I want to establish something before we begin, because I can be misrepresented on this matter by certain people.
I know where the dishwasher soap is. I always knew it and I will always know where the dishwasher soap is.
As a matter of fact, some people even call me the dishwasher soap finder, the oracle of the bubbles, or the GSPS (Global Soap Positional System).
This is not a post about a man who didn&#39;t know where the soap is. This is a post about a man who knew, with complete and unwavering certainty, where a thing was, and who then spent two hours and eleven minutes trying to make sure that it was nowhere else besides where he <del>thought</del> knew they were. Those are different stories.
The dishwasher soap is under the sink, as it was known from time immemorial, and as shall henceforth continue to be known by generations to come. It has always been under the sink. Under the sink is where dishwasher soap lives, in the same way that Tuesday comes after Monday and how it is fine to forget the toilet seat up. You do not &quot;look for&quot; the dishwasher soap any more than you look for the floor.
<h2>09:14 The situation</h2>
The sink contained, in order of discovery: two pans, one nested inside the other in a way that suggested somebody had a healthy breakfast on his own, way more food that one should eat alone for brekky; a colander; a wooden spoon that has been slowly turning solid from the grey matter that clunked around it; four mugs; the lid of something; the pancakes maker metal frame full with burnt dough that now seemed soggy and gross with water embuing into it through time, and a plate sunk deep within the bowl used to mix the pancakes.
I turned on the tap. I reached for the bottle on the counter. Tried to clean those massive clunks of food and neglect off the pieces of ceramics to no avail. That is when it hit me! I should use the washing machine!
<h2>09:26 The plot thickens</h2>
Loading it up, while killing some flies that were smurking in the drenched waters of pancake dough, all was ready, but one minute detail... The soap was not loaded in! And as a hunter who sees his prey from up close just to find himself out of bulets, there was I with my dishes ready to load, but nothing on the chamber.
So I went under the sink. Confidently. The way you&#39;d reach into your own pocket.
Some people, at this point in the story, would say they thought the soap was under the sink. I want to be precise. I did not think it. If you had woken me at three in the morning and asked me where the dishwasher soap was, I would have said &quot;under the sink&quot; before I said my own name (which would be weird in any case, as in &quot;Why would you wake me up at 3am to ask my name? Are you having a seizure? Should I take you to the hospital?). No, it was not a belief. It was closer to a physical law. I knew where the soap was!
I mention this because everything that follows is consistent with it.
<h2>09:37 Diving inside the cabinet</h2>
I open the cabinet and more flies come out of their slumber. I fend them off, and kill them one by one with a martial technique I learned in Japan (or will learn when I go travelling there). My eyes dart into the baskets and their bags, orange, white, yellow... but no dishwasher soap.
The dishwasher soap was not visible.
I want to draw attention to the word &quot;visible,&quot; because I chose it carefully. &quot;Not visible&quot; and &quot;not there&quot; are different claims. I am making only one of them.
Meanwhile, the dishes were starting to stink more profusely inside the washing machine, as if they were souls lost into the quicksand or mud pits of a swamp, begging to be rescued from all of that filth.
<h2>09:38 Like Eminem, I am cleaning out my cabinet</h2>
I removed the contents of the cabinet to the kitchen floor, I did this systematically, front to back, and I laid everything out in the order it came out, like a man documenting a dig site.
Under the objects there was a layer of dust. Under the dust there was a pasta shape.
The cabinet was empty. There was no dishwasher soap in it.
I remained calm, just the voice inside my head started to shout in panic.
<h2>09:44 The decision</h2>
I had two options.
Option one: keep looking. This is unbounded. There is no natural end to looking. Looking can make you stare at an object in such a profound way that it makes you see their minor details and flaws. If you look at something just so attentively. you might even lose track of the meaning of the object, like when you repeat a word over and over so many times that their sounds detach from their meaning. You can look at something for so long that you might become part of that something.
Option two: go and buy dishwasher soap. This is bounded. It has a beginning, a middle, and an end, and it takes place partly outdoors.
I want to stress that this was not a retreat. I was not conceding the location of the dishwasher soap. I was procuring additional dishwasher soap, which is a completely separate initiative that happened to be running at the same time.
<h2>10:02 Shoes, umbrella and a <del>call</del> coat</h2>
The weather was threatening to rain, so I couldn&#39;t leave unprepared and come back soaking wet. Or worse, with the dishwasher soap bubbling on my hands providing me a good self-washing session. I put on the shoes, got the umbrella, now I needed just to grab my coat.
&quot;Where is my coat?&quot;
Nowhere to be found...
Now, I want to establish something before we proceed, because I can be misrepresented on this matter by certain people.
I know where my coat is. I always knew it and I will always know where my coat is.
As a matter of fact, some people even call me &quot;my own coat finder&quot;, the oracle of finding my coat, or the GECPS (Global Eric&#39;s Coat Positional System).
This is not a post about a man who didn&#39;t know where the soap is. This is... Well, you know what, maybe I should just ask my girlfriend where the dishwasher soap is, just to be sure...
And voilá! Have you ever realized how on the day you buy a red car, it seems that all over the streets all you can see are red cars, that were never there before? The same happened with the dishwasher soap. There it was, where I always knew it was, and where I will always know that it will be. But now it was visible.
Some of life&#39;s mysteries may never be solved. Like why Golden Retrievers are so easy-going and friendly? Or who is the secret creature that keeps putting the toiled lid up after I put it down? Seriously! Or how can it be, that the dishwasher soap (that, as it was established, I always knew where it was) suddenly reappeared after a brief exchange of words? Enchantments? Witchcraft? Or pure magic? I guess we will never know.
<h2>Conclusions</h2>
<ol>
<li>The dishwasher soap was under the sink. I said it was under the sink. It was under the sink. Whatever else happened this morning, these are the facts.</li>
<li>My coat is still at large. Any useful information that may lead to its capture will be rewarded.</li>
<li>The <del>dishwasher is dishing</del> dishes are washing. As expected, now the machine is loaded and firing away, with the soap that was where it was known to be.</li>
</ol>
Now, you may ask me: &quot;Who will clean the mess left on the kitchen while you were looking for the dishwasher soap?&quot;.
Well, my dear child, some of the life&#39;s mysteries may never be solved.

Where Is the dishwasher Soap

[Originally posted at <a href="https://cactus-now.com/cactus-news/vibe_coding/">Cactus</a> Blog]
<blockquote>
There&#39;s a new kind of coding I call &#39;vibe coding&#39;, where you fully give in to the vibes, embrace exponentials, and forget that the code even exists.
</blockquote>
— Andrej Karpathy
<hr>
Something’s changing in the way we build software. It’s not just new tools or faster frameworks, it’s a shift in what is possible to be achieved with the same resources. Teams are moving faster, thinking differently and achieving a fast output rate of software development that allow for startups to deliver MVP versions of their products to clients in weeks if not days in some cases. With all this output speed there will be more room for creativity, intuition and even fun.
This change is going to require a new mindset where flow and speed matter as much as process, where the best ideas often come from moments that don’t fit neatly into a sprint planning session. This isn’t just disruption by technology; it’s a deeper shake-up in how we write code and design software systems. And it’s already happening and we will start to see this outburst of more personalized and niched solutions flood the market soon.
<h2>Introduction</h2>
A lot has been said about the state of coding with AI and vibe coding. Even YCombinator, the reputed american accelerator, warns in this recent podcast about the risk of becoming obsolete if software companies don’t embrace AI-assisted development.
But what is it all about? And how does it work when we put it to the test and try to create an actual application using it? Is the quality just hype, or is it just a skill issue of those who criticize it? Let’s dive in and get some insight into how things went for us.
<h2>Our Experiment</h2>
<h3>Motivation</h3>
Companies often face a challenging choice: should they invest in a custom-built solution tailored to their exact needs, or go with an off-the-shelf product that covers most of the bases but might not solve their problem perfectly? It’s not just a technical decision, it involves budget, timing, and commercial strategy. Because at the end of the day, cash is the bloodline of a company and every company must keep a close eye on both costs and revenues. Making the right choice means balancing long-term value with short-term constraints. And that’s where best practices in software development and new technologies such as AI can give companies a competitive advantage.
In a world of limited resources, tight timelines, and relentless competition, speed is everything.
A faster, more reliable way to build software can be a game-changer.
That’s why we’re embracing AI-driven development to accelerate prototyping, boost flexibility, and cut the usual overhead.
With AI, companies can:
<ul>
<li>Move faster
</li>
<li>Test ideas more efficiently
</li>
<li>Deliver value sooner
</li>
</ul>
It’s not just about changing how we code. It’s about transforming how companies interact with customers, adapt to feedback, and evolve their products at speed.
So, we decided to put the hype to the test. We picked a pilot project and committed to building it entirely by riding “the vibe”, fully powered by AI, creativity, and lean execution.
Let’s see where it took us.
<h2>Goal of the project</h2>
The goal of our project was to create a platform that allows users to upload images taken from specific locations on selected beaches. Using computer vision, the platform analyzes these images to provide valuable insights into human presence and the distribution of objects in the area.
To achieve this, we built the application with a modern tech stack: React with TypeScript for the frontend, FastAPI (Python) for the backend, and SQLite3 as the database. The entire system is deployed within a Docker container to ensure portability and ease of deployment.
<h2>Setup</h2>
In our approach, we explored several tools during development, including Cursor, Lovable, V0, Windsurf and Claude. However, in this article, we focus specifically on our experience using Cursor, V0 and Claude 3.7.
We used the Pro version of Cursor, which provided us with sufficient Agent actions to carry out the tasks required for the project.
From our experience, the frontend code generated in Cursor didn’t meet our quality expectations. To address this, we turned to V0 by Vercel, a tool designed specifically for React and TypeScript. V0 impressed us with its ability to generate dynamic and well-structured designs right from the start.
Another highlight of our frontend development was the automatic inclusion of a dark mode option, a feature often considered a “nice-to-have” but typically left for the later stages of a project. In this case, it was integrated from the beginning, saving us time and enhancing the user experience.
<img src="/images/posts/ai-coding/vibe-coding-schema.png" alt="Vibe Coding Schema">
<h2>Process followed</h2>
To develop a comprehensive Product Requirements Document (PRD), we used Claude Chat to generate a detailed specification based on our initial functional goals. The application was expected to include user authentication, image upload capabilities, people detection via image analysis, and an admin dashboard to display the results.
Once the PRD was finalized and reviewed, we used it as the foundation for development and imported it into Cursor to guide the implementation.
During our initial iterations with Cursor, we made a common mistake: we submitted a complete prompt and expected the large language model (LLM) to handle the entire codebase accurately. While the LLM quickly produced a functional first version of the application, issues emerged when we attempted to scale the project, adding new features or fixing those that didn’t work as expected proved challenging. This highlighted the limitations of relying solely on the LLM for complex or evolving codebases.
<h2>The Challenges</h2>
This is where the real challenges began to surface. When we asked the tool to fine-tune or extend features, the AI often revisited and altered parts of the code that had already been resolved, sometimes reintroducing bugs, duplicating functions, or creating redundant and conflicting logic. The result was a progressively chaotic and inconsistent codebase.
But why does this happen? And is there a way to prevent it?
The root of the issue lies in the context window of modern large language models (LLMs), which is currently around 250,000 tokens. While that may sound sufficient, it quickly becomes limiting once the AI has to juggle a frontend, a backend, and a growing chat history. Once the model runs out of context, it simply “forgets” parts of the codebase. As the saying goes, “out of sight, out of mind” and for the AI, anything outside its current context might as well not exist at all.
<h2>The Solution</h2>
Vibe coding and AI-assisted IDEs can significantly accelerate development, especially when building prototypes and MVPs.
However, AI assisted IDE does not solve yet complex software architecture challenges. Human Expert insights are still essential. A seasoned software developer must ensure that any solution is scalable, secure, performant, and built following best practices. The landscape is evolving rapidly, and at Cactus, we are keeping a very close eye on these trends.
We like to think of our AI-assisted coder as a fearless junior developer, it boosts speed, scales our efforts, and helps us deliver faster. And yes, we have definitely decided to ride this wave!
Our application was successfully deployed thanks to the oversight of our senior development team, who thoroughly reviewed the codebase and requested corrections where needed, even in areas where we encountered difficulties, such as with Docker. Ultimately, this project became a valuable learning experience, teaching us how to effectively guide AI using the tools available today. The key takeaway? It’s absolutely possible to build solid applications with AI, as long as knowledgeable humans are involved in the process.
Much has been said about AI not yet being ready to code independently, and that’s still true. However, with the right approach and thoughtful engineering, AI can dramatically enhance your productivity and output quality. By following a disciplined, step-by-step method, yes, even if it’s a bit repetitive, you can get far better results than those who rely on ad hoc prompting and hope for the best.
If you’re curious about how vibe coding can help accelerate your prototype, validate your use case, or assess market potential, we’d love to hear from you. Reach out through our channels, and let’s explore how AI-assisted development can work for your team.
<h2>Yet Unsolved Drawbacks</h2>
One of the major challenges we faced involved working with specific APIs or tools for which the LLM had outdated training data, Docker being a prime example. The model frequently generated incorrect or deprecated configurations, and after multiple failed attempts to get something functional, we ultimately had to rewrite the entire Docker implementation manually.
This experience highlights an important reality about the current state of “vibe coding” with AI. Working with an LLM often feels like collaborating with an enthusiastic and fearless junior developer, eager to help, but lacking the discipline and caution of a seasoned engineer. If left unchecked, it will push untested changes to production, break the entire codebase, and force-commit without hesitation or remorse.
This makes the role of a senior developer more critical than ever. While AI can accelerate development, it also introduces a new kind of risk: the ease of deploying seemingly functional code that lacks proper validation, security, or adherence to best practices. What used to be difficult due to a lack of technical know-how is now deceptively easy to build, and potentially dangerous if not carefully reviewed.
<img src="/images/posts/ai-coding/crowd-shore-example.png" alt="Application Example">
<h2>Conclusion</h2>
Our application was successfully deployed thanks to the oversight of our senior development team, who thoroughly reviewed the codebase and requested corrections where needed, even in areas where we encountered difficulties, such as with Docker. Ultimately, this project became a valuable learning experience, teaching us how to effectively guide AI using the tools available today. The key takeaway? It’s absolutely possible to build solid applications with AI, as long as knowledgeable humans are involved in the process.
Much has been said about AI not yet being ready to code independently, and that’s still true. However, with the right approach and thoughtful engineering, AI can dramatically enhance your productivity and output quality. By following a disciplined, step-by-step method, yes, even if it’s a bit repetitive, you can get far better results than those who rely on ad hoc prompting and hope for the best.
If you’re curious about how vibe coding can help accelerate your prototype, validate your use case, or assess market potential, we’d love to hear from you. Reach out through our channels, and let’s explore how AI-assisted development can work for your team.

From Oblivious to Obvious!

Where Is the dishwasher Soap

An honest experiment in AI-assisted development

Pages

Get in Touch