Learning to Play Together, AI Integration in Teams.

Earlier this year, I wrote about how a music jam session at La Peligrosa Restaurant in Madrid reminded me about exceptional teamwork. Watching musicians seamlessly blend individual talent with collective rhythm, embracing diversity and adapting organically to constant change felt like a masterclass in collaboration.

As I reflect on that lesson in collaboration, the current conversation about AI and work feels starkly different. The headlines are full of dire predictions: Ford's CEO claiming AI will replace half of all white-collar workers, Anthropic's leader warning of a "white-collar bloodbath" that could spike unemployment to 20% within five years. Microsoft, CrowdStrike and others are already citing AI as a factor in their latest rounds of layoffs.

Yet I keep returning to that evening (yes it was a fun night) and what it reminded me about collaboration. I believe I am not alone in thinking we might be asking the wrong questions entirely. Instead of "Will AI take our jobs?" what if we asked "How do we learn to collaborate together?"

In my university, days, we looked at Human Computer Interaction (HCI) and how this was evolving in the workplace and through organisation design (the impact of the early days of the Internet!). Through my career, using these optics has been key to how I frame technology as an enabler for businesses. Now, it is exciting again to be amongst this new opportunity unfolding at such a rapid pace.

When AI Joins the Band

AI agents and digital workers are no longer just sophisticated tools. They are becoming team members with their own capabilities, limitations and ways of processing information. They are showing up to our metaphorical jam sessions, but they do not yet know our language, our cultural shortcuts or our unspoken rhythms.

This transition creates fundamental challenges around four critical dimensions:

• creating learning mechanisms that evolve with both parties

• developing shared communication protocols

• establishing governance frameworks for accountability

• preserving essential human culture whilst enabling meaningful AI contribution.

Learning the Rhythms

One of the most intriguing aspects of this transition will be watching how human-AI teams develop heuristics, those practical rules of thumb that help teams work efficiently without having to repeat every interaction from scratch. In the early days of HCI, we learned that the most effective human-computer interactions emerged when both sides developed reliable patterns for communication and task allocation.

Today's AI integration follows remarkably similar principles. Some teams are already developing sophisticated heuristics for when to rely on AI analysis versus human judgment, how to structure prompts for consistent results and when AI recommendations should be questioned or overridden. As AI literacy grows and digital workers become more embedded in team structures, these emerging patterns will become the practical wisdom that allows hybrid teams to function smoothly.

Unlike traditional software interfaces with fixed interaction models, AI agents learn and adapt, meaning the heuristics themselves evolve. What worked last month may need refinement as the AI's capabilities expand or as teams discover more effective collaboration patterns. Teams will need to invest in evolving this relationship.

Speaking the Same Language

Beyond developing working heuristics, teams face the deeper challenge of creating shared language with AI agents. In that Madrid jam session, musicians communicated through subtle cues: a glance signalling a key change, a rhythm telling the bass to take the lead next, the nod to the drummer to wrap up the song, etc. Human teams develop similar languages over time: the shared vocabulary, the understanding of what tone indicates real concern versus routine discussion, when silence from certain team members signals disagreement and how the pace of communication reveals the true urgency of a situation.

The integration challenge here is substantial. AI agents need to learn not just what we say, but what we mean. When a team has spent years developing its own vocabulary of gestures, inside jokes and cultural references, the digital team member requires explicit instruction to join that conversation.

This forces organisations to make implicit knowledge explicit, a process in itself that often reveals how much of team effectiveness relies on unspoken understanding.

Redefining the Ensemble

What will it mean to the team, or members in that team, when when some participants never experience fatigue, emotional states or the need for encouragement. What constitutes effective collaboration when team members operate from completely different frameworks of understanding?

In that Madrid session, each musician brought not just technical skills but personality, mood and creative interpretation. They responded to the energy in the room, the audience's reaction, their own state of mind. AI agents contribute different strengths: consistency, processing power, availability, but through entirely different mechanisms.

Preserving the Soul

Perhaps most critically, how does team culture evolve when some members experience it completely differently? Culture often lives in the spaces between formal interactions. In the casual conversations, the shared frustrations, the moments of unexpected connection all built and developed through accumulated experiences and stories.

The question organisations will need to address are which aspects of team culture are essentially human and must be preserved, versus which elements can be shared or adapted when digital intelligence joins the conversation.

Beyond Replacement to Integration

Unlike the organic flow of that music session, most organisations are still figuring out how to integrate AI agents meaningfully. While we see companies deploying AI for discrete tasks such as content generation, data analysis and customer service responses, true integration as collaborative team members remains largely experimental. The pressure to demonstrate AI ROI is driving many towards automation of existing processes rather than reimagining how human-AI teams might work together.

The humans and digital workers just aren't jamming yet.

We are still in the early stages of this transition, largely experimenting without established frameworks. The few concrete examples that make the news cycles suggest the path forward is not straightforward.

What seems clear is that successful integration will require treating this as a cultural and collaborative challenge, not just a technical implementation. Questions remain largely unanswered, creating both risk and opportunity for organisations willing to approach this transition systematically.

Building the Operating System

Rather than hoping human-AI collaboration will emerge organically, organisations need to deliberately build the operating system that enables this workforce evolution. This requires establishing the four foundational components that support teams learning to play together effectively.

Communication Protocols and Interfaces

Just as musicians develop signals for key changes and transitions, human-AI teams need explicit protocols for information flow, decision-making authority and feedback mechanisms. This involves establishing clear handoff points between human and AI work, defining when human oversight is required and creating systematic approaches for both humans and AI to signal when collaboration isn't working effectively.

Teams will develop sophisticated communication frameworks that go beyond simple prompt engineering to include context-setting protocols, quality assurance checkpoints and escalation procedures when outputs don't meet expectations.

Learning and Adaptation Mechanisms

The operating system must enable both humans and AI to improve performance through interaction. This means capturing what collaboration patterns work well, identifying when and why human-AI handoffs fail and systematically refining approaches based on outcomes.

Unlike traditional training programmes with fixed curricula, these mechanisms will need to evolve continuously as AI capabilities advance and team dynamics mature. Teams will require structured ways to experiment with new collaboration approaches whilst maintaining productivity and quality standards.

Governance and Decision-Making Frameworks

Clear accountability structures will become critical when team members operate from different cognitive frameworks. This involves establishing who holds responsibility for AI-generated outputs, how quality assurance operates across human and digital work and what escalation paths exist when human-AI collaboration produces unexpected or problematic results.

Effective governance frameworks will also address bias detection and mitigation, ensuring that AI team members contribute to rather than undermine team effectiveness and organisational values.

Cultural Integration Systems

Perhaps the most complex component will involve preserving essential human culture whilst enabling digital workers to contribute meaningfully to team dynamics. This will require systematic approaches to onboarding AI agents into existing team cultures, processes that maintain human connection and creativity, and mechanisms that prevent digital efficiency from overwhelming human judgment and intuition.

The most thoughtful organisations will develop cultural frameworks that treat AI integration as an evolution of team identity rather than a replacement of human elements.

The Conductor's Questions

As leaders work to build these operating systems, several key actions deserve immediate attention:

• Audit your implicit knowledge. What team wisdom currently exists only in heads and habits? How will you make this explicit enough for AI agents to understand and contribute?

• Design your governance model. Who will be accountable for human-AI team outputs? What quality assurance processes will ensure effective collaboration rather than just efficient automation?

• Establish learning mechanisms. How will your teams systematically improve their human-AI collaboration over time? What feedback loops will enable both humans and AI to adapt and enhance performance?

• Preserve cultural foundations. What aspects of your team culture are essential to maintain? How will you ensure AI integration enhances rather than erodes the human elements that drive innovation and resilience?

  
  

The jam session continues, and we are all still learning to play together. The organisations that approach this transition by building deliberate operating systems for human-AI collaboration, rather than hoping it emerges naturally, will likely develop the most sustainable competitive advantage.

What operating system components are you building in your organisation? How are you balancing the systematic infrastructure needed for effective collaboration with the flexibility required for innovation and adaptation?