Less Is More: Avoiding AI Overload and Making Smarter Choices

By Richard Sebaggala

The flood of AI tools coming onto the market every day is overwhelming. Every hour, a new tool is announced that promises to be better, faster and smarter than the last. Even for existing tools, there are constant updates, new integrations and add-ons. The result is a choice overload, leaving many unsure of which tools to use or constantly switching between them to try and keep up. But the real problem isn't finding the best AI tool — it’s understanding our cognitive limitations when dealing with an excess of information.

Alvin Toffler popularized the term “information overload” in Future Shock (1970), describing how the massive flow of data in modern societies can overwhelm individuals and institutions. Economic models now attempt to explain how decision-makers process large amounts of data despite limited cognitive resources. Milord and Perry (1977) define information overload as a situation in which “the amount of input to a system exceeds the processing capacity of that system.” The human brain is simply not designed to absorb, analyze, and process an infinite stream of information. When confronted with too much information at once, we are forced to prioritize — processing some information and ignoring others, often resulting in inefficient decisions.

Research in the field of cognitive psychology has practically highlighted these limitations. Miller (1956) proposed the theory of the magic number seven plus or minus two, which states that our working memory can only hold about seven separate pieces of information at a time. This limitation explains why, in a famous experiment by Kaufman et al. (1949), subjects could accurately count up to five or six dots flashing on a screen, but had difficulty when the number exceeded seven. Their cognitive process shifted from precise recognition (“subitizing”) to a rough estimate, illustrating the limits of our ability to process multiple pieces of information simultaneously.

This cognitive bottleneck is of great importance for the AI landscape. Every new tool, feature, or update competes for our attention, and decision-makers - whether researchers, professionals, or everyday users — must constantly filter, evaluate and select the most relevant information. However, as our processing capacity is limited, the constant introduction of new tools often leads to hesitation, doubt, and even paralysis. As Klingberg (2000) found, our performance deteriorates rather than improves when we try to process multiple tasks or inputs at once.

 

Back when I was doing my master’s, my econometrics professor had a simple but powerful lesson about learning and decision-making. He told us, “You don’t have to drive all the cars to be a good driver.” At that time, the market was flooded with statistical software — SPSS, EViews, Stata, R, and more. I chose Stata and considered it the Benz of statistical analysis. Over the years, I stuck with it, mastered its functions, and honed my skills. I have never regretted that decision. The same principle applies to AI. You don’t have to try every tool to be effective. You just need to choose the right ones for your needs and focus on mastering them.

Instead of chasing the latest AI trends, it's smarter to first ask yourself what challenges you want to solve. Are there tasks in your work or personal life that feel slow and repetitive? Do you spend too much time researching, writing, summarizing or organizing information? The right AI tool should help you streamline these tasks instead of making your workflow even more complex. In my experience, two tools are enough to handle most research-related tasks — Avidnote and ChatGPT. Avidnote is designed for researchers and makes it easy to  work on various research tasks given its expansive AI templates designed to streamline research process. ChatGPT, on the other hand, is a powerful, creative research assistant. If you use it with well-structured prompts, it can outperform many specialized tools on the market. The key to unlocking its full potential lies in developing strong prompt engineering skills. With the right prompts, ChatGPT can generate ideas, summarize content, improve writing and help with almost any research task.

One of the biggest mistakes people make is to constantly jump from one AI tool to another without fully exploring what a single tool can do. The reality is that most AI tools offer overlapping functionalities. The difference between one tool and another is often minor, and the time spent learning a new tool could be better spent refining the capabilities in an existing tool. Therefore, avoiding AI overload is not about finding the perfect tool, but about settling on a few good tools and mastering them.

 

AI marketing thrives on hype. Every new tool claims to be revolutionary, but does it really change the way you work? Before you add a new AI tool to your workflow, ask yourself if it really solves a problem or is just another distraction. The best way to filter through the noise is to focus on the practical benefits. If a tool doesn’t significantly improve efficiency or add value to what you already have, it’s not worth the time.

The approach to AI should be minimalist. Less is more. A well-chosen and well-mastered toolset will always be superior to an ever-growing collection of half-explored tools. Just as I didn't have to learn every statistical software to master data analysis, you don’t have to use every AI tool to benefit from AI. Focus on what will help you work smarter, invest time to understand it thoroughly, and resist the temptation to constantly switch tools. The real power of AI lies not in its novelty, but in how effectively you use it.