In American folklore, there’s a legendary railroad worker named John Henry who raced a steam-powered drill through a mountain. Spoiler The story goes that John Henry beat the machine in that race by sheer grit, only to die from the stress of the effort.
The legend’s power has transcended its time: It’s been retold in songs, children’s books, murals, animation (Disney 1995), and labor movements. Unlike that 19th-century contest, we have an opportunity to change the ending.
John Henry is an early metaphor for automation anxiety — especially for Boomers and Millennials who grew up with his legend. Today, that same story feels strikingly relevant in the age of AI and robotics and we should talk about how to learn from it.
I confess: sometimes I “race” an AI agent at work.
I’ll queue up an AI (say, in a GPT-5 agent mode) to synthesize research or crunch data, and then I’ll tackle the same problem myself in parallel. It’s a friendly competition with me on one screen, the AI on another, to see who gets there first or who finds the better insight.
I’m not trying to prove I can beat the machine at all costs. Instead, I’m exploring how to work alongside it. In fact, on many days the AI is less an opponent and more like a teammate on a second monitor, handling one task while I focus on another. This approach flips the script on multitasking.
Research shows that when humans try to juggle tasks simultaneously, we pay a steep price, switching between tasks can sap up to 40% of our productive time, and it takes over 23 minutes on average to regain focus after an interruption.
In other words, traditional multitasking (constantly alt-tabbing between projects) is a productivity killer. But what if, instead of splitting our attention, we delegate one of those tasks to an AI agent? It’s like having a capable colleague or assistant on a second screen, working in parallel.
In fact, just having an extra computer monitor alone has been shown to boost productivity by about 42% on average in self-reported studies because it reduces context-switching. Now imagine one of those monitors isn’t just displaying information, it’s an intelligent agent actively doing work for you.
Offloading Tasks – The New Way to Multitask
Instead of trying to multitask in real-time, think of this as multi-tasking through delegation. An AI agent can handle time-consuming or routine work in the background, while you zero in on the work that truly requires your human touch.
For example: an executive might have an AI comb through market data and draft a report while they spend time on high-level strategy; a designer could let an AI generate a batch of layout suggestions while they refine the core creative concept; a software engineer could have an AI run tests or summarize documentation while they focus on critical architecture decisions.
In each case, the human isn’t idle – they’re doing what they do best, while the AI does the rest. This division of labor means less context switching and more deep focus for you, with the AI handling the background noise.
Here are a few practical tips to make the most of an AI “second monitor” approach:
- Identify Offload-Friendly Tasks: Pick tasks for the AI that are research-heavy, repetitive, or structured – the kinds of tasks that eat up time but don’t require creative intuition or complex judgment. This could be data analysis, initial drafts, research summaries, or QA testing.
- Maintain Human Oversight: Just as you’d check a teammate’s work, plan to review the AI’s output. You don’t need to stare at it constantly (avoid the temptation to micromanage your digital assistant), but be ready to course-correct if something looks off. Remember, an agent can go down the wrong path without guidance, a quick check can save time if it’s veering astray.
- Stay Focused on Your Task: While the AI works, resist the urge to constantly task-switch. Use that interval to push your own work forward. For critical projects, you might even pause and do “single-tasking” alongside the AI, both of you working on the same problem from different angles. You’ll reconvene with the results, much like teammates bringing inputs to a meeting.
- Leverage the Comparison: When you “race” the AI and later compare results, use it as learning feedback. Did the AI find an insight you missed? Did you catch a nuance it didn’t? This isn’t about declaring a winner; it’s about discovering how human insight and AI output can complement each other. You might even merge the best of both results.
This collaborative mindset addresses a common fear: Will AI replace us? The evidence so far suggests that while AI is becoming incredibly strong in many domains, it still lacks distinctly human qualities like creativity, empathy, and contextual judgment. Rather than displacing us, AI can augment our productivity – handling the busywork and allowing us to focus on more complex, higher-level tasks.
In the spirit of John Henry’s legend, our “human advantage” comes not from out-digging the machine, but from avoiding a direct fight on its terms. In other words, don’t swing the hammer until you drop – hand it to the steam drill when it makes sense, and move on to the next challenge.
Humans and AI: A New Kind of Team
What does this look like in practice for teams and leaders? It means reimagining AI as part of the workforce. Forward-thinking executives are already treating AI tools as digital team members – assigning them tasks, integrating their outputs into workflows, and even measuring how much routine load they can take off human employees.
Design teams use AI to generate drafts or parse user research, freeing the humans to focus on creative synthesis and decision-making. Engineering teams deploy AI agents for testing, code suggestions, and incident monitoring, so that human engineers can concentrate on innovation and problem-solving.
In each case, the AI agent serves as a “second monitor” to the human team, catching things or producing results in parallel, but the humans still drive the overall vision and ensure quality.
Crucially, embracing AI as a helper rather than a threat taps into what makes us most human: our adaptability. One author, reflecting on John Henry’s story, put it best
“the essence of our humanity lies not in our ability to compete with technology, but in our unwavering ability to adapt.”. - Mark Clifford
John Henry’s mistake (understandable in his time) was thinking he had to defeat the machine outright to secure his place. In our time, securing our place means learning when to collaborate with the “machine.”
In the end, the goal isn’t to have a race where either you or the AI wins. The goal is a faster, better outcome for the task at hand. John Henry won his battle but lost the war. In the modern workplace, we can choose a different path: let the AI handle the hammering in the background while you design the blueprint, analyze the strategy, or craft the narrative.
When your second monitor is an AI agent, you’re not really racing against the machine, you’re racing ahead with it. And that makes all the difference.
