To Build or Not to Build
In a vibecoding world, everyone can develop apps, websites, and services. But should they?
My buddy Patrick Ashamalla at SXSW. He is a sketch artist, and yes, someone who leads product development initiatives.
Vibecoding tools like Lovable, Bolt, vO, Claude Code, Google Code, and Codex have put the power of building in the hands of non-developers. A well-discussed story and benefit of the LLM boom, vibecoding gives every human being the ability to build. All that is required is a computer and a subscription to one of the many coding tools.
But just because they can, should everyone build?
To build or not to build, that is the question, to riff on the old bard.
It’s a complicated question, and really, who is to tell anyone that they shouldn’t build?
The Creative Case to Build
Vibecoding helps so many creatives, entrepreneurs, and everyday people like me express a vision for work. In some cases, they can execute all the way to the finish line. In other instances, they are able to create a vision and share it with a designer or developer who can finish the product, saving several iteration cycles and helping them get their concept to market faster than ever before.
Author Jeannie Walters mentioned this exact reason for using Lovable in the most recent AI Hope podcast. She said, “I can take that now and put it all into basically a working prototype and then say, ‘Okay, the last part of it where we’re connecting and all that, I’ll let somebody else do that;’ but they have exactly what I want. It just speeds up getting my vision out there, which has been phenomenal.”
I remember, as a child, being told I shouldn’t be an artist, that I would end up a starving artist in Brooklyn (impossible today, given current NYC real estate values). Fast forward 40 years, and yes, I am a consultant, but I write and photograph as a storyteller and to market my own business. I am not starving.
In many ways, I see building as a tool for creative expression, much like a camera or an article-creation platform. I have built my website and an app using Lovable and am helping clients do the same. But though I build, I am not a builder, per se.
It’s just not my creative calling. I don’t wake up with thoughts of new apps and tools and sites I can create. That’s not to knock it. Again, this toolset is helping so many other creatives achieve their visions. And for me, when there is a business need, I feel quite comfortable using the tools to achieve my goals.
Yet, it doesn’t always make sense to DIY. I think that’s particularly true the larger the company and team.
The Business Case Not to Build
Speed to market for developers who use coding tools to assist in their development makes a ton of sense. What the right balance of human oversight, handcoding, and intervention is best left to those who understand quality development.
Similarly, using these tools to help express creative vision and shorten product cycles makes sense. The NY Times literally vibecoded a website live on a 20-minute episode of The Daily podcast. What a great example of rapid development and deployment.
For some, the speed-to-market benefits of vibecoding make perfect sense. That is until it doesn’t.
Certain aspects of bringing products and services to market benefit from quality assurance and vetting. For example, vibecoding is known for creating sloppy code on websites, products, services, and other software. Thus, the above comment about needing experienced developers to manage code deployments.
Another aspect of that is understanding product impacts on customers. Does it have the necessary context to serve its intended audience? Does it have inherent biases? Can it handle the load if well served? Are there bugs or security risks in the code?
A small startup may better be able to navigate the issues a quick but not well-thought-out vibe-coded deployment can bring. But the more established a brand is, the greater the risk such rapid deployments pose. This Forbes article does a great job of examining some of the product-development checks and balances that a more established brand should consider. Creating the process to vet vibecoded projects before deployment is just good business sense.
In other cases, it may simply be a nice idea to vibecode, but not a good business decision. Like all AI products, these tools can trigger a bit of shiny object syndrome even among the wisest professionals. Businesses have to decide what the best allocation of their resources is to achieve their mission.
Often, I think my time is best spent working on other projects rather than demonstrating that I can create apps with coding tools. What’s the best value for my business? Just because I can vibecode doesn’t mean I should.
Eighteen months ago, that choice didn’t exist. What a testimony to the power of AI.