Love it or hate it, AI does have the ability to make you considerably more ‘productive’ in the traditional sense of the word. i.e, you can get a lot more done in the same hours than you were used to doing a couple of years ago. And for many whose proficiency with language was found wanting, AI is a great way to bridge the gap with native speakers, allowing you to craft letters and emails and even documents that read well, are grammatically sound and may even have a certain tone or voice to them that you may find suits you well. And yes, easy access to powerful tools from OpenAI and Claude amongst others has democratized this technology enabling the average person to write haiku, and even code a website quickly.

What is fast getting lost is the differentiation normally evidenced by individual human thought. In a weird way, a global mind-meld seems underway because when everyone starts using the same LLMs (which are aggregated/trained on global data and information) the output (as long as prompts are similar enough) from someone in Kaula Lumpur is not particularly different from that of a person in Zagreb or Cleveland. So much is lost in this genericization of our work output because we are no longer using our individual brains except perhaps to ask the right question/prompt. The answers are being generated based on the larger compendium of human output over millennia.

I have found this especially irksome in some work-related situations which I highlight as examples of areas that people should NOT use AI for – because when you do it, you don’t set yourself apart. Yes, if you were well below average earlier, you are able to make your way up to a boring average now, but that’s just not enough anymore.

Emails – whether as part of a cold, sales outreach campaign or even internally from known business partners. When they all start sounding the same, none of them stands out or breaks through the clutter, leaving the reader cold and unmoved.

Interview follow ups. As part of our hiring process, we often ask candidates to submit written responses that are meant to go beyond the superficial, to allow us to get a glimpse of how they think, and formulate plans that are rooted in their reality and lived experience so we understand how they would do in a particular role. What we’re seeing now is generic AI-generated swill that looks the same from person to person, and tells me nothing about whether the person will actually be able to do the job. It also suggests to me the person might be a bit ‘lazy’, which AI tends to make of the best of us. Because it is tempting to get 80% of the desired output with 10% effort.

The problem now is that everyone is getting to the 80% output easily, but it’s only those who put their heads into the job and get to the 90% and 95% level that will genuinely impress.

And win the sale. Or land the job.

NOTE – this post was entirely human generated.