Don’t Let AI Kill Your Brain

Written by Taryn de Meillon

I used to work at a B2B tech company focused on healthcare in Wyoming, USA. It was a seriously interesting position and I learned a lot but I decided I wanted a change. I had an opportunity to move to the UK, to spent time with my English family and look for a job based in Oxford or London. During the final week of the job, I discussed my future plans with my mentor. I mentioned that I was planning to take a website development course as I felt this was an important skill to master. He was adamant that taking this course would be an absolute waste of my time, and advised me that instead, I should learn how to use AI to build a website. ‘This is what everyone is doing now,’ he said. ‘Knowing how to build a website, how to code, how to analyze the content … all that will soon be obsolete.’

This was probably good advice, and it was coming from an industry expert. Nonetheless, I decided to roundly ignore it. These kinds of choices don’t have to be binary. Why not do both?

So that’s what I did. I spent two months learning the ins and outs of the website backend, and I used AI to enhance the process, to make my workflow more efficient. I eventually began using ChatGPT to write code for custom CSS and HTML adjustments. Once I began implementing changes, there was a glaringly obvious benefit to this: if AI wrote code that didn’t fix the entire problem, I was able to identify the issue myself and prompt ChatGPT to improve on that specific issue. Had I not spent the time and energy learning these concepts, it would have been an endless merry-go-round trying to resolve issues, with new ones constantly popping up.

MIT recently conducted a study concluding that students who used a large language model (LLM) when writing an essay demonstrated lower memory recall and weaker brain activity compared to students who wrote an essay without using AI. This is an over-simplified explanation of the entire experiment, and to learn more, I would highly recommend reading the full MIT AI study here.

What we can gather from my own anecdote and MIT’s study is that if we begin to rely too heavily on AI to perform for us rather than use our own skills, we will lose out in two ways:

We won’t know how to correct it when it gives us unsatisfactory results.
We will quite literally become dumber, as our brain’s neural pathways will be less developed. Norman Doidge’s amazing book The Brain That Changes Itself explains succinctly that the brain operates on a ‘use it or lose it’ basis.

While I fully endorse using AI as a tool, i.e. to shorten the timespan of a project or to reduce tedious or automated work, I caution readers not to blunt your own skills by blankly using AI. Your know-how will set you apart from those who only know how to write a good prompt; they will allow you to collaborate with AI for a far more powerful outcome; developing your skills will make you a more intelligent, well-rounded human being.