I spend a lot of time thinking about things I don’t understand. Which means I spend a lot of time thinking about new uses for AI — things like Character.AI (one of the fastest growing online companies today) where you can have your therapist’s voice sound like Harry Potter.
That’s something I never even conceived of. Yet here we are.
Where? Here. At the peak.
On one hand, that’s very exciting. AI is absolutely spilling out everywhere, flooding the market and filling our heads with potential ways to make our lives better. But that’s the key word: potential. We have reached peak potential … not productivity. Yes, some things have absolutely gotten faster, simpler, better, but they’re still not perfect. So the hype is sky high, and yet I bet your life feels more or less the same as it did a year ago. Mine does. Which means the only place to go from here is down.
We’re headed towards this valley — the trough of disillusionment, as Gartner calls it — where expectations are being trumped by reality. We have all of these fancy new AI tools, but the truth is most of them aren’t actually making the world markedly better or different. At least not yet. Your therapist talking like Harry Potter definitely isn’t.
The real issue is the perception of why AI even exists in the first place. It’s starting to feel more like a box to check — a neat parlor trick — that exists simply because people think it’s supposed to, rather than actually being designed to solve specific problems. AI is not a new pair of shoes or Taylor Swift tickets. People don't want AI for the sake of AI. They just want to be more productive.
AI is certainly a way to achieve that. Perhaps the best way, if used correctly.
So as we plummet from the peak of potential and slide through the trough of disillusionment, it’s going to be critical to stop thinking about all the cool things AI can potentially do, and start thinking about how people — like recruiters and hiring managers — can be more efficient. Faster. Smarter. Better.
If we start there and work backwards, eventually we’ll reach paradise: the plateau of productivity — a place where stores are staffed, candidates are happy, recruiters are fulfilled, and nobody talks like Harry Potter.
Start with what people know.
Recently, my son came up to me holding a floppy disk he found in the closet and asked, “Dad, why did you 3D print the save icon?”
Which made me laugh, obviously. But it also made me think about why the save icon looks like that in the first place. And it’s because of this concept I love called skeuomorphism, where you design a graphical interface to mimic its real-world counterpart in how it looks and how you interact with it. In other words, making what’s new look familiar. There’s absolutely no reason the keyboard on your phone needs to be laid out in the QWERTY format other than it's what we’re used to.
People are funny that way. We often think that regulation is the enemy of innovation, but really, it’s us. Humans are incredibly resistant to change. We don’t like things that we don’t understand. And that’s our challenge right now — technology is accelerating faster than people's ability to accept change. That shiny, new tool you purchased for your recruiting team isn’t worth the paper the contract was printed on if your recruiters don’t actually use it. And they won’t use it if they don’t get it.
That’s been a huge driving force behind the way we design products at Paradox. Make it simple, fast, and conversational. Stick to what people know.
Most talent professionals are not experts on AI. Most candidates don’t know what the heck a large language model is. It doesn’t matter. Can you text? Great, you can use our product. Do you struggle with getting back to candidates fast enough? Great, you’ll gain value from our product. Honestly, I find that “innovation” and “obvious” often exist on a razor’s edge — the reason conversational AI is so effective within the hiring process is because it’s a new way to do the most familiar thing, which is talk to people.
We can’t just innovate for the sake of innovation. It has to make sense, and people have to get it. The best way to leverage all these new AI tools is to solve old problems with them, and take things that used to be slow and tedious and make them fast and easy.
That’s how we go from potential to productivity.
Solving the ‘lens problem.’
Now, not all productivity is created equal. Sometimes productivity for me can be extremely inconvenient for you. In the case of TA, a recruiter’s automation can be the candidate’s spam.
This is what I call the “lens problem.” Productivity depends entirely on which perspective you’re looking from.
For example, you could decide you want to interview 1,000 candidates for some open roles. So you send out an automated text to these 1,000 people and invite them to do an AI automated interview — something that takes time and effort from a candidate, even though it took just a few minutes for you. Wow, that was super easy. Meanwhile, you just inconvenienced 1,000 candidates.
Was the automation good for everyone or was it good for just you?
These are the questions we have to start thinking about if we truly want to maximize efficiency. What do candidates actually want? What are the areas ripe for improvement through automation, and what are the things that are still better handled by a person? I can tell you that if you automate anything right now, it should be interview scheduling — it’s the one area of the process where both sides clearly benefit from just making it simpler and faster. And the best part is that it remains true whether you’re hiring 100 package handlers or one senior engineer.
But it’s not always that cut and dry. There are other stages of the process — sending offers and onboarding paperwork, for instance — where you need to be more strategic about using automation. As you layer in more and more AI to your processes, just remember that there’s always someone on the other end of that automated message who is looking to maximize their productivity, too.
As tech changes, people need to change.
And really, as much the discourse is centered around AI technology, all roads lead back to people. Because, again, this isn't about AI — it’s about making people better versions of themselves.
AI is evolving rapidly and it’s an undeniable fact that it can now do certain tasks better than a person can. That’s just reality. We’re seeing some of our clients who do high-volume hiring accurately schedule thousands of interviews in minutes without any human intervention. That might be scary if you’re a recruiter who spends a lot of their time coordinating interviews — but it’s also unbelievably liberating. Imagine what you could do with several hundred extra hours every year? What new tasks could you take on? What new skills could you learn? Is your role even the same or does it become something else entirely?
We have reached this pivot point where the tasks that need to be done by a human have changed. And that means humans have to change, too. The job of talent leaders — and quite frankly, tech companies like Paradox — will be to make that change feel as familiar as possible.
Some of it will just be getting back to the basics (letting recruiters focus on recruiting again), but there will also be a manifestation of new roles, new teams, new structures, to help maximize talent within the organization. For instance, as AI empowers store managers to hire more autonomously, it means that organizations can shift recruiting coordinators who used to support with screening and scheduling into new roles focused more on internal mobility and employee training. This isn’t wishcasting, either — 7-Eleven already did it.
At Paradox, we’ve always said that our goal is to help people spend time with people, not software. It’s pretty amazing that in the last few years it’s gone from a mission statement to reality.
When people can spend more time with people, that’s when productivity happens. That’s when magic happens. And no, I’m not talking about Harry Potter.
