Ever since the emergence of AI, back in 2022, as we know and define it nowadays, people have dreamed and dreaded of working with it on a day-to-day basis. The promise of AI is a magnificent one: to have an all-knowing, all-capable, never-tiring colleague – one that is always on your side, always listens, and can do whatever you tell them to do in an instant. But is it truly so?
The current trend of AI is closely following the Gartner hype cycle curve, in that, over time, hopes and dreams have been replaced with fear and disappointment, and then, with casualness. But this devious, ever-advancing entity is creeping deeper into every aspect of our lives, like a fungus.
This collaboration raises serious questions of philosophical and moral aspects, though. Is it good for us in the long run? What will happen to us, our minds, and, most importantly, our bonds with each other?
Let’s explore various aspects of AI’s effect on our lives and interactions within the workplace and try to untie this knot.
Productivity booster
AI is undoubtedly a fantastic tool for productivity. It can turn a 10x engineer into a 100x engineer overnight. It’s extremely capable of automating problems where the solution is already well-known, reducing the cost of execution to almost zero. As a result, tasks that previously required time and mental resources are now solved with a few taps, leaving time and resources for even grander things. This allows humans to make a few weeks’ leap in a matter of a few hours.
One of the simplest approaches, yet undoubtedly still effective, is to basically use various AI chatbots to solve problems and provide code for multiple tasks. However, this method can be improved upon much further.
At PlantIn, we have observed that AI performs best in coding tasks not through tricky prompting, but rather through providing it with the correct information, context, and patterns. To do so, we are using AI-centric code editors, like Cursor, which can be used over almost any domain and language. These tools provide an easy way to supply as much context as needed and then act upon it.
A typical workflow would include pointing out all the files we assume are relevant to a task, then providing AI with context on how the system works in general, what the desired outcome is, and guidance on what to look out for. It’s best not to go into too much detail on the specifics of the execution – unless AI goes sideways – as it might propose elegant solutions that will surprise you.
Another great application is solving the blank page problem. Sometimes you just need to get past this initial nothingness and then build onward from there. This is the best task to be handled by various AI agents, like Claude Code, which can spin up almost anything from scratch, guiding and asking for details, but building autonomously and in the background. This leads us to another important topic.
AI agents and the future of interfaces
While agents are themselves a highly overhyped theme, there is a lot to learn about ourselves and the future of our interactions through them. Essentially, agents are a reflection of what we want from AI: to be able to handle things on our behalf, as we want it, but without bothering us too much. It’s pretty clear that at the current stage of technology, it’s still impossible to build a single entity to handle everything. As a result, various tasks are being split among agents, which then communicate with each other to achieve results.
But here’s an important thing to consider: agents are, by definition, zero-UI. Their whole existence and purpose are to make our lives easier and do things autonomously. They don’t need any buttons, forms, cards, or any other UI elements. All they need is a task and the tools to accomplish it. If anything, they might go over existing interfaces and use them without our involvement.
One can’t help but wonder about the future of the interfaces themselves. If we don’t need to see or press anything, and all the interactions with machines are done using plain language, then what place does UI have in this upcoming world? And while it’s too soon to answer for sure, as so many usages are still not covered by AI, we can already see that one day, all we might need is a single chat window.
But even given these futuristic expectations, one thing is sure and should always be emphasized: the human is still the central point of the whole system. The focus has just moved from writing to thinking, validating, and planning. AI tools merely enable us to use more of our mental and time resources on what we were always meant to do.
AI beyond productivity
If a person can do in a day what used to take a month, does it actually help this person in particular? Unfortunately, at least as history tells us, it’s implausible. What really happens is that this level of efficiency becomes expected – a baseline from which any deviation signals that something is wrong. I am, of course, referring to the Industrial Revolution, where instead of granting people free time to do what their hearts desired, it drastically raised the expected output per hour, and people became entangled in even more work than they’d seen before.
In the meantime, one should also consider the long-lasting impact on critical skills: memory, problem-solving, and independent thinking. One of the most common usages of AI in the workplace might look something like this: “I have a half-formed thought, help me finish it,” or “here’s the context of the situation, help me decide who is right,” or “help me write a message about <topic>.” While using AI for such tasks might significantly reduce time spent and maximize outcome, it can also lead to the atrophy of critical soft skills like clarity of expression and candor.
These use cases might also raise valid questions about the value of the information exchanged. If a text was generated in a few seconds, is it even worth reading? And if it can be summarized, then what was the point of writing it in the first place? It’s easy to forfeit our humanity in such interactions, but here’s a good compass to be guided by: when applying AI to any task, ask yourself first: Would doing it myself make me better? Am I trying to save time or use a shortcut to the result? Would I be able to reproduce it given more time?
If the answer to any of these is no, then you should not use AI in this case, not because of fear or rejection of the technology, but because you will miss out on learning something important, something that will stay with you for the rest of your life, no matter how the world changes. That said, AI is a fantastic tool for learning, but you must make sure that you understand what it does and always double-check the references.
It’s even more critical for junior specialists. You can’t bypass years of knowledge and experience without paying a significant price. Using AI in the early stages of a career can mean missing the acquisition of valuable knowledge and practices that distinguish high-level professionals from others. Ultimately, these can only be acquired through years of hard work solving challenges. This shift in expectations also fuels a growing myth in the modern workplace.
AI-first Companies
There are reports of companies halting hiring if people can’t prove that the work can’t be done using AI. While this is a valid approach in itself, one can’t help but wonder – how does the effort compare to the result? How many hours and how much effort should one spend to automate something with AI? And what would be the cost of support?
An answer to that can be found very easily: just explore job openings at the companies that actually create AI systems. Those companies still hire humans. So if they are not able, or not willing, to replace the workforce with machines, there must be a good reason.
At PlantIn, we strongly believe in people and their value. We would rather have people achieve ten times their productivity using AI tools instead of spending enormous effort to build an AI replacement and still fall short of human quality, while also creating systems that are harder to maintain. It’s only natural that when people can drastically boost their effectiveness using AI, the AI itself would never be able to replace anyone or be a preferred choice over a human.
This doesn’t mean that we aren’t using AI – quite the contrary. Most of our product workflow, except for hypothesis creation, validation, and analysis, is performed virtually autonomously. We have dedicated automations that move tasks into appropriate states, copy information among platforms, gather results, assign executors, and compare metrics.
Such an application saves us tens of hours of manual button pressing and data manipulation, but we do it to empower people to achieve even greater things, not to pretend that AI is the core of our values.
Summary
AI has already impacted our lives, it’s not going away, and it will only grow more influential. As with everything in our lives, the key to success is moderation. Using AI to amplify your skills while also providing a backbone and line of thought unique to you is the most effective way to navigate the modern technology landscape.
The journey still comes before the destination, and no shortcuts, even AI-generated ones, can change that, because there is nothing to alleviate in the first place. Sometimes, the journey itself is what builds us. And while it’s doubtful that AI will completely replace humans, we should learn to live alongside it without sacrificing our identity or our future.
