AI x Biz Collaboration

working with AI

“I was in an AI meeting, and honestly, I struggled to follow what the AI team was saying”

This is one of the most common remarks from game publishing professionals after introducing AI into their workflow.

When terms like prompt, model, accuracy, and parameters… start dominating the conversation, it can feel like you need a translator just to follow along.

However, failing to bridge this gap can quickly derail an AI project. The distance between technology and business planning is often greater than expected, and without mutual understanding, collaboration rarely works as intended.

This final chapter in the series answers a practical question:

‘If you don’t build AI yourself, how can you still work effectively with the AI team?’

ℹ️ Connecting Technology and Business Planning: The Innovation Workshop

When discussing successful digital innovation, Amazon’s “Innovation Workshop” is a textbook example.

Before launching new initiatives, Amazon brings the technical and business teams together in a structured workshop – called “Innovative Workshop” - to turn ideas into actionable plans.

The process typically follows three steps: '

① Define the problem you want to solve

② Visualize the ideal user experience without technical constraints

③ Brainstorm the technologies and data needed to make it happen

This framework is highly applicable to AI in gaming. For example:

“Can we analyze PvP win rate data to detect player dissatisfaction early?”

“Can we review NPC dialogue logs to improve quest quality?”

By first defining the problem as mentioned above, envisioning the user experience, and then mapping the necessary technology and data, collaboration between the AI team and the business team can proceed effectively from the very start.

ℹ️ Great Collaboration Starts with a Goal, Not Technology

When working with AI teams, the first question that often arises is, “What can the AI team build?” But the better question to start with is, “What do we truly need right now?”

The AI team brings expertise in technology, while the game business team brings expertise in players and markets. Collaboration begins when both sides acknowledge and respect this domain knowledge.

For example:

The AI team designs how to solve the problem and which algorithms to use.

The business team defines the specific business outcomes to achieve.

A practical framework for this is Amazon’s “Working Backwards” approach. It starts by defining the end-user experience in detail and then planning in reverse. One way to do this is by writing a mock press release that describes the ideal scenario. Instead of saying, “Let’s build a churn prediction model,” frame it as, “We will create a system that automatically offers rewards to at-risk players to reduce churn.”

By clarifying ‘why something is being built’ rather than just ‘what to build,’ both the AI and business teams can establish trust and move toward shared goals.

ℹ️ From One-Way Request to a Feedback Loop

The reasons are clear. Most failures come from treating AI like any other IT task, following a one-way process of “request → development → delivery.” AI development rarely succeeds in such a linear model—it is inherently iterative.

But, AI doesn’t work that way.

It is built on continuous experimentation. Rarely does the first attempt succeed. Midway, directions often need to shift. Questions such as “How accurate were the model’s predictions of actual player behavior?” or “Did the anomaly logs AI detected connect to real in-game issues?” must be revisited repeatedly throughout the project.

This makes collaboration a circular process, not a linear one. Regular meetings, interim reviews, and feedback loops allow AI teams to refine and improve models in alignment with business needs.

And here lies the critical element: a shared language. When terms like accuracy, error rate, or quantitative metrics are mapped directly to business KPIs, business teams can grasp their meaning more easily. At the same time, AI teams gain clarity on “how much the business side understands,” which strengthens collaboration.

With that in mind, let’s revisit some essential AI concepts that game professionals and AI experts inevitably face together.

ℹ️ AI Is, Ultimately, a Co-Creation Experience

Effective collaboration is shaped as much by culture as by structure. Workshops, ideations, and open Q&A sessions where planners, developers, and AI engineers all participate serve as the foundation of AI collaboration.

It must be natural to ask: “Which data was this model trained on?” or “Which player behavior does this metric represent?” Only when such questions flow freely does AI truly integrate into everyday work.

AI does not deliver perfect results at the start.

The essence of collaboration lies in agility: experiments small, fail fast, and expand gradually. Start with a simple churn prediction, then move toward personalized promotions. The clearer the role division, the stronger the collaboration. The AI team interprets data; the business team reviews models. That is why this structure works best.

ℹ️ AI Is Both a Tool and a Culture

AI-driven collaboration extends beyond technology into a new way of working.

Real-time dashboards, prompt-based task automation, and meeting tools that summarize discussions and extract action items are not just convenient features. They transform how teams collaborate and communicate.

In game operations, QA, and marketing — where repetitive tasks dominate — AI is no longer just a tool. It is the work style itself.

ℹ️ The Core of Collaboration Is Trust

Technology and tools matter, but collaboration ultimately depends on people.

A culture that allows for experimentation and failure

A mindset that acknowledges contributions and shares success

An environment where questions can be asked freely

When these elements accumulate, AI stops being “just a technology” and becomes the way the team works together.

It may sound idealistic, but here is a practical way to ground it: an AI Collaboration Kickoff Worksheet.

Structured around “working backwards” and “innovation workshops,” this worksheet can be adapted

It ensures both AI teams and business teams start with the goal in mind, not just the technology.

Because at the end of the day, the most critical factors in live service games are not technical achievements but player experience and business impact.

ℹ️ In Closing: We Are All One Team Working With AI

AI teams and business teams speak different languages, but they share the same goals.

Even if you don’t build AI yourself, if you can define “why,” “by what standard,” and “what kind of experience we want to create,” you are already part of a team working with AI.

I hope that this post in the series serves as a realistic standard and starting point for every game professional considering AI adoption. Because the future of game publishing does not lie in choosing between human or AI — but in learning how to succeed together.