Efficiency, AI and Meaning

Recently, I’ve been trying to make things more efficient. After all, who doesn’t want to create more with fewer resources? But “efficiency” means different things to different people.

I work in tech, where companies are pushing AI-first workflows, not necessarily because they benefit employees, but because they scale faster, cost less, and look better for marketing.

Here’s what concerns me. As Avery beautifully put it:

“One day, you may find that the part of your job you loved is now the part AI does better and faster. What’s left for you? The boring bits.”

That’s why, although I value efficiency, I refuse to give up what makes me human: creativity, critical thinking, and a sense of ownership. I want the code I write still feel like mine: the style, the decision, even the bugs.

To me, AI is like a pair programmer, assisting me with suggestions, arguing about techniques (many times it ends up saying “You’re absolutely right!”). At the end of the day, I’m the one who decides what goes into the final product.

The same applies to companies I want to join. I skip those that don’t resonate with my values, especially the ones forcing AI buzzwords into their products just to sound innovative.

“The best AI products don’t start with “Build me an AI tool!” They start with real customer problems. If the solution happens to be AI, that’s when it works.” - someone wise

Recently, I came across a company providing platform to train robot brains through VLA models. As I understand it, their platform helps robots learn faster through simulation, feedback loops, and continuous improvement. They abstract away the infrastructure complexity so people can focus on real-world behavior and creativity. By accelerating innovation and lowering materials cost, that’s when efficiency feels meaningful to me.

“Efficiency itself isn’t the enemy. Efficiency without meaning is.” - Avery

Another concern with training AI is its environmental footprint. Data centers consume massive amounts of water and energy. However, when compared to training robots in physical environments and wasting materials, virtual simulation still feels like the more sustainable path. And if it makes complex technology more accessible, that’s a win-win.

The robots I imagine building

vs the robots people are talking about

At the end of the day, real efficiency is earned, not assumed. Although I value it, I’m aware we can’t optimise what we haven’t explored. So, things are messy right now, and that’s okay. Through trial and error may we find meaning in the process.