Let's be honest. You've probably poured a ton of money into AI, and you're still not seeing the results you want. I get it. I've been there. I've made the same mistakes. I've seen it in my own companies, and I've seen it in the 200+ startups I've invested in.
Everyone is so focused on the tech—the latest models, the fanciest algorithms. But they're missing the most important piece of the puzzle: the people. The culture.
A recent study shows that a staggering 43% of companies are failing at AI. And it's not because they're using the wrong tools. It's because they haven't built a culture that can actually use them.
My First Big AI Failure
I remember it like it was yesterday. At MovieLaLa, we were trying to build an AI-powered recommendation engine. We had a team of brilliant engineers, the best data scientists, and a mountain of data. We spent months, and a small fortune, building what we thought was the perfect system.
And it was a complete flop.
Our users hated it. The recommendations were generic, and the whole thing felt…creepy. We had to scrap the entire project. It was a painful and expensive lesson.
I was so focused on the technology that I forgot about the humans on the other end of the screen. I didn't think about how they would interact with it, how they would feel about it. I just assumed that because the tech was cool, people would love it.
I was wrong.
The Culture Shift: From Tech-First to People-First
That failure was a wake-up call. I realized that AI isn't just about code and algorithms. It's about people. It's about building a culture where AI is a tool to empower your team, not replace them.
So, I changed my approach. I started talking to my team. I asked them what they needed, what they were afraid of, and what they were excited about. I started to think about AI not as a product, but as a partner.
And that's when everything changed.
AI in Onboarding: Your First Impression Matters
Your company's culture is set from day one. That's why it's so important to introduce AI into your onboarding process. But not in a creepy, Big Brother way.
At RemoteTeam, which was later acquired by Gusto, we built an AI-powered onboarding buddy. It wasn't there to watch you or grade you. It was there to help. It would answer questions, point new hires to the right resources, and even introduce them to other team members with similar interests. It was like having a friendly guide in your pocket.
The result? Our new hires felt more connected, more supported, and got up to speed faster. They weren't afraid of the AI; they saw it as a helpful tool. And that set the tone for our entire company culture.
Performance Reviews: The Most Broken Process in Business
Let's talk about something everyone hates: performance reviews. They're biased, they're infrequent, and they're usually a giant waste of time. I've always despised them. The traditional way of doing them is just awful.
So, I thought, why not use AI to fix them?
My first attempt was a disaster. I tried to build a system that would analyze an employee's work and give them a score. It was cold, it was impersonal, and my team hated it. They felt like they were being judged by a machine. And they were right.
I had made the same mistake all over again. I was so focused on the data that I forgot about the people.
A Better Way: AI as a Coach, Not a Judge
I went back to the drawing board. I realized that AI shouldn't be the one giving the review. It should be a tool to help managers and employees have better conversations.
We built a new system at one of my later ventures. This one didn't give scores. Instead, it did a few key things:
- It surfaced highlights: The AI would pull out key accomplishments from the past quarter, based on public conversations in Slack, project management tools, and code check-ins. This helped jog everyone's memory and made sure good work didn't get forgotten.
- It identified coaching opportunities: The system would spot areas where an employee might be struggling, not by looking for mistakes, but by seeing where they were asking a lot of questions or where projects were repeatedly stalled. It would then suggest resources or people who could help.
- It fought bias: The AI was trained to spot common biases in performance review language. If a manager was using loaded words or focusing too much on personality instead of performance, the system would gently nudge them to be more objective.
It wasn't a perfect system, but it was a huge improvement. The conversations were more productive, the feedback was more specific, and the team felt like the process was fairer. We turned a process everyone dreaded into something genuinely useful.
Building a True AI Culture: The Raw, Unfiltered Truth
So how do you actually build a culture where AI thrives? It's not about buying the latest tools or hiring a bunch of PhDs. It's about a fundamental shift in mindset. Here are the hard-won lessons I've learned over two decades of building companies.
1. Stop Calling It "AI"
The term "AI" is loaded. It scares people. They think of robots taking their jobs. I found that as soon as I started calling it something else, the fear went away. We called it "our smart assistant" or "our internal toolkit." The name doesn't matter. What matters is that you frame it as a tool to help people, not replace them.
2. Find the "Annoying Work" and Automate It
Don't start with some grand, moonshot AI project. Start small. Find the most tedious, repetitive, soul-crushing tasks in your company and use AI to automate them.
At one of my portfolio companies, the sales team spent hours every week manually updating the CRM. It was a nightmare. We built a simple tool that used AI to automatically capture and log their calls and emails. The sales team was ecstatic. They got hours back in their week, and they could focus on what they do best: selling. We didn t just save them time; we earned their trust. That one small win made it so much easier to roll out other AI initiatives later on.
3. Make It a Two-Way Conversation
Don't just force AI tools on your team from on high. That top-down approach is a recipe for failure. You need to create a constant feedback loop. Some of the most successful AI applications we built came directly from ideas from junior employees who were deep in the weeds of a problem.
Create a dedicated Slack channel. Hold regular office hours. Ask people what's working and what's frustrating. When you give people a voice in the process, they become partners in building the solution, not just passive users. They'll tell you what they actually need, not what you think they need.
4. Celebrate the Wins, Especially the Small Ones
When an AI tool saves a team a few hours a week, that's a big deal. Celebrate it. When it helps someone close a sale or solve a customer's problem faster, make sure everyone knows about it.
This isn't just about patting people on the back. It's about building momentum. It shows the rest of the organization that this AI stuff isn't just some fancy executive pet project. It's actually making their lives better. These small, tangible wins are what build a groundswell of support and get people excited for what's next.
5. Be Radically Transparent, Especially About Failures
Your AI is going to mess up. It will have biases. It will break. It will give weird recommendations. When this happens, you have two choices: you can try to hide it, or you can be open about it.
I've seen leaders try to sweep AI failures under the rug. It's a disaster. It erodes trust and makes people suspicious of everything you do. The only way to build a real, lasting AI culture is to be radically transparent. When the AI makes a mistake, own it. Write a public post-mortem. Talk about what you learned and how you're going to fix it. This shows your team that you're not trying to pull a fast one. It builds the psychological safety needed for people to experiment and take risks.
Stop Chasing the Hype. Start Solving Problems.
The AI hype cycle is deafening. Every day there's a new model, a new breakthrough, a new technology that promises to change everything. It's easy to get distracted and feel like you're constantly falling behind.
But I'm telling you, most of it is just noise.
The real, durable advantage with AI doesn't come from having the absolute latest model. It comes from how you weave it into the fabric of your company. It's about using it to solve real, human problems. It's about making your team's work more meaningful, your customers' lives easier, and your business fundamentally smarter.
So, stop chasing the hype. Stop trying to build some perfect, all-knowing artificial brain.
Start with your people. Start with their problems. Start with your culture.
That's the secret to winning with AI. It's the lesson that cost me millions and years of frustration to learn. You can have it for free.
Frequently Asked Questions
How has this view evolved over time?
My thinking on most topics has changed significantly over the years. Early in my career, I held many conventional views that experience proved wrong. I try to update my beliefs when the evidence changes.
What experience informs this perspective?
This perspective comes from over a decade of building companies in Silicon Valley, two successful exits (RemoteTeam to Gusto, MovieLaLa to Gfycat), and investing in 200+ startups including Anthropic, OpenAI, and Scale AI. I write about what I've lived.
What's the most common pushback you get on this?
People often push back by citing exceptions or edge cases. And they're usually right that exceptions exist. But building a strategy around exceptions rather than patterns is a losing game for most founders.
Do all experts agree with this view?
No, and that's fine. The best ideas in business are often contrarian. I share my perspective based on my experience and data, but I encourage you to seek out opposing viewpoints and form your own conclusions.