AI works best as a helper—but we’re crossing a line where it’s quietly becoming the middleman. From call screening to hiring, AI is increasingly talking to AI, removing human signal from the process. Résumés are written by AI, read by AI, and ranked by AI. Nothing about this is unethical—but it is deeply disconnecting. The real question isn’t whether this is efficient. It’s whether we’re willing to pause long enough to rethink how humans stay in the loop.
If I can reflect for a moment on a topic that keeps resurfacing for me: AI will mature.
Not in some sci-fi way—but in the same way every once-fun technology giant has. Google became Alphabet. Facebook became Meta. Twitter became X. AI isn’t new, and that’s about as far as I’ll go in defining what it is or isn’t.
What interests me more is AI’s impact—and how unclear its return feels for people and companies alike. In many cases, it resembles a Super Bowl commercial: expensive, impressive, and difficult to measure.
I often think of AI as an agent—or better yet, a highly trained dog. A good one can do work you can’t, but its purpose is still to support the owner. Reaching that level requires good “genetics,” training, feedback, and experience. So far, so good.
A simple example is call screening. I let my call screener handle an unknown number while I keep drinking my coffee. Most of the time, the call doesn’t proceed and the system has done its job. But there’s an assumption baked in: the caller is likely unwanted. If they hang up, I’m left without enough context to know whether I should call them back.
Now extend that scenario to robocalls—or more interestingly, to calls you would benefit from. At some point, your screener needs to know friend from foe.
That’s the moment we allow AI to think, decide, and communicate without us.
Professional platforms have been around even longer than social media. LinkedIn is the largest player here, and it highlights a growing problem. Today, nearly every résumé is AI-enhanced at a minimum. My AI helps generate or refine my résumé. Your AI reads it. Another AI ranks it against thousands of others—also AI-generated.
Where does that leave us?
This isn’t hypothetical. It’s gravity. If AI can generate a résumé faster than I can write it—and review it with more confidence—it’s a no-brainer. If a company can use AI to narrow an overwhelming applicant pool, that’s also a no-brainer.
There’s nothing unethical about either choice.
But where does it leave both sides?
Disconnected.
At this point, résumés may not even represent the people behind them. They’re no longer written—or read—by the humans involved. If that’s true, then we’re overdue for the next phase of professional matching.
The question is simple, but uncomfortable:
Are we willing to take a step back, so we can all actually move forward?