I once sat in a boardroom where a CEO, sweating under the pressure of his board, approved a $50 million plan to "integrate AI across the enterprise." Twelve months and $20 million later, all they had was a buggy chatbot and a team of demoralized engineers. The project was quietly killed. Around the same time, I watched a small team at one of my portfolio companies use an open-source model to automate a single, tedious data entry task. It took them two weeks and cost less than a new server. That tiny project saved them 20 hours a week, and the momentum it created completely reshaped their product roadmap.
What was the difference? The first company followed the official playbook. The second followed the unspoken rules.
I’ve seen this game from every angle. As a founder who built and sold two tech companies, RemoteTeam and MovieLaLa. As an angel investor in over 200 startups, including foundational AI companies like Anthropic, OpenAI, and Scale AI. I’ve seen the hype, the fear, the catastrophic failures, and the quiet, game-changing wins. The consultants and the airport business books will sell you a clean, five-step plan for AI transformation. It’s all nonsense.
Change isn’t a checklist. Especially not change driven by a technology as potent as artificial intelligence. There are rules to this game that no one talks about publicly, the messy, human truths that actually determine success or failure. I’m breaking the silence.
Rule #1: Your “Why” is Probably Wrong
The first question I ask any CEO who pitches me their AI strategy is: “Why are you doing this?” The most common answers are the worst: “Our competitors are doing it.” “The board is asking about it.” “We need to be seen as an innovator.”
This is just fear masquerading as strategy. Adopting AI because you’re afraid of being left behind is like learning to swim by jumping into a hurricane. You’ll just drown faster.
A real “why” is specific, selfish, and tied to a painful problem. It’s not “We want to leverage AI to optimize synergy.” It’s “Our customer support tickets are taking 48 hours to resolve, and it’s killing our retention rate.”
At RemoteTeam, we didn’t have a grand “AI strategy.” We had a payroll processing problem. Our team was spending hundreds of hours a month manually verifying international payroll data, and it was a nightmare of spreadsheets and time zones. Our “why” was simple: kill the spreadsheets. We focused a small team on using AI to automate the verification of that specific data. It worked. It cut our processing time by over 70% and freed up our team to handle actual customer issues. That single, focused win gave us the confidence and the blueprint to find the next problem, and the next.
Don’t start with the tech. Start with the pain. Find the most expensive, time-consuming, soul-crushing bottleneck in your business and aim AI directly at it.
Rule #2: Stop Obsessing Over the Tech
This might sound strange coming from someone who invests heavily in deep tech, but the technology is rarely the hardest part of an AI transition. You can buy the models, you can hire the PhDs, you can rent the GPUs. The real challenge is the people.
Your employees aren’t stupid. They read the headlines. They see the demos. And they are terrified. They’re worried about their jobs, their skills becoming obsolete, and their value disappearing. When you announce an “AI initiative,” they don’t hear “exciting new capability.” They hear “your replacement is coming.”
Ignoring this fear is managerial malpractice. The single biggest obstacle to AI adoption is internal resistance, both active and passive. It’s the manager who subtly sabotages the new system because it threatens their domain. It’s the team that keeps using the old, inefficient process because it’s familiar and safe. You can’t solve a human problem with a technical solution.
Your job as a leader isn’t to pick the right algorithm. It’s to create psychological safety. You have to address the fear head-on. Be brutally honest about how roles will change. More importantly, show them a path forward. Invest in retraining before you deploy the new tools. I’ve seen companies offer to pay for AI-related certifications for their entire staff. The cost is trivial compared to the cost of a failed implementation and a culture of fear.
Rule #3: The “Chief AI Officer” is a Trap
Here’s a controversial take: appointing a Chief AI Officer is often a huge mistake. On the surface, it looks like decisive leadership. You’re creating accountability! You’re centralizing expertise! In reality, you’re creating a silo.
AI is not a department. It’s a foundational layer, like the internet was 25 years ago. You don’t have a “Chief Internet Officer.” You expect every leader in your business—from marketing to finance to product—to understand how to use the internet to do their job better. AI is the same.
When you crown a CAIO, you give everyone else an excuse to not take ownership. AI becomes “that person’s problem.” Initiatives get stuck in a centralized queue, disconnected from the business units that actually understand the problems. I advised a Fortune 500 company where the CAIO had a brilliant team and a massive budget, but they couldn’t get anything done. Every project was a political battle to get on his roadmap. The real progress was happening in the shadows, where a product manager in the logistics division was using a simple AI tool to optimize their warehouse routes, without ever talking to the CAIO’s office.
Instead of a CAIO, create a culture of distributed AI literacy. Empower your product managers, your engineers, and your business analysts. Give them the tools, the data, and the permission to find and solve problems within their own domains. The leader’s job is to set the guardrails—on ethics, security, and budget—not to be the central gatekeeper of all innovation.
Rule #4: Small Wins, Big Momentum
The $50 million “enterprise-wide transformation” is doomed from the start. It’s too big, too slow, and too abstract. By the time you’re ready to show results, the technology has changed, the business priorities have shifted, and the budget has been cut.
Momentum is the most valuable currency in change management. The only way to create it is with quick, tangible wins. Forget boiling the ocean. Find a puddle and boil that.
At MovieLaLa, my second startup, we were competing with giants. We didn’t have the resources for a massive recommendation engine overhaul. So we picked a tiny target: the “you might also like” feature on our movie pages. We used a simple collaborative filtering algorithm—nothing fancy—to see if we could lift engagement by just a few percent. It took one engineer about three weeks. The result was a 5% increase in click-through on those recommendations. It wasn’t earth-shattering, but it was real. It was a win.
That small victory did more than a hundred PowerPoint slides ever could. It proved to the team that this AI stuff actually worked. It gave us the confidence to tackle a slightly bigger problem. And it gave me the data I needed to convince our investors to fund a more ambitious AI-driven discovery feature. We built our AI strategy one small, undeniable win at a time.
Rule #5: Transparency is Your Only Currency
Your team is not an adversary to be managed. They are adults. They know that AI will change things, and trying to sugarcoat it is insulting. The only way to lead through this transition is with radical transparency.
This means being honest about the good, the bad, and the ugly. Which jobs will be automated away? What new skills will be required? What is the company’s plan to support people through the transition? If you don’t have all the answers, say so. The uncertainty of the truth is always better than the false confidence of a lie.
Share the data. All of it. If you’re implementing an AI to improve efficiency, show the dashboard with the efficiency metrics to the whole company. Explain what the goals are and track progress in public. When your team can see the same data you see, they can become partners in solving the problem rather than victims of the solution.
I remember holding a town hall to announce a major automation project. I put up a slide that listed the three job roles that would be eliminated in six months. The room went silent. Then I put up the next slide, which showed the five new, higher-paying roles we were creating, and the fully-funded training path to get there. It was a tough conversation, but it was an honest one. We lost some people, but the ones who stayed did so with a sense of trust and purpose.
The Real Work Begins Now
Forget the utopian promises and the dystopian fears. The reality of AI integration is much more mundane and much more difficult. It’s not about one big, glorious revolution. It’s about a thousand small, difficult, human-centric decisions.
This transition will be messy. It will be uncomfortable. You will face resistance. But the alternative is to stand still and watch the world move past you. The companies that win in this new era won’t be the ones with the fanciest algorithms. They will be the ones who understand that change management is not a project plan. It’s the hard, daily work of building trust, creating clarity, and empowering your people to navigate an uncertain future. Now, go get started.
Frequently Asked Questions
How often is this guide updated?
I revisit and update my guides regularly as I learn new things and as the market evolves. The core principles tend to stay stable, but specific tactics and tools get refreshed based on what's working right now.
Who is this guide designed for?
This guide is written for founders and operators who want practical, actionable advice rather than theoretical frameworks. Whether you're just starting out or scaling an existing business, the principles here apply across stages.
How should I work through this guide?
Don't try to absorb everything in one sitting. Read through once to get the big picture, then go back and work through each section as it becomes relevant to your current challenges. Bookmark it and return to it regularly.
What if I disagree with some of the advice?
Good. That means you're thinking critically, which is exactly what a good founder should do. Take what resonates, test it, and discard what doesn't work for your specific situation. No advice is universal.