5 MISTAKES FOUNDERS MAKE WITH AI

1 Oct 2026 · 1 min

Watch on YouTube

1:09

In this concise instructional video, Eric Siu outlines five common errors founders commit when integrating AI into their businesses. 1) Delegating AI ownership to others – founders must personally champion AI change management. 2) Treating AI initiatives as theater rather than revenue drivers – hackathons and trainings must translate into measurable impact. 3) Purchasing disparate AI tools without creating a unified "company brain" that enables shared context and memory across tools. 4) Hiring non‑AI‑savvy staff and expecting them to catch up – AI‑pilled talent builds workflows and systems at speed, while others lag. 5) Failing to make every team member a builder – using leaderboards to surface shipping and review activity encourages a culture where everyone contributes to AI‑enabled product delivery. Siu ends with a call to follow for more AI, business, and marketing insights.

Chapters

The first is believing that they can delegate AI to somebody else. You have to own it yourself so you can drive that change management throughout your organization. So it doesn't matter who you hire, if you're not forcing it through, it's not going to happen. Number 2 is theater versus revenue. And so you might do these hackathons, you might do all these trainings, but if you don't actually have people show that they're driving impact, they're driving revenue, whether it's for you or driving it for their clients, it ultimately just becomes a bunch of theater. Buying a bunch of tools, But not building a brain. You have to build a single brain, a company brain where all your different tools can talk to each other. You have shared context, you have shared memory as well, and the team can continue to grow with this. Believing that you can hire non-AI-pilled people and train them up. They might have all the experience, but trust me, if you hire these people and they fall behind, they won't be able to catch up to the AI-pilled people who are building workflows, they're building systems, and they're building, building, building. When you have people that are working in the Stone Age, that's not going to work out for you. Number 5 is not having everyone become a builder. We have a leaderboard that shows how often people are shipping and how often people are reviewing. And then if people need help, how do we get out there and how do we help these people? Follow for more on AI, business, and marketing.