two weeks, and add a buffer for setup and learning.

One tool, one task, one pilot.

Step 4. Decide what success looks like

Before you start, write down the result you want. Hours saved per week, faster replies to clients, or fewer errors are all good measures.

Name one person who’ll review the AI’s work.

Measure the result you want, and keep an expert in the loop.

Step 5. Start your pilot within 48 hours

Put the first session on your calendar in the next 48 hours.

Tell your team what you’re testing, why, and what’s staying the same for them.

The best AI plan is the one you start this week.


Want Help Building AI Into Your Business?

Sometimes it’s hard to know where AI fits when you’re inside the business every day. That’s a big part of what I do with my coaching clients. We find the tasks worth automating, the ones worth protecting, and a plan your team can actually follow.

Book a discovery call

No pitch. Just a real conversation about where AI can save you time and where you can grow.

Start small, measure what matters, and let the laws guide you. You’re doing just great.

DOWNLOADABLE WORKSHEET

Your Ai Pilot Plan

Your AI Pilot Plan

This worksheet helps you identify repetitive tasks, choose one AI tool to test, and create a two-week pilot with clear goals. Start small, measure what matters, and keep human judgment in the process.

Get the worksheet

your next step

downloadable worksheet

Your Ai Pilot Plan

One page to put this post to work. Tell us where to send it and it’s yours.

free toolthe brand story workbookWork through your why, vision, mission, and values, and write a brand story your team can use.get the workbook

interested in collaborating?

Or maybe you just have a couple questions? Say hello, we’d love to hear from you!

support@citrusstudios.com(310) 717-8201

A smiling woman in an orange sweater working on a laptop