When doing more with AI isn't enough - Teamwork Lab

AI has made it easier to move faster, produce more, and check off tasks at a pace that would have felt impossible just a few years ago. 

But there’s a catch: more output doesn’t automatically translate into more visibility, trust, or impact. In fact in some cases, it quietly does the opposite.

In our recent webinar, Work Smarter, Not Harder: 3 AI Traps to Avoid, Atlassian’s Teamwork Lab unpacked this “AI productivity paradox” and heard from hundreds of attendees who feel stuck in it.

This article distills those insights into a practical guide for leaders and individual contributors alike. Read on to learn why “doing more with AI” isn’t enough, how to spot three hidden AI patterns that quietly drain visibility and impact, and how to turn AI from a simple productivity booster into a true engine for team and career growth.

Three hidden AI patterns that quietly undermine team visibility and impact — and how to fix them

In our research and hands‑on work with teams, the Teamwork Lab has spotted three AI patterns that often backfire.

Pattern 1: The Quality Sacrificer

When speed replaces intention.

This shows up when AI makes it incredibly easy to produce a lot of work, very quickly. How invigorating!

But when does speed start standing in for quality? A shipped document might have pristine grammar, expert structure, and a professional look. Upon a closer look, the document doesn’t answer any new questions. The key context is entirely missing or buried in niceties.

This is “AI slop”: work that looks polished but lacks deeper context and judgment.

In one recent survey, 37% of executives say AI has wasted their teams’ time or led them in the wrong direction due to low-quality or poorly framed outputs.

“With AI, the cost of doing work drops,” says Teamwork Lab researcher Ben Ostrowski. “At an individual level, that sounds great. But when everyone is suddenly turning in five to ten times more content, the reality is that everyone is drowning.”

Early warning signs you’re in this pattern:

  • You’re pasting AI’s answer with only light edits instead of adding your own judgment and context
  • People reply with questions like, “What are you really trying to say?” or “Can you TL;DR this?”
  • Your team is shipping more, but engagement is dropping — fewer comments, fewer questions, less collective energy around next steps

If you’re buried in AI-generated drafts and hearing crickets when it comes to feedback, your team might be in the Quality Sacrificer trap. This gap between effort and recognition isn’t just frustrating. It’s a real reputation risk. If your team’s work is invisible (or worse, mistrusted), it doesn’t matter how much you complete.

Reverse the Quality Sacrificer: Use AI to sharpen goals around impact, not just output

One of AI’s most powerful (and overlooked) uses is stress‑testing your goals. Instead of just shipping faster, use AI as a strategic partner to tie work to clear, outcome‑based goals that map to team and org priorities. Visibility and recognition follow naturally.

Start / Stop / Keep: Goals + AI

Start: Using AI at kickoff to define success, share the intent behind goals, and iterate until they’re impact‑focused.
Stop: Measuring by “things shipped” or setting secret, static goals in isolation.
Keep: Using goals to prioritize weekly work and asking AI, “Given these goals, what should I focus on this week to maximize impact?”

Result: You’re not just busy — you’re visibly moving organizational goals forward. (Try the OKR Play)

One of AI’s most powerful (and overlooked) uses is stress‑testing your goals. Instead of just shipping faster, use AI as a strategic partner to tie work to clear, outcome‑based goals that map to team and org priorities. Visibility and recognition follow naturally.

Pattern 2: The Rogue Optimizer

When individual speed outpaces team progress.

In this scenario, you — or any AI super user on your team — genuinely moves faster with AI. The problem? The rest of your team and systems haven’t caught up.

On an individual level, this can feel like success. In reality, individual speed doesn’t automatically equal team progress.

As Ostrowski explains, “This is a bottleneck problem. A team needs to collaborate to produce high-quality work that leverages its members’ unique expertise.”

Without shared norms and workflows:

  • Work gets stuck at handoffs (“I want to move this forward, but I’m still slogging through a huge review queue.”)
  • Projects are duplicated or re‑done from scratch
  • Knowledge stays trapped in one person’s AI habits

In our research, teams that focus only on individual productivity are significantly less likely to drive real innovation than teams who share their ways of working.

If your personal speed is leaving your team behind, it’s time to shift from individual optimization to collective impact.

How co-creating AI working agreements drives confidence and clarity

Reverse the Rogue Optimizer: Create AI working agreements so teams (not just individuals) can shine

AI working agreements are team‑created guidelines for how you’ll use AI together. They turn vague “be smart about AI” expectations into shared norms and best practices.

For a Rogue Optimizer, an AI working agreement is the perfect place to surface your favorite AI use cases and turn them into team habits instead of solo shortcuts.

After Teamwork Lab’s AI Working Agreement experiment, 82% of participants said the exercise made them more aligned on how their team should use AI to drive progress.

Start / Stop / Keep: AI working agreements

Start: Meet with your team to discuss how each person is using AI today. Capture concerns (security, quality, bias, stakeholder expectations), and create 4-6 simple norms.
Stop: Hiding or not sharing how you’re using AI.
Keep: Maintaining (and updating!) clear team working agreements and sharing good prompts and use cases in a common space (like a Confluence page or team Slack channel).

Result: When teams have explicit norms, AI impact snowballs. (Try to AI working agreement play)

Pattern 3: The AI Consumer

When AI becomes a shortcut, not a thought partner.

This pattern is more subtle. You’re using AI regularly, but mostly for fast, transactional to‑dos like drafting an email, summarizing a meeting, rewriting text in a different tone, or doing quick fact‑finding.

The risk here is that if you regularly outsource your problem-solving, critical skills can erode. You might be the person or team who can “get things done,” but your cognitive muscles atrophy and you stop being the leader others look to for direction.

Teamwork Lab researcher Alissa Yu explains, “Humans are cognitive misers. We don’t want to think hard if we don’t have to.”

If you’re only using AI to knock out busy work, you’re missing a critical application. The ideal is an active partnership with AI, using it to stretch your thinking, not to avoid thinking for yourself.

Reverse the AI consumer: Treat AI as a thought partner, not a shortcut

If you mostly use AI for convenience, you’re leaving value on the table. A great AI session should feel like a tough working session with a colleague, one with infinite patience and a fresh perspective.

Start / Stop / Keep: Thought partner mode

Start: Asking AI to poke holes in your ideas, generate counterarguments, help you see second‑order consequences, and simulate feedback from specific personas (e.g., “review this as a senior engineer who’s skeptical of AI”).
Stop: Treating AI like a one‑and‑done answer vending machine, prompting with little context , and accepting the first answer as the answer.
Keep: Keep yourself as the final boss and arbiter of quality. Refine your prompts as you move along and iterate with AI to explore alternate options. 

Result: You’re building a reputation for consistently sharp, well-tested, and grounded thinking.

AI isn’t going anywhere, and neither is the pressure to do more with less. But your team’s impact isn’t built on how many documents you ship or how many prompts you run. It’s built on the trust people have in your ability to meaningfully move work forward. So use AI for what actually earns that trust: clearer goals, stronger decisions, and work that unmistakably moves the needle.

teamwork lab

When doing more with AI isn’t enough: Three patterns quietly undermining your team’s impact