In my previous column, I outlined a three-step approach to optimizing workflows before integrating AI tools. I showed how capturing relevant information, developing a process, and consistently executing it can lead to effective AI deployment. Now, I'd like to take a step back into the significance of cultivating your own intelligence in the face of rapidly advancing AI tools.

According to the Oxford English Dictionary, intellect encompasses the ability to think logically and comprehend concepts. Breaking it down, you find three distinct components: the capacity to learn, logical deduction, and the ability to make connections. This three-part system shapes and sharpens our minds.

However, the advent of AI tools has gradually shifted this mind-forming system. AI tools can now generate images, summarize text, write lines of software code, and even complete entire PhD theses.

Consider this: if you allow every idea you can't complete to be finished by an app, every sketch you can't draw to be auto-completed, and every knowledge gap to be filled by a prompt, you'll gradually cease developing the very mental muscles that generated those ideas in the first place.

In AI meeting tools, experimentation matters more than automation

So, how can you regain control? The solution isn't Luddism, but rather deliberate rebalancing. Begin with a short AI detox by selecting a daily window of at least ninety minutes where you disable all generative AI tools and force yourself to work with paper, a whiteboard, or a blank screen.

During this time, capture raw observations, draft rough arguments, and sketch connections without relying on suggestions. This is not punishment but a training regimen to build attention, recall, and associative thinking muscles that only grow when exercised.

Next, develop a simple triage framework that distinguishes between tasks suited for your intelligence and those that can be handled by the machine. Tasks that demand learning, synthesizing novel connections, ethical judgment, or persuasive explanation are your domain; repetitive transformations, surface research that can be verified, and layout or formatting are where AI can assist.

Practice explicitly switching modes by dedicating an hour to "you-first" session where you resist autocomplete and another hour to "AI-assisted" session where you refine and scale your projects. Over time, you'll rediscover the value of mental friction as a productive force.

Establish a daily habit of focused, incremental practice. Select a narrow subject and allocate ten to twenty minutes each day to reading primary material. Then, spend another ten minutes explaining it in writing without references. Utilise teach-back to explain the concept aloud as teaching forces organization, uncovers gaps, and enhances long-term retention in ways passive consumption never can. After each creative session, allocate five minutes to note your decisions, the reasoning behind them, and where you relied on intuition versus explicit rules.

Has AI finally solved your email problem?

Maintain a concise "decision log" that tracks your reasoning evolution. This creates evidence of improvement and highlights habits you want to retain or eliminate. Complement this with small projects that require integration, such as writing a short essay, designing a simple product, or building a tiny script where you intentionally avoid using templates or AI from the outset.

Only after you've completed a comprehensive human draft should you invite the tool to critique, summarize, or polish your work. Employ AI tools as amplifiers by requesting them to challenge your premises, provide counterexamples, or suggest overlooked edge cases. By consistently practicing this detox, you'll not only survive the rise of AI tools but also enhance their value as machines tend to amplify whatever intelligence they are paired with.

Do you need an AI detox?