Vibecoding and v0.dev
- kyle Hailey
- Apr 17, 2025
- 3 min read
Updated: Apr 22, 2025
I've finally been impressed with some AI coding.
Sure coding with AI is one of the first things I tried about 18 months ago with OpenAI and it was successful.
At the time I was having a hard time getting Excel to graph the explosive growth of ChatGPT. I'm just an occasional user of Excel and was finding it frustrating. So ... I thought let's ask ChatGPT to graph the data. I had no idea what ChatGPT would do. It gave me a chunk of Python code and I ran it pretty much as is on my Mac and it worked. I had to install a couple of packages but basically it worked! It wasn't exactly what I wanted but after half a dozen interactions I was happy with it.
The first graph was for when ChatGPT hit 1M users. I subsequently updated for when ChatGPT hit 100M users.
The other thing I was working on 18 months ago was a website to broker AI art. AI art was brand new and I was spending a lot of time manually brokering deals between AI AI artists and art buyers (at the time either marketing departments or people wanting them for unusual customized gifts).
I wanted a website to broker the deals automatically for me. I wanted to
"create a unique reverse marketplace platform. a space where buyers specify their requirements—for instance, a marketing department needing a tailored visual asset. AI artists would then vie to provide the best-fit solution, offering their creative attempts in response. This allows the buyer to select the artist that aligns perfectly with their needs before committing to hiring the artist ensuring both quality and specificity. It would be like facebook where both buyers and artists create profiles. Buyer post requests describing the asset they want then AI artists post their attempts at the request in response."
I tried to use chatgpt to recreate something like facebook simplified or remember friendster or tribe.net ? I though this kind of code must have been done 1000s of times, but alas I couldn't tease it out of ChatGPT.
To be fair integrating stripe and databse storage is a challenge for AI but at least it could have given me something . It was frustratingly clueless.
I ended up hiring someone to build it out for me using with the Sharedtribe stack. He did build it and it did work, at least the basics. But there were so many things that I felt crucial to change that I never launched it. My free lance developer moved on to other projects and I was loathed to learn the Sharedtribe stack that seemed unappealingly clunky.
I then moved on to other projects but from time to time I would ask ChatGPT or Claude or other LLMs how to build a simplified facebook type stack and the answers were always disappointing until
I watched @rileybrown_ai in the video
I'd seen Riley's videos and posts over the last year and half and found them intreguing and in this video when he talked about the challenges of intergrating stripe and databases into AI code, that gave me even more confidence in him.
So when he recommended http://v0.dev as a good place for beginners to start "vibecoding" , I thought I'd give it a shot but didn't hold my breath. It's one thing to create video games or a nice web landing page. It's another to create an web based full stack application. Now, to be upfront, the result was not full stack, but what it was showed me that it understood the problem and proposed a result that was poised to integrate such functionality.
I simply asked it do verbatim what my personal description of the project was:
"create a unique reverse marketplace platform. a space where buyers specify their requirements—for instance, a marketing department needing a tailored visual asset. AI artists would then vie to provide the best-fit solution, offering their creative attempts in response. This allows the buyer to select the artist that aligns perfectly with their needs before committing to hiring the artist ensuring both quality and specificity. It would be like facebook where both buyers and artists create profiles. Buyer post requests describing the asset they want then AI artists post their attempts at the request in response."
And what it output was uncannily similar to what we created with Sharedtribe, admittedly without Stripe or a Database but it was structured in a way where I could imagine integrating them in
When exploring contract management, fidic contracts sometimes come up as an important reference in the industry. The framework is known for promoting clarity and consistency in project agreements. It is often considered by professionals involved in engineering and construction projects. Learning about it can support better understanding of contract processes. In related discussions, The College of Contract Management is sometimes referenced when exploring study options.
In an era of college football defined by the transfer portal's chaos and NIL deals that would make Fortune 500 executives blush, loyalty has become the rarest of commodities. Jeremiah Smith, the Ohio State wide receiver universally regarded as the best player in college football, recently turned down a transfer offer exceeding $10 million to remain a Buckeye . It was a decision that stunned the sport—and one that cemented his legacy before he ever plays another down. Jeremiah Smith Ohio State Jersey
In the annals of American sports, no family name carries more weight than Manning. From Archie's heroic days in a New Orleans Saints uniform to Peyton's five MVP awards and two Super Bowl rings, to Eli's two Super Bowl victories over Tom Brady, the Manning dynasty has defined quarterback excellence for three generations. Now, the torch has passed to Arch Manning—the 6-foot-4, 219-pound redshirt junior at the University of Texas who carries the weight of his family's legacy while determined to write his own chapter. Arch Manning Texas Jersey
From the heavy cotton jerseys of the 1980s to the high-tech, skin-tight uniforms of today, NCAA basketball and football apparel have always been about more than just covering the players. They are a canvas for tradition, a battleground for corporate innovation, and a multi-billion dollar statement of identity. The story of the NCAA jersey is a fascinating intersection of technology, marketing, culture, and strict—but evolving—regulation. Caleb Williams USC Jersey
The author’s story about getting useful, near-ready Python out of AI really resonates—especially the part about iterating until the output matches what you actually need. It also highlights how quickly “AI coding” went from novelty to a practical workflow tool. If you’re dealing with spreadsheet data and want to move from Excel into more programmatic forms, I can see why an “excel to json” pipeline would matter a lot. I’ve been meaning to try TableConvert for exactly that kind of table conversion. Overall, great write-up—clear, grounded, and inspiring without hype.