“The boss built a thing and he wants you to take a look.” These are the words echoing through hallways around the world, as executive leaders get their hand on vibe-coding tools. 

This kind of statement invariably sends a shudder through the engineers, designers, and product managers who know that they’re about to get shown something hacked together with Claude in a weekend—as they’re simultaneously asked why it’s taking them so long to ship their software. We’re hearing about it everywhere: There’s a plague of vibe-coding CEOs spreading across industries.

I’m going to go against convention and make a case to engender sympathy and love towards those same CEOs. Yes, I know I’m unlikely to convince you. And yes, I am the Chief Executive Officer of Aboard. Nonetheless, I believe I have a strong case.

As my co-founder Paul Ford says, “Most people don’t know what business they’re in” (his words—you can reach out to him at hello@aboard.com). But the CEO is the business; They live and breathe it. They think about stability, growth, and risk reduction. They’re permanently anxious because they know that no matter how well things are going today, in a few months, everything could fall apart.

One thing people don’t realize about business leaders is that the more people work for them, the more powerless they become. CEOs often start their careers with actual skills. They might be engineers, or at least make good PowerPoint decks. But as a business scales and hiring increases, the CEO is carefully and methodically marginalized. They must stay out of the way. They are distracting and disruptive.

Old-school CEOs had their own way of handling this. They played golf, read reports on planes, and occasionally fired their leadership teams and reorganized their companies. Everyone basically despised them, and they were fine with that. When they needed to communicate, they would send a memo. Brief, and often tactical: “I did some thinking. Here’s where I’m at and here’s what I want.” In an ideal world, those thoughts would be born from a broader strategic context. Of course, just sending a memo means nothing; the memo would usually come with a task force. “I’ve put Sally in charge of the new effort. She will work with the leadership to implement the Purple Plan by Q4, and the highest performers will be our next generation of leaders.” Then back to the golf course.

The modern CEO is different. They’re supposed to be peer-mentors. If they play golf, they don’t talk about it. Organizations are flatter; transparency is key. As a result, the modern CEO doesn’t send a memo: They load up Tableau (or Snowflake or Databricks). They want to share their insights. And, importantly, they want to show that they still have the skills—that they can do the work.

Both kinds of CEOs have their advantages and disadvantages, but there’s one big difference between them: For the latter kind of CEO, this new wave of AI empowerment is a dream for them, but a nightmare for their employees. In the past, CEOs might have rejected digital opportunities—social media, blogging—but they’re bought in on this one. That’s because a vibe-coded demo is the ultimate executive memo. 

A message to my fellow CEOs

Now look. We can talk all about how great it is that you’re able to prototype, all the ways you can exercise your leadership using AI, and how it’s bringing you closer to the core product. But all I can say is: Cut it out. Seriously, cut it out. 

Your interactive C-suite demo memo is not going to get you any results. Here’s why:

• Your demo is not impressing anyone. They may pretend they’re impressed (you’re the CEO, after all) but there is so much stuff being made by so many people that nobody thinks anything is great anymore, ever. Yes, you made it in a weekend. So can anyone.
  
  
• When you drop a demo, it’s usually the opposite of clarifying. It’s confusing. When an executive says something or writes something, people’s brains go to managing expectations, de-risking, and due dates. None of that is included in your demo. Its fidelity and complexity is little more than the corporate equivalent of a bad acid trip.
  
  
• “Why don’t we just do what I did here?” is demoralizing. I get why you’re frustrated. Nothing seems to get done, regardless of all these wondrous advancements in AI. It trivializes all the actual work that needs to get done to produce something real and reliable. It dismisses people’s expertise and skills. You also don’t look good—you even look just a little worse.

We are all CEOs now

Everyone today has convinced themselves that there is clarity and salvation in making software fast and showing it to other people. This is not the case. What I keep seeing from vibe coding is too much software, all at once, and not enough product. 

I am utterly convinced that a tightly specified piece of software—a CMS or CRM—can be built in a tiny fraction of the time that it used to take. That is amazing. I am, so far, utterly unconvinced that it will be good.

The more optimistic view is that while we’re imbued with all these superpowers, our first instinct is to show our creations to other people. This confirms that consensus and social validation still very much matter. We want to connect with one another and we’re using these creations to engender enthusiasm and support. 

But ultimately, it isn’t working. Quiet rooms and blank stares abound because nobody’s really sure what to do next. Demos are great for sparking conversations, but the hard work of making useful, intuitive software still lies ahead. I’ve heard it in one too many meetings: “Don’t take this literally. This is just a bunch of ideas…” The skepticism and mild confusion surfaces almost immediately. Even CEOs end up feeling deflated (poor things). In a world where everyone can create anything rapidly and with very little effort, how do you stand out?

The answer is the same answer before AI showed up: Make a great product. AI can produce outputs all day long, but it can’t guarantee quality. Quality software is borne out of creative problem-solving and lots of little, tiny moves, not one-shot magic prompts.

I love much of the AI progress I’m seeing. I’m vibe coding, too. But I’m standing my ground: A good product takes deep thought, and that’s a slow process. So far, it’s something that only skilled humans can do. 

For what it’s worth, I’m absolutely willing to be convinced otherwise. Send me your great, vibe-coded examples, built in a day or so, of really novel, exciting app experiences that lots of people love to use.

Crafting something that stands out in this age of digital abundance takes deep thinking and hard work. “More” is often worse, not better, requiring us to make tough decisions about what to take out and what to leave in. This requires a lot of trial and error and a lot of refinement. It’s a bittersweet lesson: Making great things is still not easy. This is why we have product companies, with product managers to coordinate, engineers and designers to build, and CEOs, for better or for worse, to lead them.