Is Your HR Vocabulary Teaching AI to See Employees as Headcount?
The Words You're Using at Work Are Shaping More Than You Think
When's the last time you actually questioned the words that came out of your mouth at work?
Not the language on your careers page. Not your company values.
The automatic ones. The ones in your emails and your Slack messages and your Monday morning meetings that nobody ever put there on purpose.
Most HR professionals are working with a vocabulary they inherited.
Headcount. Leads. Prospects. Right-sizing. Streamlining.
Those words were already in the room when you arrived. And the longer you use them without thinking, the more invisible they become.
That invisibility is exactly the problem.
Joshua Berry spent two years turning that problem into a TED Talk. And this conversation, part of our Greatest Hits series because some episodes genuinely deserve a second listen, is where that idea started.
(00:00) The Guy Who Turned a Podcast Conversation Into a TED Talk
Joshua Berry was on the podcast back in June of 2024 for episode 183, "Challenging Beliefs: Dare to Be Naive."
That conversation apparently did not leave him.
He returned with news: a TED Talk delivered to a full audience in Omaha, Nebraska in November of 2024. He came back, in part, to say that this podcast played a role in shaking something loose that needed to be shaken.
Joshua is the director of Econic, an author, speaker, entrepreneur, and facilitator of change with two decades of experience working with Fortune 500 companies and venture-backed startups.
His work sits at the intersection of adaptive leadership, overcoming limiting beliefs, and the innovation systems that actually move organizations forward.
His question, the one that runs through everything he does, is deceptively simple: What beliefs are quietly running the show? And what are they costing you?
(05:12) "Nobody Actually Wants to Be Put in a Funnel"
Earlier in his career, Joshua led sales and marketing teams. The vocabulary was completely standard.
Prospects. Leads. Funnels. Targeting.
Nobody questioned it. It was just the language.
And then one day Joshua started asking: who actually wants to be prospected? Who wants to be targeted?
The answer, of course, is nobody.
His mentor gave him a framework that stuck: words are a symptom of our conscious or unconscious beliefs.
When you say "right-sizing" instead of "layoffs," that's not just a cleaner phrase. It might be a signal of something underneath. A belief about the replaceability of people inside a system.
That's what Joshua explores in his TED Talk. Not a moral lecture. Not a vocabulary lesson.
A genuinely provocative question about what happens when we stop noticing the words we reach for automatically, and what those words might reveal about what we actually believe.
The ask isn't to overhaul your language overnight. It's to get a little more intentional.
Notice what you're saying. Ask why.
(12:24) What Your HR Language Is Teaching AI Right Now
This is where the conversation shifts from philosophical to urgent.
Here's what you need to know about how large language models work:
They don't generate original ideas. They predict responses based on patterns in their training data.
That training data is built from enormous amounts of existing text -- emails, documents, blog posts, internal communications.
As publicly available training sources become exhausted, companies are turning to private and gated sources to train their models.
Which means your internal emails, your performance review language, your Teams messages could eventually be part of what a model learns from.
If those sources consistently describe people as headcount, expenses, or resources to be controlled, that's the model that gets built.
Joshua puts it plainly: if we've described people in dehumanizing ways in our communications for decades, what do we expect AI to reflect back at us?
We're at an inflection point.
The language choices HR professionals make right now aren't just about internal culture. They're potentially shaping what generative AI believes about the value of people in the workplace for years to come.
That's a different kind of reason to pay attention to the words you choose.
(18:27) What Would You Do With Infinite Interns?
The "is AI going to take my job?" conversation is everywhere. Joshua doesn't sidestep it.
His answer isn't reassuring in a hollow way. It's actually more useful than that.
Someone described AI's current capabilities with a thought experiment Joshua shares in this episode.
What would you do if you had infinite interns? That framing changes everything. Here's how to work through it:
Step 1: Audit what your role actually requires. If your value sits in information retrieval, policy lookup, and analysis tasks, those are the areas most exposed. Knowing that is more useful than resisting it.
Step 2: Identify what AI genuinely cannot replace yet. Relationships. Judgment. Context that lives beyond what's in the data. The human side of HR is still the hard part, and it's the part that matters most to the people you serve.
Step 3: Start experimenting before you feel forced to. Use a tool as a thinking partner, not just an output generator. Ask it what you're missing. Ask it why something won't work. Ask it to challenge your assumptions.
Step 4: Let the tool help you map your own risk. Paste your actual job description into ChatGPT and ask which tasks are most likely to be automated and why. Let the technology help you see the terrain. Then decide what skills are worth building next.
The point isn't to be naive about what's changing. It's to stop being reactive and start being curious.
(28:33) The ChatGPT Setting You've Probably Never Opened
Joshua shared something in this episode that Traci genuinely had never seen. And if you haven't either, here it is.
What it is:
Inside ChatGPT, there's a customization option that lets you set standing instructions. Things the tool factors in every single time you use it. Things like: treat me as an expert. No moral lectures. List sources at the end of your response. You control what goes in there.
How to use it to humanize your work:
Add an instruction like: if I ever refer to people in a dehumanizing way, flag it. One line. Now every draft you run through the tool, every memo, every message you ask it to review, it's watching for that pattern.
The other practical move:
Before you send an email you're unsure about, paste it in and ask: how might this land? Can you suggest a more empathetic version? It's the old 24-hour draft rule, with a built-in second opinion.
And maybe the most useful application of all: you can ask the tool to help you become a more intentional, more human communicator.
Which is a pretty good use of technology, all things considered.
The Real Reason This One Made the Greatest Hits List
The language we use at work is not neutral. It never was.
It carries assumptions about people. About their value. About how replaceable or irreplaceable they are.
And as Joshua lays out in this conversation, those assumptions are now getting baked into the systems we're all starting to rely on.
The full episode gets into so much more, including how Joshua's TED Talk came together, the specific connection between unconscious beliefs and word choice, and a genuinely fun moment where he shares his screen to walk Traci through the ChatGPT settings live.
If you want a conversation that makes you think differently about something you do every single day without noticing, this is the one.
Connect with Joshua Berry: Econic | LinkedIn
Connect with Traci here: https://linktr.ee/HRTraci