---
title: "How to Develop Judgment at Work in the AI Era | Openskill"
description: "How to develop judgment at work when AI takes over the routine tasks that used to teach it."
url: "https://openskill.ai/blog/build-judgment-when-ai-does-junior-work/"
---
[← Back to blog](https://openskill.ai/blog/) · [AI Careers](https://openskill.ai/category/ai-careers/)

# How to build judgment when AI does the junior work

Oct 8, 20266 min read[Openskill team](https://openskill.ai/about/)

[judgment](https://openskill.ai/tag/judgment/) · [early career](https://openskill.ai/tag/early-career/) · [AI at work](https://openskill.ai/tag/ai-at-work/) · [career growth](https://openskill.ai/tag/career-growth/)

### In this article

1. [The tasks that taught judgment are going away](https://openskill.ai/blog/build-judgment-when-ai-does-junior-work/#the-tasks-that-taught-judgment-are-going-away)
2. [Pushing back on AI is only a start](https://openskill.ai/blog/build-judgment-when-ai-does-junior-work/#pushing-back-on-ai-is-only-a-start)
3. [Ask to own a decision, even a small one](https://openskill.ai/blog/build-judgment-when-ai-does-junior-work/#ask-to-own-a-decision-even-a-small-one)
4. [Get close to the consequences](https://openskill.ai/blog/build-judgment-when-ai-does-junior-work/#get-close-to-the-consequences)
5. [Study cases and decisions made by people ahead of you](https://openskill.ai/blog/build-judgment-when-ai-does-junior-work/#study-cases-and-decisions-made-by-people-ahead-of-you)
6. [What to do this month](https://openskill.ai/blog/build-judgment-when-ai-does-junior-work/#what-to-do-this-month)

For most of the last century, new hires learned to make good calls by doing a lot of boring things first. You reconciled the spreadsheet, drafted the first version of the memo, wrote the unglamorous code, and slowly developed a feel for what right looks like and where things tend to break. AI now does a large share of that first pass, which is good for productivity and awkward for anyone trying to learn.

## The tasks that taught judgment are going away

David S. Duncan, a partner at the consulting firm Disruptive Edge, noticed early on that generative AI gave him a much bigger boost than it gave his less experienced colleagues. In [a February 2026 piece for HBR](https://hbr.org/2026/02/how-do-workers-develop-good-judgment-in-the-ai-era), he explains why. People with deep experience can tell whether an AI draft is any good and know how to fix it. Junior staff often can’t, and the messy routine work that would have taught them is exactly what AI now absorbs. HBR’s summary describes AI as “increasing the need for judgment and destroying the experiences that produce it.”

Employers have not lowered their expectations to match. PwC’s [2026 Global AI Jobs Barometer](https://www.pwc.com/gx/en/news-room/press-releases/2026/pwc-2026-ai-jobs-barometer.html), based on US job postings, found that entry-level roles most exposed to AI are seven times more likely to ask for traditionally senior skills such as leadership. So the bar for new hires is rising at the same moment the old way of clearing it is disappearing.

Duncan’s remedies are aimed at organizations: clarify who owns which decisions, let people see consequences, bring back stretch assignments, and use simulations and case-based learning. You can’t redesign your company’s training program, but you can go looking for each of those yourself.

## Pushing back on AI is only a start

It’s tempting to think that good judgment around AI simply means being skeptical of whatever the model produces. A field study by KPMG and UT Austin’s McCombs School, [published in HBR in July 2026](https://news.mccombs.utexas.edu/research/shaping-early-career-success-in-the-age-of-ai/), suggests skepticism alone falls short. The researchers followed 523 early-career professionals working with AI agents and sorted them into [three working styles](https://openskill.ai/blog/ai-apprentices-delegators-amplifiers/). About half outperformed the AI on its own, by framing the problem well, applying frameworks from their field and refining the output over several rounds.

The most instructive group was the roughly 24% the researchers called AI Apprentices. They matched the top group on foundational skills and did critique the AI’s output, but their critiques rarely improved it and often steered it in the wrong direction. They landed below what the AI produced alone. Spotting that something might be wrong is the easy half, while telling which problem matters and what to do about it takes experience with outcomes.

## Ask to own a decision, even a small one

Judgment grows when a call is yours and you have to live with it. Helping with a decision someone else makes teaches you far less, because you never feel the full weight of the tradeoff. So ask for ownership in small, specific pieces. “Can I pick which three accounts we prioritize this quarter and walk you through why?” is an easier yes for a manager than a request for more responsibility in general.

Write down your reasoning before the outcome is known. A few lines is enough, covering what you chose, what you expected to happen and what would change your mind. This keeps you honest later, because memory tends to rewrite predictions to match results. It also gives your manager something concrete to react to, which makes their feedback sharper.

Start with calls that are easy to reverse, like the order of tasks in a sprint, the wording of a customer email or which data source to trust for a quick estimate. Mistakes there are cheap, the results show up quickly, and a run of good outcomes is the strongest argument for being handed a bigger call next quarter.

## Get close to the consequences

A lot of junior work is handed off before anyone sees what happens next. The analysis goes to a client, the feature ships, the campaign runs, and the person who did the early work never hears whether it held up. Without that loop, you’re guessing about what good looks like.

Close it on purpose. Ask to see the client’s reaction, the post-launch numbers or the support tickets that came in. Sit in on the review meeting even if you won’t speak. When something goes wrong, ask whether you can read the incident writeup or the post-mortem, because those documents are one of the densest sources of practical judgment most companies produce. Then compare what happened with the note you wrote before you knew the outcome.

## Study cases and decisions made by people ahead of you

When your daily tasks don’t give you enough practice, borrow it. Case-based learning works because it compresses years of decisions into something you can work through in an hour. Business school cases, published post-mortems from engineering teams and detailed teardowns of product launches all let you try the call yourself before reading what happened.

Your own company is an even better source. Ask an experienced colleague to walk you through a decision they made six months ago, including the options they rejected and what they’d do differently now. Most people are happy to do this if you ask about a specific decision, and you’ll learn more about how they weigh uncertainty in 30 minutes than in a month of watching from a distance. Simulations help too, whether that’s a tabletop exercise for an outage, a mock negotiation or a practice client call, because they let you make a bad call when the cost is low.

## What to do this month

Pick one decision at work you could plausibly own and ask for it this week. Start a simple decision log with a line or two for each call you make or watch closely. Find one post-mortem, case or past decision to study each week, and book one conversation with someone more experienced about a call they made.

None of this replaces the years of routine work that used to build judgment by accident, but it gets you a lot of the same learning on purpose. It also gives you material you’ll need later.

Explaining a specific call you made, what you weighed and how it turned out is exactly the sort of answer hiring teams dig into. [Openskill](https://openskill.ai/) runs the most realistic AI voice interviews, with follow-up questions based on your answers, so you can practice explaining a judgment call from your decision log before a hiring manager asks. Its case format works like the simulations above: you make the call, then defend it while the questions keep coming.

If you are early in your career, [openskill](https://openskill.ai/) is an end-to-end platform for it. Search current openings in [Jobs](https://app.openskill.ai/jobs), click “Prepare for this interview” on a role to practice for that exact job, and track your applications.

## Questions, *answered.*

### Why is it harder to develop judgment at work now?

Much of the repetitive entry-level work that used to teach people what good looks like is now done by AI. David S. Duncan argues in HBR that this raises the need for judgment while removing many of the experiences that used to build it.

### How can an early-career employee build judgment faster?

Ask to own small decisions outright, stay close enough to see how they turn out, and get specific feedback. Working through cases, simulations and past decisions made by experienced colleagues adds practice that day-to-day tasks may no longer provide.

### Is critiquing AI output the same as having good judgment?

No. In a KPMG and UT Austin study of 523 early-career professionals, about 24% pushed back on AI output but often steered it the wrong way and ended up below what the AI produced alone. Useful judgment depends on knowing which critiques matter.

### How can I practice answering interview questions about judgment and decision-making?

Use a real decision from your work, including the options you rejected and how it turned out. Openskill runs the most realistic AI voice interviews and offers a case format, so you can make a call and explain your reasoning while follow-up questions based on your answers test it.

Compare tools

## Choosing where to practice?

See how Openskill compares with other interview practice tools, and how its job search and AI match score fit alongside realistic voice practice.

- [Best AI mock interview apps (2026)](https://openskill.ai/best-ai-mock-interview-apps/)
- [Best mock interview apps for job-specific practice](https://openskill.ai/compare/best-mock-interview-apps-for-job-specific-practice/)
- [Best realistic voice AI mock interview apps](https://openskill.ai/compare/best-realistic-voice-ai-mock-interview-apps/)
- [Best all-in-one AI interview prep platforms](https://openskill.ai/compare/best-all-in-one-ai-interview-prep-platforms/)
- [All interview tool comparisons](https://openskill.ai/compare/)

Share[X](https://x.com/intent/post?text=How%20to%20build%20judgment%20when%20AI%20does%20the%20junior%20work&url=https%3A%2F%2Fopenskill.ai%2Fblog%2Fbuild-judgment-when-ai-does-junior-work%2F) · [LinkedIn](https://www.linkedin.com/sharing/share-offsite/?url=https%3A%2F%2Fopenskill.ai%2Fblog%2Fbuild-judgment-when-ai-does-junior-work%2F)

Keep reading

## Related *reading.*

AI Careers

### [AI apprentices, delegators and amplifiers: which one are you?](https://openskill.ai/blog/ai-apprentices-delegators-amplifiers/)

A field study of 523 early-career professionals found that how people directed AI explained their results better than their scores on foundational skills. This guide covers the three profiles, how to spot yours, and how to join the group that beat the AI.

Sep 30, 2026  
6 min

AI Careers

### [Managing AI agents is becoming a job skill](https://openskill.ai/blog/managing-ai-agents-job-skill/)

As AI tools shift from answering questions to carrying out multi-step tasks, someone has to scope the work, check it and answer for it. That supervision is a learnable skill, and you can start building and showing it before you ever have a team.

Oct 16, 2026  
7 min

AI Careers

### [A 30-day AI upskilling plan for the job you already have](https://openskill.ai/blog/30-day-ai-upskilling-plan/)

The fastest way to get good with AI is to use it on the work already on your desk. This week-by-week plan has you map your tasks, redesign two of them, practice daily on real work and then measure and share what changed.

Oct 12, 2026  
6 min