You're probably using AI all wrong. I was too. Here's the painful story of how I learned to do it right.
It was 2022. RemoteTeam was growing fast, post-acquisition by Gusto. We were onboarding engineers, marketers, and support staff across a dozen time zones. I thought I was ahead of the curve. We had the best AI tools, the slickest integrations, and a mandate from the top to be an "AI-first" company.
Six months later, it was a mess. Productivity was flat. Morale was dipping. Our best people were getting frustrated. The expensive AI tools were gathering digital dust. We had all the tech but none of the culture. We were on track to becoming another statistic.
And it's a big statistic. Recent data shows that a staggering 47% of companies fail to build a successful AI culture. They spend millions on software, but they fail on the human element. They buy the tools, but they don't build the team.
I’ve seen this movie before. As an angel investor in over 200 companies, including foundational players like Anthropic, OpenAI, and Scale AI, I get a front-row seat to what works and what doesn't. I’ve managed over $2M in remote team payroll. I’ve seen the hype cycles and the painful reality checks. The problem is almost never the technology. It's the culture.
The All-Too-Common Mistake
Most leaders think throwing a new AI tool at their team is the answer. They see a slick demo, get sold on the promise of 10x productivity, and push it out with a company-wide email. “Go use this!” they say. “It will make your life easier.”
That’s not a strategy. It’s a prayer.
What happens next is predictable. A few early adopters play with it. The majority of the team, already busy with their actual jobs, ignore it. They see it as another task on their to-do list, not a tool to make their work better. There's no guidance, no shared understanding of why they should use it, and no integration into their actual workflows.
This is how you end up in the 47%. You create a graveyard of expensive, unused software licenses. More importantly, you create cynicism. The next time you introduce a tool, the team will just roll their eyes. You've burned their trust.
My Painful Lesson with RemoteTeam
After we were acquired by Gusto, the pressure was on to integrate and scale. I made the classic mistake. I was so focused on the “what”—the AI tools for coding, for content, for analytics—that I completely ignored the “how.”
I bought a cutting-edge AI-powered project management tool. It was beautiful. It promised to automate status updates, predict project risks, and even write meeting summaries. I was in love. I announced it in a team-wide meeting, shared the login details, and expected magic.
Of course, the magic never came. Our engineers, who lived in Jira and GitHub, saw it as a redundant chore. Our designers, who thrived on visual collaboration in Figma, found it restrictive. It was a tool built for a generic “team,” not our team. I had failed to understand their world. I was trying to force a top-down solution onto a bottom-up problem.
It was a humbling and expensive lesson. We had to scrap the tool, but the real cost was the time and trust we lost. It forced me to stop and rethink everything I thought I knew about building a culture, especially in a remote, AI-driven world.
Building an AI Culture That Actually Works
So how do you get it right? It’s not about a single tool or a magic bullet. It’s a systematic approach to integrating AI as a partner, not just a piece of software. It’s about building a culture of curiosity, experimentation, and shared learning.
Here are the pillars that I've found to be essential.
1. AI Onboarding is Non-Negotiable
You can't just give people a login and expect them to figure it out. You need to make AI a core part of your onboarding process from day one.
- Show, Don't Tell: For a new engineer, this means showing them how your team uses GitHub Copilot to write better code, faster. Pair them with a senior dev who can walk them through real-world examples. Show them the prompts that work, the pitfalls to avoid, and how it integrates with your code review process.
- Create AI “Playbooks”: Document your best practices. What are the top 5 prompts your sales team uses to research leads with AI? What’s the best way to use an AI writing assistant to match your brand’s tone of voice? These playbooks turn tribal knowledge into a scalable resource.
- The Buddy System: Assign every new hire an “AI buddy”—someone on their team who is already proficient and can answer those “stupid questions” people are often afraid to ask in a group setting.
2. AI Coworking: The New Collaboration
The goal is not to have humans working for AI, or even with AI. The goal is to create a collaborative environment where AI is a third member of the team. It’s a concept I call “AI Coworking.”
Think about it like this: when you’re stuck on a problem, you grab a coworker and brainstorm at a whiteboard. You should be able to do the same with AI. This means creating spaces and processes for it.
- Shared Prompt Libraries: Create a shared repository where team members can save and share their most effective prompts. This is the new water cooler. It’s where people learn from each other and build on each other’s successes.
- “AI Office Hours”: Host a weekly, optional call where people can drop in to share a cool trick they learned, ask for help with a specific task, or just see how others are using the tools. It makes the learning process social and collaborative.
- Integrate, Don't Isolate: The tools should live where your team works. If your team lives in Slack, the AI should be in Slack. If they live in Figma, it should be a Figma plugin. The more you can reduce context switching, the more likely people are to adopt the tool.
3. My Unpopular Opinion: Most AI Tools Are a Waste of Money
I get pitched new AI tools every single day. Most of them are just thin wrappers around an OpenAI API call. They solve a niche problem that could be solved with a well-written prompt.
Don't get distracted by the shiny objects. Before you buy any new tool, ask yourself these three questions:
- Can we do this with our existing tools? Chances are, the powerful models from companies like Anthropic, Google, and OpenAI can already do what you need. You just need to learn how to ask.
- Does this solve a real, recurring pain point? Is this a “nice to have” or a “need to have”? If it doesn't solve a problem that your team faces every single week, it's probably not worth the investment.
- What is the total cost of ownership? The sticker price is just the beginning. You also have to factor in the cost of implementation, training, and maintenance.
Focus on a few, powerful, general-purpose AI tools and teach your team how to master them. A team of expert prompters with access to a powerful model will outperform a team with a dozen niche AI tools every single time.
The Future is Not About AI, It's About Augmented Teams
I learned my lesson. After the initial failure, we went back to the drawing board. We started small. We picked one team—our content marketing team—and we worked with them to build an AI workflow from the ground up. We didn't force a tool on them. We asked them about their biggest challenges and then explored how AI could help.
We built a custom AI writing assistant, trained on our own content and brand voice. We created a shared library of prompts for everything from generating blog post ideas to writing social media copy. We celebrated the small wins and shared them with the rest of the company.
Slowly but surely, things started to change. Other teams saw the success of the marketing team and they wanted in. The adoption was organic, driven by pull, not push. We weren't just installing software; we were changing mindsets.
Building an AI culture is not a technical challenge. It's a human one. It requires empathy, patience, and a willingness to admit when you're wrong. It’s about empowering your people, not just implementing new technology. The 47% of companies that fail are the ones that don't understand this. The 53% that succeed are the ones who put their people first. Don't be a statistic.
Frequently Asked Questions
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.
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.