This Isn't the First Tool That Was Going to Ruin Us

What do the calculator, AI, and your paycheck have in common? Show your work.

When the calculator showed up in classrooms, people panicked. Kids will lose the ability to do simple math, they said — a fear that's been around since the 1980s, when parents worried their kids would become dependent on the machine and lose their conceptual skills (Historical Analysis of Attitudes Toward Calculators). The panic never really went away. But the evidence never caught up to it either — research reviews have found calculators can actually boost pupils' fluency and understanding of math rather than harm it (Maths — No Problem!).

The working theory, pieced together later: those kids were learning two things at once. The math. And the tool. They hadn't automated their number facts yet, so the calculator wasn't lightening the load. It was splitting their attention across two unfinished jobs.

Still — 62 x 78 = 4,836. There's no mental math going on there anymore. Excel does it, your calculator does it, even Word just filled it in for me as I typed this sentence.

What actually happened is this: we didn't get worse at understanding math. We changed how we tested the understanding of math.  NCTM pushed schools toward a "calculator aware" approach — not banning the tool, not ignoring it, but teaching kids when to reach for it and when not to.  The SAT moved toward real-world scenarios and interpreting graphs and tables instead of long calculations (Manhattan Review — SAT Changes). Common Core now requires students to show their work in multiple steps and justify their answers, not just write the final number (Khan Tutorial — Common Core Math Guide). The tool changed. So the measurement had to change with it.

Now we're doing it again. AI can research and write the paper for you:  that's even better than a calculator.  But this time, the testing for mastery is already evolving in step with the technology. Teachers are increasingly breaking assignments into stages:  proposal, outline, rough draft, revision, final submission, and grading each one, because a student's real writing leaves a visible trail of drafts and revisions that AI-generated text simply doesn't (Educators Technology — AI-Resistant Assessments). Show your work. Same lesson, new tool.  Just like with the calculator: choose WHEN to reach for AI and when not to. 

So here's the real question. If we know this pattern: the tool changes, the old measurement stops meaning anything, and we build a new one, what else are we still measuring with a broken ruler?

Compensation, for one.

Ford Motor Company built the 9-to-5, Monday through Friday, in 1926 — one of the first companies in America to adopt it, and the rest of the country largely followed (History.com — Ford's 40-Hour Week). It built the worker bee. And it came with a message: you're lucky to have this job. Your life is this job. Flash forward to 2020: Covid cracked that open. Corporate America was forced to reckon with remote work, and once that question was on the table, employees started asking a bigger one: what does quality of life actually mean to me? Before Covid, that question wasn't allowed. Not really. You didn't push back if you wanted to keep the job. Now, we do.

There's no doubt anymore that work can be done just fine from home or remote. If we’re home, we can watch our kids grow up instead of missing it. That's one of my favorite evolutions we've lived through, because we miss so much life, so many people, so much of the actual world, when we're chained to a schedule built for a factory floor a hundred years ago. But ultimately we're still grading compensation on attendance, not on the work.

Think about it: two weeks. The average American worker earns just 11 paid vacation days after their first year on the job,  barely more than two weeks (BLS — Average Vacation Days).   The modern vacation benefit was shaped less by concern for employee well-being than by economics: when wartime wage controls limited cash raises, employers used paid time off and other benefits to compete for scarce workers. (BLS- Compensation Trends)

 I say let compensation evolve the way math testing did  because the old ruler doesn't fit the new tool.   We still need to be paid, but that is no longer the only value or benefit an employer can offer - or what an employee is producing.

When I left a corporate job deliberately, intentionally and purposefully, my first step was redefining what compensation actually meant to me. Yes, I need money to survive. But what does money mean? For me, it  meant freedom. The freedom to choose. Agency over my own time. So when the big salary gave me money and took my freedom, I wasn't getting more. I was actually getting less.  I didn’t like that equation any longer. And I know I’m not the only one - many people have chosen entirely different methods to make a living and where they live (ie tiny homes etc)  So why hasn't the employee benefits package advanced or evolved too? 

Is 2 weeks paid vacation right for employees in 2026? 

Paid vacation policy has barely moved in over three decades. A federal benefits analysis found paid vacation allowances stayed essentially flat from 1992 to 2013, even as employers reshaped nearly everything else about how and where we work (BLS — Paid Leave in Private Industry Over the Past 20 Years). Two weeks was the offer then. It's roughly the offer now. 

Compensation is due for an evolution too. 

The shift is already visible. Americans have filed more than 5 million business applications in each of the last several years, far above the roughly 3.5 million filed in 2019. Not every application becomes an operating company, but the sustained surge shows that millions of people are trying to create new ways to earn outside the traditional employment model. But it does show that millions of people are no longer waiting for a traditional employer to define the terms of work, compensation, time, and freedom for them. (U.S. Census Bureau, Business Formation Statistics)

The calculator threatened our ability to learn and use mental math. Yet, we didn't lose math abilities. Instead we tested math at a higher level and changed when and how the calculator was used in teaching math. What we realized with calculators still applies here with AI: decide when and how to use it. If mental math testing evolved to become word problems, then AI writing has to evolve the same way: we can’t grade it just on the answer, we have to show our work - our process. Human-in-the-loop researchers found that when students had to check sources, catch hallucinations, and justify their edits, that oversight itself became the thing that built critical thinking, not a chore layered on top of the real work. The vetting is the learning. Ask AI to quiz you on what you wrote. Push back on its assumptions. That’s how you show your work.