BYOAI: Bring Your Own AI

I live at a crossroads. Literally. A four-way stop with slightly obscured signs and just enough traffic to make things confusing. Most days it works out fine. Drivers hesitate, wave each other through, inch forward. But every so often, confusion collides with impatience—and metal collides with metal.

One day, I watched a mechanic’s pickup get clipped in that intersection. The impact scattered his tools across the pavement—wrenches, sockets, screwdrivers sprayed everywhere. As he scrambled to gather them, still shaken from the crash, he worried: “These tools are my livelihood.”

That moment stuck with me, and made me think of how AI tools are becoming the new livelihood for knowledge workers.

Bring Your Own AI

We’re entering a new era where knowledge workers may be best served—or even expected—to bring their own AI tools into the workplace. Call it BYOAI: Bring Your Own AI.

This shift could actually be beneficial. If workers adopt and master AI on their own terms, they’ll shape how it affects their work. Better for workers to build AI fluency themselves than to have it imposed top-down by management. Otherwise the tools may define the workers, not the other way around.

One possible road ahead is that those who control the AI tools will control the labor power. Workers who come to work equipped with their own AI tools will have an advantage—not just in productivity, but in defining how their contributions are valued.

For the inspiration on this point, I recommend Prof G’s podcast episode One AI to Rule Them All, with economist Justin Wolfers (specifically here:  https://youtu.be/EklEzXBQP9U?t=1949).

History Rhymes

Of course this isn’t the first time that technology has had an impact on labor, and one can overreact with the analogies. The industrial revolution and robotics later displaced muscle power, but they didn’t eliminate work entirely. Instead, new specialized roles emerged. Our mechanic, and machinists, riggers, and technicians are now not valued for their brawn—they’ve valued because they have the right tools, and the expertise to wield them.

Those tools aren’t just implements. They’re also signifiers of expertise. To show up without them is to show up unprepared and appear unskilled. That was part of our mechanic’s worry.

Software engineers already live in a version of this world. Many bring to work their own advanced editors, debuggers, profiles, and scripts. Those who pick their own additional tools and become proficient in them stand out as more productive with their peers and managers. In the same way, adopting and mastering AI will become the next signifier of quality—the proof that an engineer is experienced and future-ready.

Building the Muscles

At this year’s Sage Futures, Katie Drummond, Global Editorial Director at WIRED Magazine, made a compelling case for using AI in our personal lives. She suggested simple, everyday experiments—like planning a travel itinerary with AI—as a way of building the muscles we’ll need in an AI-empowered world.

Just as our mechanic might use his tools at home or on side projects, knowledge workers can strengthen their AI “muscle memory” by applying it in low-stakes, personal contexts. Planning a vacation itinerary. Organizing a dinner menu. Drafting a home budget. Each exercise makes us more fluent, more confident, and more skilled in the interface between human and AI.

The workplace applications will follow. But they’ll be far easier if we’ve already built comfort with the tools in our own lives. BYOAI starts at home.  And while the educational applications of generative AI are, um, problematic today, one can fully expect future students (our future colleagues) will be building those AI muscles in their classes and internships in the near future.

The AI-Human Interface

In manufacturing, there’s constant discussion about the robotic-human interface—how people and machines share space and tasks. In software, we should be talking about the AI-human interface.

Prompt engineering, fine-tuning models, chaining workflows, integrating APIs, understanding bias—these aren’t abstract topics. They’re the new specialized tool-using skills. The same way a rigger knows which sling to use, or a machinist knows how to align tolerances, the worker of the future will know how to coax the right result from an AI system.

Those who master this interface will become the machinists and riggers of the digital age. They won’t be replaced by AI. They’ll be the ones who can make AI work. More than just capable—they’ll be valuable, adaptable, and future-proof.

And as signifiers go in any form of knowledge work, one stands out above all: the ability to learn over time. That has always been critical to a successful modern career, and AI only heightens its importance.

I like to say: use humans for what humans are best at, and computers for what computers are best at. But technologies also change us—personally and societally. Social media and algorithms reshaped how we consume information, for better and for worse. AI will have similar unexpected impacts.

In my AI 1.0 days with applications of neural networks, we always observed emergent properties of the AI – classification or predictive capabilities that were side-effects of the application they were trained for. I’ve noticed something interesting with generative AI: our engineers who are practicing better prompts with generative AI are reporting that they’re asking better questions of users during requirements analysis. The same muscles they’re developing in AI carry over into other tasks. Mastering AI is improving their abillities to listen, probe, and clarify more effectively. That’s a positive emergent property of the AI-human interface.

The Crossroads Ahead

Over the course of my career, I’ve lived through the PC era, the rise of the internet and the web, midrange systems and UNIX, the first wave of AI, search and social media, open source, Web 3.0—and now Generative and Agentic AI. Every one of those moments felt revolutionary and filled with positive potential (ok, maybe not Web 3.0).

But this time it feels different. This time the excitement is mixed with a host of side concerns that have nothing to do with technological potential: environmental impact, copyright and ownership, security, truthfulness and accuracy, productivity and job loss, confirmation bias, mental health, ethics. I find myself wondering not just where we’re going, but whether we even know where we are today. Maybe that’s great – thinking about future implications of technology beyond economic benefits before negative consequences emerge (<cough>social media</cough>).

So let’s land the analogy: we’re at a crossroads. The signs are obscured, the traffic unpredictable. No one really knows where the roads lead—not companies, not policymakers, not AI users themselves.

When the mechanic’s pickup was hit at the crossroads, his tools scattered across the road. And in that moment, it was painfully clear—without his tools, his livelihood was gone no matter what path he took.

Knowledge workers now face a similar future. AI is not optional. It’s the new set of tools. And you don’t want to find yourself without them when you need them most.

The question is whether you’ll wait for someone else to hand you those tools—or whether you’ll show up with your own.

Whatever path we’re on: Bring Your Own AI