I have been interviewing data and analytics candidates recently. Same pattern every time.
I ask what feature they would add to improve the customer experience. The answers I keep getting: RAG pipelines. Evaluation frameworks. Model fine-tuning. One candidate went straight into retrieval architecture before I had finished the sentence.
The answer might have been a better help link.
Not because the technical ideas are wrong. Because they did not ask what the customer was actually struggling with before deciding how to fix it. There is a version of this problem where users cannot find documentation and a well-placed link solves it entirely. There is another version where the product is genuinely broken and needs something more. You cannot know which one you are dealing with until you understand the problem. Nobody was asking.
This Is Not a New Problem. AI Just Made It More Visible.
When everyone has access to the same tools, execution becomes a commodity. Anyone can spec a feature. Anyone can describe a tech solution. Anyone can generate a product roadmap in the time it used to take to run a meeting. The hard part is no longer the doing. It is knowing what is worth doing in the first place.
That distinction is reshaping what hiring managers are actually looking for. A December 2025 survey of over 1,000 U.S. hiring managers found that 60% say soft skills matter more now than five years ago. Cangrade analyzed 200 real AI job postings and found 83% consistently required the same core skills across roles, industries, and seniority levels: critical thinking, strategic thinking, and collaboration. Not tools. Not certifications. Thinking.
LinkedIn's Skills on the Rise report for 2026 shows cross-functional collaboration climbing faster than almost any technical skill. The market is not asking for different credentials. It is asking for different humans.
The Gartner Stat That Should Stop You
In October 2025, Gartner's chief of research made a prediction that deserves more attention than it got. Half of global organizations, he said, will require "AI-free" skills assessments by the end of 2026. Not because they are anti-AI. Because daily AI use is atrophying critical thinking skills fast enough that the humans who can still think independently are becoming rare.
The skill most valued in the job market right now is the one that AI use is actively eroding.
A June 2025 study published in PNAS added another layer. Researchers ran twelve studies across 13,000 participants and found that when candidates know they are being screened by an AI tool, they instinctively suppress their emotional and intuitive qualities and push forward analytical ones. They hide the things that make them good collaborators and perform the things that make them look like machines.
The tools companies use to screen for human skills are training candidates to hide them. Nobody is winning that game.
What Problem Definition Actually Looks Like in a Room
The candidates who stood out in my interviews were not the ones with the best feature ideas. They were the ones who slowed down before answering.
They asked who the customer was and what they were trying to get done. They asked what friction currently exists and where people drop off. They questioned whether a new feature was even the right solution, or whether the existing product was just confusing. One candidate pushed back on the question entirely and said the better investment might be fixing something already broken rather than adding something new. That conversation was more useful than any feature spec they could have written on the spot.
That is the skill. Not solutioning. Problem definition. Understanding what the customer actually needs before deciding how to serve it. The best product decisions come from people who obsess over the problem, not the solution. The tech stack is secondary. It always has been.
If you are preparing for interviews right now, stop rehearsing your answers. Start practicing your questions. When someone asks what you would build, your first instinct should be to ask why. That pause, between the prompt and the solution, is exactly where the best candidates live.
Three Things You Can Do This Week
Before your next interview, practice answering product and problem questions without naming a technology. Force yourself to describe the customer's experience, the friction point, and the desired outcome before you ever say "we could build." If you cannot do it for two full minutes, you are not ready to talk about the solution.
When you are in the room, treat the interviewer's question as a hypothesis, not a directive. Ask what data they have on the problem. Ask who has complained about it and what they said. Ask what a good outcome looks like six months from now. These questions do not make you look unprepared. They make you look like someone who has done the job.
After the interview, reflect on the ratio. How much time did you spend defining the problem versus describing solutions? If the answer is mostly solutions, that is the thing to work on. The candidates who are separating themselves right now are the ones who can hold the problem longer than everyone else in the room.
Thank you for reading. If you like the newsletter feel free to forward along and help drive new eyes to Syllabi!
