How AI is Impacting Knowledge Work: Observations and Thoughts
TL;DR
- AI simplifies routine, repetitive tasks, making work easier and enjoyable.
- It may limit innovation by encouraging reliance on ready-made solutions.
- AI helps improve overlooked areas, such as design and user experience.
- AI-generated work currently has imperfections and subtle oversights.
- Solo workers benefit significantly by maximizing productivity with fewer resources.
- Corporations might not see major productivity gains due to difficulties in measuring creative output.
- Cross-domain productivity still presents challenges.
- Long-term unemployment due to AI may not be as severe as currently predicted.
- Pure knowledge no longer offers a significant competitive advantage.
- Timeless skills like decision-making, communication, and critical thinking remain essential.
- Competitive advantages now hinge on time and effort rather than knowledge alone.
- Curiosity and observation are increasingly important skills.
Introduction
I’ve been trying to write down my thoughts on AI and how I feel it’s impacting the present – and possibly the future – based purely on what I see today.
Its capabilities may change drastically in the future, and since I’m not an expert, it’s hard for me to make accurate predictions. But I’ll still take a shot at it from the perspective of what I observe now.
AI and Knowledge Work
There are a few upsides I see, especially in the kind of work I do (which I’ve written about briefly here). Not just in programming – many parts of any job involve repetitive or mechanical tasks. For example, when I want to write a specific logic, I might not remember the exact syntax. I might need to write documentation, create test cases, or repeat certain steps – all of which can be tedious.
AI makes the boring parts of the job easier, even fun.
However, AI can also automate some of the interesting parts of the job. The joy of creating your own solution never loses its charm. People love re-inventing the wheel, and sometimes in doing so, they improve something or discover something entirely new.
AI can inhibit innovation, as people may stop asking questions and just rely on its solutions.
AI also helps polish areas of work that are usually ignored. For instance, take an admin panel – the backend of a web app. It’s often built quickly, with less focus on design or UX, since it’s used only by internal teams, who can be taught to live with it. But with AI, I can improve and refine the admin panel as well. Building UIs has become easier too, as HTML is full of repetitive code.
AI helps improve the often-neglected parts of your work.
That said, there are some downsides. One of them is the subtle deviations you might not notice if you give AI full control. I’ve caught several small issues in the code it generates – especially when using it in “agent” mode. It may not be as meticulous as a careful programmer. For example, it might create a new column with a similar name instead of recognizing that a similar (unused) column already exists – something a thoughtful programmer would catch and confirm with the team.
I’ve also seen it duplicate code or get stuck in loops when running commands. These are minor issues for now and could be resolved with time.
AI tools are not perfect yet.
AI and the Productivity Boost’s Impact on Jobs
I believe individuals who are working alone will benefit the most. These are people who want to do more with fewer resources and maximize their output per hour.
AI allows individuals to do more with less.
Businesses, however, might not benefit as much, because employee productivity can’t be measured accurately. Many employees today believe, “If I can finish my work in 4 hours, why should I work for 8?” With AI, they might complete it in 2 hours and then spend the rest of the day doing something else (outside of work). Creative productivity has never been easy to measure and likely never will be – unless a manager works alongside each employee, which is unrealistic.
Large corporations may not see significant productivity boosts, as people already have a set idea of what’s “enough.”
Cross-domain productivity will still be limited. For example, in a small 2-3 person team, if one person handles the business side, they can’t fully rely on AI to handle tech work. They would still need to spend time talking to the AI, taking time away from their business responsibilities. But the tech person, thanks to AI, can now do more. The business person, in turn, can focus on sales and marketing. So productivity within the team improves.
Since AI allows more people to build and scale businesses, we’re likely to see more companies emerging. These businesses will eventually need people.
So, my prediction is that long-term unemployment may not be as bad as some fear today.
The Future of Knowledge Work
The future of knowledge work definitely looks different. Some of the things I used to be proud of are no longer as valuable. The advantage of simply having more knowledge is fading.
For example, I once read the Laravel documentation end-to-end. That gave me an edge – I knew which feature to use in a given scenario. I’d say things like, “Oh yeah! I can push this as a job with a delay,” or “Let’s add a middleware to log all user actions.”
But now, I can just ask AI for all available options and get pros and cons for each. I can then pick the best one for the situation. Just knowing a programming language is no longer a guarantee of a bright future (just ask those who built careers in the 2000s by just knowing Java).
Just having knowledge isn’t enough anymore.
This makes me believe that the skills that were valuable 50 years ago will still be valuable in the future. You’ll still need to make decisions based on the options AI provides. You’ll still need to communicate your problems clearly and have the judgment to understand and evaluate the responses.
Also, AI has taken away the time advantage that knowledge once gave. Just knowing something and being able to solve it quickly no longer sets you apart. Unless you can build something unique with your knowledge, knowledge alone isn’t a competitive edge.
Now, it’s mostly a battle of time – how much time you’re willing to put in. Earlier, you could work an hour a day while your automations did the rest. Today, anyone with the will to use a chatbot can build those same automations. But if someone is willing to work 8 hours instead of 1, they can do 8 times more.
That’s why we’ll see small, focused teams outpacing large corporations. They’ll have their heads down, doing the work. They won’t just be “jiggling the mouse” to make it seem like they’re working.
In conclusion, it’s a strange time to be alive. We’re witnessing incredible things, but there could also be chaos if this change isn’t managed well. Will the people who lose jobs accept that they were unlucky? Or will they be willing to re-skill? (Maybe I’ll be one of them.)
Will we take this shift in good spirit? Will we be ready to adapt? It’s hard to predict – but we’ll need to stay curious and observant.
That’s why cultivating curiosity and observation is more important than ever.
