I recently had the pleasure of catching up with Mike — founder of Zurp, Juice, and now Maya, a wellness brand he co-founded with his wife Jess. He’s also a General Partner at Magic Fund, where he invests in early-stage companies.

Mike and I worked side-by-side for two intense years while building Zurp, and I can say with full confidence: Mike is one of the most resourceful and hardworking people I’ve ever met. A true builder, a relentless learner, and always on the bleeding edge of innovation. Even though he’s super busy, he’s generous with his time and always willing to help others level up.

This conversation was packed with insights — from practical ways to use AI in daily work to mind-expanding thoughts about how it will reshape marketing, learning, and human creativity.

You can find & contact Mike at @mikeliso and @_Liso_

Everything in <60 Seconds

• Mike’s AI “wow” moment? He laughed at Jess’s idea that AI could build a website. Months later, GPT-4 proved her right.

• He uses AI as a thought partner: not just for content, but for research, strategy, and scaling team output.

• The true edge? Memory. The more it knows about you, the more powerful it gets. “Context compounds.”

• What most ignore? Overconfidence. “It fools you into thinking it’s excellent—it’s just good.”

• Mike’s go-to AI protocol: ask it to take a role, prompt a direct question, prompt opposite question, then ask it to check its work.

• Humans will be orchestrating more and creating less initial outputs — but creativity in how to refine, use tools, and orchestrate will become more valuable than ever.

• AI introduces a third mode of learning: personalized dynamic. Humans learn better when they can ask questions. It’s like having a personal tutor with infinite patience.

Tools & Resources

Here are some tools Mike uses or is currently exploring, as well as some experimental tools that I came across this week:

🧠 Research & workflows

Perplexity Pro
AI-powered research assistant with up-to-date sources and deep reasoning capabilities.

GPT Operator
Automates prompt loops and workflows—ideal for chaining tasks or building AI agents.

🎨 Creative

Ideogram
Text-to-image generation tool great for ad creative, visuals, and quick concept mockups.

⚡️ Prototype Quickly

Base 44
Rapid prototyping platform for building AI-enhanced apps with minimal code.

🧪 Experimental Tools

Glato
Early-stage AI interface for content generation and task execution. Still evolving.

Plus AI
AI slide creation tool that turns text and data into polished presentations in minutes.

How Mike Turns AI Into a Second Brain

The Wow Moment

We’ve all had that first “holy sht*” experience with AI. I asked Mike when that hit for him — and his answer says a lot about how fast AI evolves:

For Mike, that moment flipped a switch — from I can use AI to I need to use AI. For me, it was when I built a full web app in a weekend. It’s these exponential jumps that keep surprising us. The innovation outpaces our ability to adapt, each leap arrives before the last one fully settles.

The Real Unlock

The real unlock is Memory.

I fully agree, and this is where I see a big change. Before ChatGPT, one of the biggest levers for stickiness was network effects — what we now have are engagement effects. The more you use the system, the better it becomes for you. That’s why I sometimes default to GPT, even when other models seem more advanced.

Mike put it perfectly:

AI’s Blind Spots (and Ours)

Mike was very honest about the limitations and dangers. AI is overconfident, and so are we.

This matches my experience with AI, but Mike said this one thing that really hit me.

He recalled having to convince a friend to take a Grok-generated contract to a real lawyer. AI can miss crucial legal provisions — but it always sounds totally confident.

How to Use It Well

Mike doesn’t prompt once and hope for the best. He’s built a repeatable method:

Example:

Assign a role: “Act as a world-class customer experience strategist.”

Ask the direct question: “What are the best ways AI can improve customer support at scale?”

Then invert it: “In what scenarios would AI actually make customer support worse?”

Check your answer: “Compare both perspectives and check your work. What are the best ways AI can improve customer support at scale?”

Where are we headed

One of Mike’s core domains is growth and marketing — and he sees massive transformation ahead.

He also sees ideation and inspiration happening faster than ever.

Understanding the psychology, strategic thinking, and orchestration becomes key, the execution is now AI-enhanced from day one.

I agree that there will still be a human component. I brought up an analogy: a private banker isn’t really doing much technical work anymore — they’re building human connection, trust, and reassurance. A nod. A nudge. A voice that says, “You’re doing the right thing.”

Mike added:

That’s the part AI hasn’t replaced — and maybe never will.

Final Thoughts

Talking to Mike always sharpens my thinking. This wasn’t just about AI tools — it was a philosophical and visionary discussion about AI. I got a glimpse into how Mike uses AI to learn faster, work smarter, and lead better.

What I loved the most was how he uses AI as a second brain that enhances his thinking and learning curve, not just as a shortcut. His 360 degree approach of prompting, inverting, and checking — is intentional and strategic to avoid the pitfalls and get better outputs.

AI is powerful when used as your “co-founder”. Like with any co-founder, the risk is blind trust. The edge is thoughtful use with a 360 degree check mechanism. And most importantly, we’re all still humans who need connection and psychological reassurance from another human.

Hope this left you inspired and with something new to try.