United's CEO Vibe-Coded a Betting App Between Gym Sets. Here's What That Means for Designers

United Airlines CEO Scott Kirby can now add "vibe coder" to his resume, and he is proud of it. On the "Cheeky Pint" podcast with Stripe cofounder John Collison, as reported by Business Insider on October 27, Kirby described the moment AI clicked for him: he hired a tutor for three hourlong sessions, and then in a two and a half hour gym session he built a working sports betting webpage by typing prompts on his iPad between sets.

The page has "four or five" dropdown menus for different sports, dates and bets. It links to four sportsbooks, pulls up the best odds, and lets him enter a dollar amount and place the bet. That tool was his "AI epiphany," he said. Then he made the part that should get every product team's attention: "I think I called them on that Saturday, our team, and like, 'We're all in on this.'"

Image: United Airlines via Wikimedia Commons, CC BY 4.0

What the coverage keeps missing

The gambling jokes write themselves, so everyone is writing them. Casino stories, card counting bans in Las Vegas, the weird novelty of an airline executive with a betting habit. It is all filler, and it buries the actual story.

The story is the prep time. Three hourlong tutoring sessions. One iPad. One Saturday phone call to declare the whole company "all in." The distance between "I have an idea" and "I made a working thing" just collapsed to a single weekend, at the very top of a Fortune 100 org chart, for the busiest kind of executive there is.

Kirby also swiped at the use case everyone else keeps selling him. "Summarizing documents and all that crap, I think it's ridiculous," he said. "I can read my calendar. I don't need to summarize my day." While half the enterprise world is buying AI to generate more meeting summaries, the CEO of United looked at all of it and picked building tools over chatbots. That is a product judgment, and honestly, a correct one.

My take: this is good news for designers, if you treat it as an invitation

For twenty years the pattern was identical everywhere: someone with budget and authority had an idea, and then handed it down a ladder to the people who could actually make it. Entire careers were built on standing on that ladder.

Kirby just publicly stepped over the ladder. He is not going to be the last. Every executive who hears that story will try it on Saturday, and most weekends from now on, some executive somewhere will ship a working rough draft of the thing they keep asking your team for. It will be ugly. It will have bad defaults, no accessibility pass, confusing states, and in this particular case, some compliance questions that a gambling odds aggregator probably deserves. But it will work, and "it works" changes what people believe is possible.

Designers who define their job as "the one who makes the thing" are about to compete with CEOs on Saturdays. That is a race you do not want.

Designers who define their job as the judgment layer, knowing what is worth making, why version one will hurt someone, what to cut, what has taste and what is noise, just watched their leverage go up. When everyone can produce output, the scarce skills become everything around output: framing the problem, designing the system, and having the taste to tell good from fast.

What to actually do about it

1. Run Kirby's experiment yourself this week. Three hours of tutoring, then build one small tool you personally would use. Not a portfolio piece, not a client demo, something that does a job for you by Sunday. You cannot coach a process you have not lived, and most designers have not lived it.
2. Stop selling output, start selling judgment. Rework how you present your work and your services: problem framing, system design, design quality standards, the "why" behind choices. That is the layer a Saturday prototype cannot touch.
3. Build the "prototype to production" path before you need it. Exec prototypes will start landing in your inbox uninvited. The team that answers with "here is exactly how we take rough demos to shippable, and here is what we check along the way" owns the pipeline. The team that answers with "please follow the intake process" loses the meeting.
4. Coach in public. The tutor Kirby hired is the template for a new, high-value design role: the person who teaches non-designers to build well. The designers who become that resource inside their companies will have more influence over what ships than they ever had as the maker.

If you want the fastest version of that experiment for yourself, Base44 is the AI app builder I recommend first (affiliate link: it supports this newsletter and costs you nothing extra). You describe the app, it builds and hosts it, and you focus your three hours on the product judgment that is suddenly the valuable part.

The takeaway

The funniest part of this story is that it is not about gambling at all. It is about the moment the ability to make software stopped being a profession's moat and became everyone's default setting, all the way up to the corner office.

Your CEO will try it. Probably soon. The question is not whether the draft they build will be any good, because it will not be. The question is whether your design team is ready to be the judgment on top of the draft, or whether you are still planning to be the draft.

What would you build if you had his Saturday?

Related on NewForm

• A Vibe-Coded Site Got Hacked 6 Days After Launch. It Was Not the AI's Fault
• Google Play Now Runs Your Personal, Vibe-Coded Android Apps. This Is a Bigger Deal for Designers Than It Looks
• Epic Games Vibe-Coded Its New Launcher. The Real Story Is the "AI First" Order