Don't Fight (AI) Gravity

As Rickson Gracie once said, flow with the go.

We're in a disruptive but ultimately incredibly exciting time right now. Perhaps, like David Foster Wallace's fish, we aren't even aware of how ubiquitous AI is becoming in our lives, workplaces, and everyday interactions. 

When the chat LLMs first blasted on the scene less than three years ago (can you believe it's only been ONLY 3 years?), I played with them as any good techno-optimist would. They were fun and novel, but I could not quite see the use case for me to get immense value from them. 

Fast-forward three years, and I think 80% of my Google searches have now shifted to LLMs. We are still in a bit of a "Dogpile search" moment of LLMs, where while we may have an LLM of choice, we probably all bounce between a few. At work, I am only authorized to use Copilot, so I use that, but for my personal life, I prefer Gemini right now. 

Regardless, this post isn't about all that it's about not fighting gravity and finding true north. By true north, I mean the real value in our work. The LLM disruption has caused me to really think about what value I bring as a product manager. I never thought it was documentation or data analysis; I always felt it was the "softer" (hate this term, but that's a blog for a different day) stuff: strategy, alignment, decision-making, taste, etc. 

So, why would I fight utilizing LLMs to complete PRDs I have created the substance for, help ideate on things I am missing, or clean up Excel documents and help prioritize features I already have ranked? Shifting what I consider to be low-value or busy work has opened me up to focus on much more strategic high-value things than I used to when I was "just so busy." I am not ashamed of that; I will shout it from the rooftops.

And why do I see other PMs online feeling ashamed for utilizing AI to be more productive? We are in the age of output, not input. If you spend 3 hours doing Excel work that an LLM can do in 38 seconds, well, in 2025, you're fooling yourself. And I have seen some PMs express that their employer might not value them as much if they knew they were using AI. I would say that you either have an employer problem or a value creation problem, not a "to use or not to use AI" problem.

Don't fight gravity. LLM copilots and agenetic AI are coming for every single low-value task possible, so the only option is to shift your focus to becoming world-class at high-value work. 

If LLMs can write faster than me, produce documents faster, or do data analysis faster, great; feel free.

I will stick to the human-centric, forward-looking, high-value work I enjoy anyway, and my advice is that you should, too.