DC70: Part 3: AI Collaboration
What Happens When You Add AI to the Mix
“We are entering a world where we will learn to coexist with AI, not as its masters, but as its collaborators.” – Mark Zuckerberg, CEO of Facebook.
When a team starts working with AI, something shifts.
Not instantly. Not magically. But meaningfully, if you approach it with the right mindset and structure.
Because adding AI into collaboration isn’t like adding another tool. It’s more like adding another participant, one that’s fast, unpredictable, sometimes brilliant, occasionally wrong, but always available.
When teams learn to work with AI together, they unlock a new rhythm of thinking. It’s faster. It’s more generative. And it often surfaces insights that wouldn’t emerge from human conversation alone.
Speed, Scale, Synthesis
The benefits show up early:
Speed: AI can write, rewrite, reframe and suggest in seconds. Teams no longer get stuck rewriting the same paragraph or summarising stickies.
Scale: It can scan vast documents, summarise long meetings, or translate multiple perspectives simultaneously.
Synthesis: Perhaps most powerfully, AI can bridge gaps in understanding. It can find the common thread between diverse inputs or expose where no such thread exists.
This doesn’t replace critical thinking, it accelerates it. It gives teams more to react to, more quickly.
It also brings some unique superpowers:
Memory: AI can retain detailed context across a session and even between sessions, depending on the tool.
Customisation: It can be tuned with your team’s specific language, documents, and frameworks (we’ll explore this in the “Build Your Own Bot” chapter later in the book).
But There Are Risks
AI isn’t neutral. It can hallucinate. It can reinforce bias. It can offer confident nonsense in a tone that sounds authoritative.
Used casually, this can derail collaboration. Teams might over-trust AI output, or default to whatever sounds polished. The result? Shallow consensus, groupthink, or a false sense of progress.
That’s why using AI collaboratively must be built on a foundation of critical engagement. Teams need to treat the AI like a junior team member with flashes of genius, helpful, but not unquestioned.
This requires a culture of:
Pausing to reflect on responses
Asking “what’s missing here?” or “does this actually help us?”
Encouraging critique without ego
Shared Prompting: A New Team Skill
Most people still prompt in isolation. They write a question, get a reply, and tweak it silently.
In team settings, prompting becomes a shared activity. The group discusses the intent. They shape the wording together. They test the AI’s response out loud. They iterate, not alone, but as a unit.
This builds alignment faster. It makes the reasoning visible. And it strengthens team learning.
Shared Prompting Tips
How to Phrase, Evaluate, and Revise Prompts as a Team
Start with Intent: Before typing anything, ask aloud: What are we really trying to find out? Clarity of purpose sharpens the prompt.
Use Roleplay: Try beginning prompts with “Act as…” to shape tone and perspective. For example, “Act as a strategy coach…” or “Act as a sceptical stakeholder…”
Be Iterative: Don’t aim for the perfect prompt on the first try. Try, evaluate, revise. Involve the group in deciding what to tweak.
Show Examples: If you want a certain tone or structure, provide a short example. AI mimics patterns better than vague instructions.
Question the Output: Ask the group, “Does this response help us?” or “What feels missing?” Keep AI in service of team learning, not as the final word.
Rotate the Prompt Driver: Let different people take turns entering prompts. It surfaces different mental models and expands group capability.
Shared prompting also helps surface the group’s assumptions. When people disagree on how to ask the question, they’re often revealing deeper differences in how they see the problem.
This is gold. It turns prompting into a diagnostic tool.
AI Doesn’t Replace People: It Elevates the Process
AI doesn’t replace curiosity, intuition, or relational wisdom. It doesn’t replace strategy or judgment.
But it does give teams a new kind of momentum. It turns hesitation into exploration. It turns silence into output. It makes it easier to keep going when the conversation gets stuck.
Used well, AI becomes an amplifier for shared intelligence.
Can Teams Collaborate with ChatGPT?
A common question is whether ChatGPT (or similar tools) can be used collaboratively by teams. Right now, ChatGPT itself doesn't offer native support for multi-user project spaces with shared context or memory across users.
That said, there are several practical workarounds:
Shared access: Teams often use a shared login or screen during live sessions.
Custom GPTs: These can be tailored to your content and used across a team, though memory and version control are still limited.
AI-augmented platforms: Tools like Notion AI, Miro + AI, and Taskade integrate AI into collaborative spaces, allowing real-time co-editing.
For more robust collaboration:
Build your own RAG chatbot: This approach supports shared memory, role-based access, and team-based training on your actual documents and language. (Covered in detail in Part 7.)
We’re still early in the journey toward fully integrated team-AI collaboration environments. But even today, teams that learn to share prompting, reflect together, and embed AI into their workflows will hold a meaningful edge.
In the next chapter, we’ll get into the practical techniques: how to structure group sessions, rotate prompting roles, and design formats that make AI a visible part of collaborative work.
Because the tool is here. The question is how we’ll use it.
I am currently looking for some contract work where I can help your organisation be even better.
Please get in contact if you would like to discuss whether I would be suitable or not.
Subscribing
Please create a paid subscription if you want to comment on this article or support the writing of these articles.
A subscription costs the same as buying one book per year.
Sharing and liking
I would appreciate it if you shared this article with your network if you think they would be interested. Please like this article if you find it valuable.
Thank you.