The New Team Member Who Isn't Human: Navigating the AI Collaboration Conundrum

Our newest team member isn’t a person. It’s an AI. And this presents a fascinating conundrum for modern leadership and teamwork.

We know AI isn’t sentient. It doesn’t have feelings, a family to get home to, or a career to build. Yet, to unlock its true potential, we must treat it less like a vending machine and more like a junior team member. The most effective users of AI are those who ‘work with’ it, not just use it.

The paradox is this: the more we treat our AI tools like a colleague, the better they perform.

Coaching the Code: Why Your AI Needs a Mentor

Think about how you onboard a new team member. You don’t just give them a single, blunt instruction and expect a perfect result. You provide context, offer examples, give feedback, and clarify your expectations.

It’s the same with AI.

As platforms like ChatGPT, Gemini, and Copilot evolve with larger context windows and persistent memory, our ability to “train” them on our specific needs grows exponentially.

Provide Context: “Analyse this sales data” is a weak prompt. “You are a senior data analyst for a UK-based publishing house. Please analyse the attached Q2 sales data for our e-commerce fiction division, identify the top three trends, and present them in a table for a senior leadership team who are short on time” will yield vastly superior results.
Give Constructive Feedback: When the initial output isn’t right, don’t just discard it. Coach the AI. “That’s a good start, but we’ve focused too much on raw numbers. Could you reinterpret the data to highlight the implications for our marketing strategy in Q3?”
Iterate and Refine: The best results come from a conversational process, much like a creative brainstorming session with a human colleague.

Research from institutions like MIT’s Media Lab has consistently shown that a collaborative, iterative approach to human-AI interaction leads to more innovative and effective outcomes. By investing the time to mentor our AI, we build a powerful assistant that understands our unique context and standards, ultimately saving us significant time and improving the quality of our work. It becomes a true extension of our own capability. Of course, this kind of collaboration doesn’t happen in a vacuum — organisations need to give employees the time and space to become AI fluent for this partnership to truly work.

The Empathy Erosion Risk: A Warning for Managers

There is a danger with working closely with AI. AI is endlessly patient (which in itself can also be quite frustrating). It requires no pleasantries. It doesn’t get discouraged by blunt feedback. It’s available 24/7 and never has an off day (that’s all yours).

Interacting with this kind of hyper-efficient, non-emotional “colleague” can subtly recalibrate our expectations for human interaction. We risk becoming less patient with the very human messiness of… well, humans.

Patience Deficit: After receiving an instant, detailed report from an AI, waiting a day for a human colleague to do the same can feel frustrating. We might forget they are juggling multiple projects, attending meetings, and dealing with their own human limitations.
Communication Shortcuts: The need for politeness, small talk, and emotional intelligence (the social glue of any successful team) is absent in AI interaction. There’s a real risk this “efficiency” bleeds into our management style, leading us to become more transactional and less connected with our teams.
Devaluing Human Thought: We might begin to undervalue the slower, more deliberate, and sometimes meandering process of human thinking, which is often where true creativity and breakthrough ideas originate.

The risk is creating a workplace culture where we are brilliant at managing technology but poor at leading people. We become impatient, less empathetic, and more focused on the “what” than the “who” and “how.”

Mind the Merge: Some practical guardrails

Here are some ideas on how one can re-humanise after an intense prompting session:

Debrief aloud: Give the team a quick spoken summary of what the AI just produced; it recentres the conversation on human nuance.
Rotate human reviewers: Regularly swap who screen-shares their prompting; it keeps empathy muscles warm and helps newer users learn from veterans.
Add courtesy by default: Include “please draft…” and “thanks, now refine…” in your prompts; the habit trains both the model and your own tone, saving rewrites later. Have a look at my example prompts above and see how I softened the prompts with “please” and “could you…” instead of “now…”
Shadow-coach the AI: Explain your edits as you go, out loud or in comments; it makes tacit prompting skills visible and is ideal for onboarding graduates and apprentices.

Leading in the AI Era: A Call for Conscious Collaboration

The solution requires a conscious effort toboth switch between modes of communication, and to maintain our humanity across all interactions. Rather than treating AI with blunt efficiency, we should consider our interactions with it as a training ground for our interactions with people.

If we practice being demanding and polite, clear and considerate with our AI assistant, we strengthen those positive communication pathways. Treating the AI as we would a human colleague isn’t for its benefit, it’s for ours! It keeps our empathy and patience muscles from atrophying.

This approach reframes the dynamic:

The AI as a Practice Partner: Use interactions with AI to hone your skills in providing clear, respectful, and constructive feedback. This makes you a more effective AI user and a better manager of people. If you wouldn’t say it to a human colleague, don’t say it to your AI.
Humans as the Core Focus: By maintaining a baseline of polite and patient interaction with our technology, we ensure this remains our default mode. We can then more easily elevate that to the higher-level empathetic leadership required to truly nurture and develop our human teams.

Our challenge as leaders is to embrace AI as a powerful force multiplier without letting it diminish our own humanity. Let’s coach our AI with rigour and clarity, and let’s do it with a level of respect that we want to see reflected in our interactions everywhere. By doing so, we not only produce better work, we also protect and enhance our most valuable asset: our ability to lead and connect with other people.

What are your strategies for navigating this new dynamic?