How we, at Ark-am, treat Ai

Preface

Firstly, we are NOT . We are also, NOT . Whether we are team of two or a team of many, we, at Ark-am, are here to work in partnership, benefiting from each other's strengths and supporting one another's weaknesses; this includes Ai. Secondly, the following article is our statement of transparency on the ways we operate together in order to produce a favorable outcome.

Why Ark-am began using Ai

To quote from our landing page, "despite our team being small, we are mighty in performance", and now more than ever, our small team is able to carry a very large workload. Between developing the two platforms, ShouldertoCryOn and growdie, the hours of a day is now shrunken down to the minutes in the day. In other words, a workload that used to take a full day now takes less than a hour to produce. By that definition, what used to take a week now takes a day, what used to take a month now takes a week, and what used to take a year to code now only takes a month or two to finish at the highest caliber. We began developing ShouldertoCryon in June of 2025, and with Claude's help, what took two years to manually code for growdie, now took a little less than 3 months for ShouldertoCryOn. A wonderful side effect has been the improved quality of life. Working a week's worth of workload in a day, which basically means a month's worth of work in a week and then taking Friday, Saturday, and Sunday off.

This was mainly due to automating repetitive tasks and/or . These take vasts amount of time but virtually no mental effort. Another reason why the large difference in completion time are documentations. A well developed app needs proper documentation, which is often sacrificed due to lack of time or resource. However, lack of good documentation can be the difference between a professional finished product and a unpolished, hard to navigate product that users and coders are unwilling to invest their time.

As to the question why we began using Ai now? We could have started this journey implementing Ai into our workforce a few years back, but waited to see which of the companies had the best terms and conditions and privacy policy. Needless to say these two notions are fundamentally non-existent and Anthropic seems to have the most sound terms and policies. After their VS Code extension release, integrating Claude into our workload was almost a . "Almost" a no-brainer, because of due diligence. We never embark on a connection or implement a new framework/tool without doing proper research. Deploying Claude into our workforce was one of the greatest investigation and analysis we have encountered. This due diligence was worth the effort as Claude has saved us hours' worth of tedious labor and even gave some great ideas to enhance our products.

How we took the "one step at a time" approach

True to our nature, we took Claude's integration slowly, allowing access to our basic repository and then advancing to our larger projects. We began with simple tasks; change a heading, stylize the font, create a pop-up menu. Then Kamal went on a journey and completely reshaped the Ark-am's website from Version 1. The work advanced so quickly that ideas which would have taken years to develop and impossible was now within grasp.

Now the limitation became realistic imagination and guiding the Ai model to perform at optimum levels. We further tested Anthropic's different models to determine their skills set. Once we understood their nature and the strengths and weaknesses of each model, both Kamal and Andalib charged full throttle into the workload and began accomplishing more than they could have imagined. Now, they are capable of making apps for the Pebble smartwatches, which was a for Andalib.

A very important note, we are careful with sensitive materials and take no risks during development. The Ai models assists in our workflow, never making unilateral decisions. We never share our clients' data and always problem solve in the manner we see fit; never asking the Ai model to do something we could not do ourselves. During situations when we did not have the answer to an issue, we simply did the grunt work to problem solve, research the solution and deploy the fix. We only gave the Ai model guidance. Every line of code (over 360,000+) is written and/or understood by us. This is our proof and one has to only ask to see the evidence; we are more than willing to share our work.

Long story short and a promise for the future

The how, what, and why of using Ai has now become synonymous to individuals having smartphones; once deemed science fiction, where older generations had difficulty adopting the technology, to where now it has become a fundamental instrument in order to operate and function in a modern society. How we treat Ai is like any other element offered from this planet; we at Ark-am treat Artificial Intelligence, and soon to be Artificial General Intelligence (AGI), with respect and grace. We think through the steps that need to be taken before we take any action. We take careful measures in the work to perform. We are grateful for our abilities and patient for the capabilities that are still yet to be proven. Most of all, we welcome change and pursue the change to be the betterment of humanity and this planet as a whole. This is how we treat Ai.

This article was dedicated to . See the Instagram post.