“Hiring assessments should be ‘wilderness camps’; places where people are evaluated to perform without AI assistance. Schools, to a large extent, should focus on core skills in the absence of AI. Because you still need those core skills to direct the agents.”
- Sanjoe Tom Jose (CEO at Talview)
As CEO of Talview and father to two teenagers, here’s why I believe this so deeply.
Picture this: instead of a take-home assignment where a candidate feeds prompts into Claude or GPT and gets a polished deliverable in 20 minutes, we send them into the ‘wilderness’. No AI. No internet copilot. Just the problem, a blank page or whiteboard (or laptop with no Copilot or Cowork), and their own mind.
Why? Because in the real world of 2026+, everyone has AI. The differentiator is who can direct it with judgment, taste, and depth.
I’ve watched brilliant young employees who are fluent in prompting but freeze when asked to explain why the AI’s output is suboptimal. They can generate code. They struggle to debug the deeper logic. They can summarize reports. They can’t spot the strategic blind spot the summary missed.
Core skills aren’t obsolete. They’re the operating system that lets you run the AI apps at full power.
My two teens are growing up in the most powerful tool era in human history. I’m thrilled about it and actively encourage it. But I also insist on large stretches of AI-free learning at home and my kids hate it :). They have to do Kumon math the old way. They write first drafts of English essays by hand. They need to argue ideas out loud without looking anything up.
It’s harder and way slower. But it’s building the resilience and mental models they’ll need to lead AI, not just use it.
I’m not saying ban AI in classrooms. That ship has sailed. I’m saying protect the developmental window where kids build raw capability. A significant portion of education should be unaided territory. The intellectual equivalent of learning to change a tire before you get self-driving cars.
Because the kids who only ever knew assisted driving will never be great mechanics when the system glitches.
At Talview we built Agentic AI proctoring and interviewing for exactly this reason to make large-scale, trustworthy “wilderness assessments” possible. Our AI agents (Alvy for proctoring, Ivy for interviewing) act like vigilant colleagues: they monitor, reason, and act in real time to protect integrity without being intrusive.
They let organizations design assessments that actually test unaided human excellence at scale. No more wondering if the code was 100% Claude. Just truth.
When we protect core human skills, AI becomes liberating instead of leveling. The people who have struggled, failed, and rebuilt their own thinking become the ones who can push AI furthest, spotting its limits, bending it creatively, using it as a true multiplier.
They become the directors. The orchestrators. The ones who matter most in the agentic future.
I want my kids to grow up knowing they can solve hard problems with nothing but their minds. I want the companies we work with to hire people who bring that same grounded capability to the table. And I want all of us to build systems in education and in hiring that honor and develop the irreplaceable human core. AI will take care of the rest.
What do you think?
Should more companies create “wilderness” moments in their hiring process?
Are schools striking the right balance?