Three Questions to Ask Before Your Business Adopts an AI Tool
For a while after I was diagnosed with ADHD in my thirties, I wondered if I was losing my mind. I'm one of the most organized people I know — voice memos, phone notes, computer notes, alerts for everything — and it still wasn't enough to keep my brain from dropping the ball. That fall, in 2023, I was teaching one of the first AI prompt-writing classes in the country, carrying a full course load and a research agenda on top of mom duties: the family's schedule, the kid's schedule, everyone's mental load. When Google's early Bard chatbot came out, I started using it to untangle my own messy thinking, and something clicked. A tool that could take a mess of ideas and hand back structure wasn't just useful to me. It could normalize what society still quietly labels a weakness, the same thing we ask for Americans with Disabilities Act (ADA) accommodations at work or a 504 plan or an Individualized Education Program (IEP) at school to manage. Used to assist rather than replace our own thinking, that kind of tool could help equalize society instead of dividing it further. I've been chasing that idea ever since.
Two Very Different Guys, One Familiar Warning
Last week, I read Bill Gates' new essay on the choices we're making about AI right now, and it sounded a lot like something Neil Postman would have written. Postman was a media theorist who, along with his mentor Marshall McLuhan, founded the field of media ecology because they'd watched earlier waves of technology reorganize society before most people noticed it happening. Postman died in 2003, so I'm pulling from his 1998 address, “Five Things We Need to Know About Technological Change," Reading Gates' essay next to Postman's decades-old warning, I kept landing on the same question: Are we going to make deliberate choices about how AI gets built into our workplaces, or are we going to let the choice get made for us?
What Gates Sees Coming
Gates' argument has three parts. First, AI will erase entry- and mid-level jobs faster than any past transition, hitting law, customer service, medicine, software and manufacturing within a decade instead of the several generations it took the United States to shift from agriculture to office work. Second, the same tools that help people can hand real power to people who mean harm, from AI-enabled fraud to attacks on hospitals and power grids. Third, he worries AI companions could quietly replace the harder, more valuable work of learning to relate to other humans, especially for kids. None of this makes him a doomer.
He's just as clear that AI could become the single biggest equalizer humanity has ever built, cutting the cost of expert-level help in health, education and government services for people who currently can't afford any of it. Whether we land on the equalizer side or the injustice side, he argues, comes down to choices made on purpose. His asks are concrete: build new institutions to manage the transition now, protect certain jobs as what he calls “Human Reserved” work we keep for people on purpose, and rebalance how we tax labor against machines so retraining has a funding source.
What Postman Warned Us About
Postman saw a version of this coming decades before generative AI existed. In his 1998 speech, he argued that technological change is never additive. It doesn't just add a new tool; it reorganizes an entire culture. And the people driving that change fastest are rarely elected officials or philosophers. They're businesspeople chasing what a new technology makes possible, even when it means overturning the institutions and traditions they claim to value. His example still lands: He argued that the single greatest force reshaping 20th-century American education wasn't a philosopher like John Dewey, but the quiet engineers who built the standardized test. The SAT and the GRE didn't just measure learning. They redefined what counted as learning, and schools reorganized their curriculum around the test rather than the other way around.
That's the pattern Gates is trying to interrupt before it repeats at society's scale. Postman's point was that we rarely get a vote on these reorganizations. The technology arrives, a handful of risk-tolerant founders decide how it gets deployed, and by the time the rest of us notice, the new arrangement already feels normal. We've watched a version of this play out before, in the early social media era, when companies raced to connect the world first and dealt with the guardrails later, well after real harm had already been done. Gates is explicitly trying not to repeat that script. He's arguing the guardrails need to be built alongside the tool, not promised for later.
My own ADHD story is the best-case version of Gates' argument: a tool that helps a brain that needs external structure, deployed with guardrails and used to assist rather than replace human judgment. The version Postman warned us about looks different — a tool rolled out at speed, with the guardrails promised for later, after the culture has already reorganized around it. Both are AI. The difference is entirely in the choices Gates says we still have time to make.
The Hiring Tools Are Already Proving Him Right
Postman's second warning was that every technology produces winners and losers, and the losers are usually the last people asked. We're already watching that happen with AI hiring tools, at a scale that touches millions of people far beyond any one office or industry. In 2023, the U.S. Equal Employment Opportunity Commission settled its first-ever AI hiring discrimination case, after a tutoring company's screening software was found to automatically reject older applicants. Since then, disability advocates and labor reporters have raised similar alarms about autistic and neurodivergent candidates being screened out by tools that score eye contact, speech patterns and résumé gaps against a model trained on what “normal” is supposed to look like.
That's Postman's standardized-test problem, running again in real time. A small number of engineers, working for companies most job seekers have never heard of, quietly define what counts as a qualified worker, and millions of people reorganize their résumés, their interview behavior and their sense of their own worth around passing a test they never voted on and can't see inside. For someone like me, whose ADHD already meant learning to translate myself into neurotypical formats just to get hired, this isn't an abstract policy fight. It's the difference between AI being the equalizer I found for myself and AI becoming one more gate that filters people like me out before a human being ever reads what I wrote.
Three Questions to Ask Before You Adopt Any AI Tool
Whose definition of “qualified” is this system quietly encoding? Postman's standardized-test warning applies to any AI tool making decisions about who gets hired, approved or believed. Ask who built it, what it was tested on and whose definition of “normal” it's scoring against.
Who's building this, and what are they willing to overturn to ship it? Gates and Postman agree that the people moving fastest aren't waiting for public permission. Before your organization adopts a tool, ask what human judgment, safeguard or accommodation it quietly proposes to remove.
Where's your Human Reserved line, and who drew it? That's Gates' term for the work you protect on purpose, not because a machine can't do it, but because a human should. If your organization hasn't named that line, the fastest-moving vendor will draw it for you.
None of these questions are anti-technology. They're the discipline Postman said a culture owes itself before it lets a new tool finish reorganizing it.
Where Prompt & Proper Comes In
I didn't grow up thinking of my ADHD as an asset. But sitting with Gates' essay next to Postman's decades-old warning, I keep landing in the same place: The tools that gave my brain external structure are exactly the kind of technology Postman warned us not to let a handful of fast-moving founders design in private. Equity doesn't happen because a tool is powerful. It happens because enough of us insist, loudly and early, on a say in how it gets built and who it's built for.
That insistence is the reason Prompt & Proper Consulting exists. I didn't start this practice to sell people on the newest tool. I started it because I watched the same pattern Postman described happening again in real time, with nobody outside a handful of companies at the table, and I decided the people who show up to my workshops and consulting calls deserve a real seat in that decision instead of just inheriting whatever the fastest-moving company built for them. If you want that seat for your own team, that's the work we do together.
Comments