Published Date:
May 15, 2025

5 Habits of Effective AI Leaders You Should Adopt Now

The best AI leaders aren’t tech-savvy; they think and act differently when it comes to working with AI. These are five ways that help "manager nerds" stay ahead in the AI era.
By
Daan van Rossum
Founder & CEO, FlexOS

Five Ways “Manager Nerds” Go Beyond the Surface of AI

Whether we call them “​SuperWorkers​” or “​Manager Nerds​,” it’s clear that leaders who run a human plus AI team are the future.

“I think it's actually going to be the era of the manager nerds now, where I think being able to manage fleets of AI agents and orchestrate them is going to make people incredibly powerful,” said Anthropic (Claude) cofounder Jack Clark this week.

And look, while ​almost every hiring manager​ is looking to hire these AI SuperWorkers, no one has fully figured it out yet.

So, if you’re one of those who feel like you’re just scratching the surface of AI, then look no further than last week’s graduating class of the ​Lead with AI Executive Bootcamp​ for killer insights.

Here are five ways these AI-supercharged leaders say they think and act differently when it comes to working with AI:

1. Default to AI.

We know from ​research​ that the more AI use cases you have, the more you get out of the technology. This is why we need to give AI a ‘​permanent seat at the table​,' as AI professor Ethan Mollick calls it, and let AI co-do everything with us.

Not always to replace us, but at least to collaborate with us and augment our capabilities.

One graduate experienced this first-hand. He asked a friend for help with interview prep and got a fantastic answer… then realized that his friend had just copied the question into ChatGPT.

That moment reframed how he should always use AI as a team member, even if he didn’t think it could help on a certain task.

2. Quality Prompting is Key.

While AI is extremely capable, beyond what most of us can picture now, it still needs direction. As Dror Poleg says, “AI can do anything, but we have to direct it.” This is where prompting comes in.

As I wrote last week, prompt engineering isn’t a job anymore; it’s a ​core skill everyone needs to possess​.

Quality prompts are like delegating your work to a supersmart human. Context matters – a lot. See my ​CODO SuperPrompting framework​ and our ​ChatGPT prompt generator​.

For many in our latest cohort, the lessons on prompt engineering were a game-changer. (Their words, not mine.) The better you get at prompting, the more you get out of AI. It’s as simple as that.

3. Trust, But Verify

It’s amazing what AI can do.

However, as one participant highlighted to the group, we need to maintain healthy skepticism about AI-generated outputs and human oversight even when AI is impressive.

Whether Perplexity or ChatGPT Deep Research, use AI to accelerate your work, but verify outputs. There’s a reason why we added an entire lesson on beating ‘hallucinations’: AI isn’t perfect and can make up entire quotes and references.

Checking sources, comparing AI outputs (like ChatGPT vs. Claude), and forcing it to use your information are all ways to ​prevent AI hallucinations​.

4. Thinking More, Not Less.

​Research​ found that younger users scored lower in critical thinking because of cognitive offloading when they delegate all thinking to AI.

AI overuse could reduce cognitive abilities, including memory retention and problem-solving. Another paper called this “​Falling Asleep at the Wheel​.”

But many of our graduates felt the opposite. One exercise we do in particular, doing an AI voice call with an advisor like Adam Grant, made them experience the power of AI as a way to think more and better, not less.

Rather than only using AI as an army of always-on interns, these leaders also use their AI team members as senior advisors to work through ideas and shift perspectives.

5. Learn Together

There’s a reason I created this executive boot camp as a ‘social learning experience.’ No matter how behind we feel, we can always learn more from each other than just by ourselves.

Multiple participants highlighted how the community aspect was key to mastering AI in such a short time frame. One person mentioned, for example, that seeing how others refined prompts and shared feedback was critical to pushing AI beyond generic outputs.

Asking the ‘stupid question’ has become part of our community's culture. What we may think everyone else already knows is probably exactly what everyone is wondering.

Becoming an AI Leader

Even after working intensely with AI for over 2 years and training hundreds of leaders, I still feel like I’m only scratching the surface.

That's normal and expected.

AI is not something you can study and ‘understand.’ As the technology is changing rapidly every single day, the best we can do is know the fundamentals and experiment, experiment, experiment.

As ​Marcus Bowen​ concluded during last week’s graduation session: “We’re still early—and that’s the opportunity. We’re just getting started.”

Wherever you are in your AI journey, these five principles should help you to get increasingly more out of it.

Lead with AI - Icon

Just Scratching the Surface with AI?

Many experiment with AI, but few go beyond the surface. Basic prompts, one or two tools, and a lack of focus.

We’re about to change that.

On June 6, we're launching our latest Cohort of Lead with AI, the Executive Bootcamp that transformed hundreds of leaders at companies like Apple, Microsoft, and McKinsey.

You could get so much more out of AI if someone showed you the way, and that's exactly what we'll do.

No matter what level of AI you are now, by the end of June, you'll be an AI expert.  

And, uniquely in Cohort 9 of Lead with AI: ✅ Our unique AI Certification to prove your expertise anywhere, instantly ✅ AI-powered workflows tailored to your role and industry  ✅ A private peer community that turns questions into breakthroughs

If you want to stay ahead, you don’t want to miss this.

Launching in just 4 days, so:

👉 Join the Waitlist 👈
Lead with AI - Icon

Category Essentials: AI for Vibe Coding

Each week, I spotlight one category and suggest the three tools that are tried, tested, and trusted by ​Lead with AI members​.

For this week: We’re clearly stepping into the vibe coding era. Last week, ​Figma launched its vibe coding features​. Google is building ​a software developer agent​. ChatGPT has ​agreed to buy AI coding assistant Windsurf​ (which likely means we’ll see vibe coding built right in soon!).

But for now, if you'd like to try this AI-enabled workflow for yourself, here are three tools our community members are already using—with real outputs, experiments, and even hands-on guides.

#1 Lovable

If your goal is to turn an idea into something tangible fast, Lovable can help you get there. Our community member ​Fredrik​, who has an engineering background, said Lovable is “fantastic at getting ideas validated quickly.”

It generates working UIs that look good, making it great for pitching, prototyping, or just pressure-testing a concept.

But if you're aiming for something more refined, Fredrik noted that it's important to be specific in your requests. And you can ask Lovable to "refactor [certain sections] as if you [Lovable] were a senior React engineer" for better code lines.

​>> Try Lovable here.​ (Lovable tutorials in our PRO library ​here​)

#2 Bolt

A few community members have been trying Bolt to bring app ideas to life, especially when they want something functional without needing deep technical skills.

It’s been used in open-source projects, business simulations, and early-stage prototypes, with members highlighting its ability to produce usable, deployable results.

​>> Try Bolt here.​ (​Helen Lee Kupp​ has a mini-course on how to use Bolt ​here​.)

#3 Cursor

Cursor came up as a tool that gives you more control, especially if you’re comfortable with code or working on logic-heavy builds.

​Helen​ shared that she’s been using it lately, though with a note: “It’s a hard first setup hill.” It’s closer to a real dev environment than other tools, which means more flexibility, but also a bit more effort. ​Here​ is her prototype using Cursor.

​>> Try Cursor here.​

‍Want me to cover a specific category and/or AI tool next? Reply and let me know here.

AI Drains Work Drive, Security Risks from AI Code, Safari Searches Drop Due to AI

I read dozens of AI newsletters weekly, so you don’t have to. Here are the top 3 insights worth your attention:

#1 AI Can Boost Output—But Drain Motivation

​A study​ of 3,500+ people found that while gen AI improves performance, it also makes employees feel less motivated and more bored when switching to non-AI tasks.

Why? AI removes the mentally stimulating parts of work. And it raises a concerning question: As AI takes on more, how are you helping your team keep work meaningful?

The authors suggest five ways that companies can do in ​this HBR article​. And the solution isn't to abandon AI!

#2 We Can Now Code Faster Than Ever—What’s Missing?

AI has made coding wildly accessible. But ​new research​ reveals a risk: LLMs often suggest fake software components—called “​packages​”—that don’t exist. Hackers can upload malicious code using those names and wait for someone to install it.

As more teams rely on AI to build quickly, we’ll need better checks. How should we design a new system to verify what's being built with AI?

#3 Apple Users Are Ditching Safari for AI

For those watching the future of AI search, here’s a real signal: For the first time in 22 years, ​Safari search volumes have dropped​. And Apple's senior vice president of services, Eddy Cue, says AI is the reason. iPhone users are turning to tools like ChatGPT and Perplexity instead of "googling" in Safari.

Cue confirmed Apple is exploring a Safari redesign to integrate AI search more directly. But here's what matters more: search behavior is shifting—people now actually ​"search" on social and AI​.

Lead with AI - Icon

Prompt of the Week

A good prompt makes all the difference, even when you're just using a core LLM.

The more building with AI becomes easier, the more your ideas matter. This week’s prompt helps you push past the obvious and surface sharp, original use cases worth testing in your next vibe coding experiment.

Generate High-Signal Vibe Coding Ideas

You’re a creative technologist who deeply understands [your role/ work context] and what sparks delight, surprise, or utility in my work.

Your task is to generate original, high-signal ideas I can test using an AI coding assistant. Don’t name or explain the tool — just focus on clever, context-aware use cases that are uniquely relevant to me.

Avoid generic, recycled, or beginner-level ideas. Think sharp, fresh, and execution-ready.

Give me 5–7 standout ideas. Each should include a short note on what makes it unexpectedly useful or worth trying.

👉 Try it, tweak it, and save it for your future use. If this prompt is helpful (or if you made it better), I’d love to hear how.

👉 Want a free prompt library template?  Reply with one thing here, and I’ll send it your way.

Lead with AI - Icon

AI for Strategy, Responsible Adoption, and Prototyping: From the Community

Every day, ​Lead with AI PRO members​ discuss practical ways to benefit from AI in their work and organizations. This week's highlights include:

Don't want to miss more insights and conversations like these?

Then it's time to upgrade to PRO:

Join The Leading Business AI Community

Lead with AI - Icon

If you made it this far, reply and tell me what you'd love AI to take over in your daily workflow.

Also, please forward this newsletter to a colleague and ask them to subscribe.

If you have any other questions or feedback, just reply here or inbox me.

See you next week,

Daan van Rossum - Lead with AI

Daan van Rossum​

Host, Lead with AI

Daan van Rossum

Founder & CEO, FlexOS

Five Ways “Manager Nerds” Go Beyond the Surface of AI

Whether we call them “​SuperWorkers​” or “​Manager Nerds​,” it’s clear that leaders who run a human plus AI team are the future.

“I think it's actually going to be the era of the manager nerds now, where I think being able to manage fleets of AI agents and orchestrate them is going to make people incredibly powerful,” said Anthropic (Claude) cofounder Jack Clark this week.

And look, while ​almost every hiring manager​ is looking to hire these AI SuperWorkers, no one has fully figured it out yet.

So, if you’re one of those who feel like you’re just scratching the surface of AI, then look no further than last week’s graduating class of the ​Lead with AI Executive Bootcamp​ for killer insights.

Here are five ways these AI-supercharged leaders say they think and act differently when it comes to working with AI:

1. Default to AI.

We know from ​research​ that the more AI use cases you have, the more you get out of the technology. This is why we need to give AI a ‘​permanent seat at the table​,' as AI professor Ethan Mollick calls it, and let AI co-do everything with us.

Not always to replace us, but at least to collaborate with us and augment our capabilities.

One graduate experienced this first-hand. He asked a friend for help with interview prep and got a fantastic answer… then realized that his friend had just copied the question into ChatGPT.

That moment reframed how he should always use AI as a team member, even if he didn’t think it could help on a certain task.

2. Quality Prompting is Key.

While AI is extremely capable, beyond what most of us can picture now, it still needs direction. As Dror Poleg says, “AI can do anything, but we have to direct it.” This is where prompting comes in.

As I wrote last week, prompt engineering isn’t a job anymore; it’s a ​core skill everyone needs to possess​.

Quality prompts are like delegating your work to a supersmart human. Context matters – a lot. See my ​CODO SuperPrompting framework​ and our ​ChatGPT prompt generator​.

For many in our latest cohort, the lessons on prompt engineering were a game-changer. (Their words, not mine.) The better you get at prompting, the more you get out of AI. It’s as simple as that.