"AI without data is like a car without gas"

“The name of the game is, of course, AI. This is the complex reality we live in. On the one hand, a massive transformation, and on the other, an overwhelming flood of data that we need to learn how to navigate,” said Playtika CTO Uri Rubin, speaking about how smart code and advanced technologies are shaping the next generation of games at the Calcalist Gaming Conference, held in collaboration with Playtika and Google.

Rubin described Playtika’s journey over the past decade. “In 2018, we acquired an AI lab staffed by data scientists, with the goal of building breakthrough solutions focused on player experience and advanced capabilities. In 2023, we released quite a few impressive tools, but it was an extremely expensive and lengthy process that didn’t pay for itself, likely because the underlying technologies, both AI and hardware, just weren’t mature enough. Every year, the number of tools we work with grows. If in 2023 it was mainly ChatGPT, today we have dozens of tools inside the organization, including many we piloted, invested time in, and ultimately realized weren’t the right fit.”

Rubin noted that the process is long and filled with setbacks. “Locking yourself into a single tool is unhealthy, it’s a recipe for failure. The most important factor is mindset. Managers and employees must recognize that the challenge is learning how to work with tools that are constantly changing. The conclusion is simple: don’t fall in love with the tool; teach people how to work with tools.”

Rubin shared several key lessons from the journey. “AI without data is like a car without gas. Some failures happened because the data wasn’t ready. You need at least a year of historical data on everything players do, and you need it for millions of players, otherwise AI can’t produce insights. Most organizations don’t collect data consistently. The second lesson is mindset. Just as a new manager must let go and trust their team, employees must learn to let go and let AI handle tasks. They have to decide where to let go, and we have to teach them that constantly. Another lesson is the learning curve. It’s expensive and takes time. People are expensive, infrastructure is expensive, computing power is expensive. But anyone who didn’t start this process yesterday won’t be ready for tomorrow.”

Playtika’s transformation approach is divided across business and organizational needs. “As a gaming company, when we talk about AI, the real story is where it meets the business and enhances the game. When talking to employees about the business potential, I need to make sure they understand I’m not replacing them. When we talk about organizational transformation, we’re transparent: some roles will change, software developers, product managers, artists. Wherever we see room for efficiency, we put it on the table and build a plan. We’ve launched a transformation project for our studio to make it fully AI-operated. This doesn’t mean fully replaced, but its operations will be fully AI-based. The key is personalization. We look at the player end to end.”

“At a technological level, AI isn’t new. What’s new is that today it’s feasible, cost-wise and speed-wise. Personalization is the ability to reach a player and deliver a tailored experience in real time. Even ten years ago, I could personalize an experience, but it required a week of data processing. Today, in real time, I can decide whether a player is turning left or right.”

To conclude, Rubin offered practical insights. “The entire organization must speak the same language of strategic alignment. Without that, change won’t happen. Second, accept that AI replaces tasks and people. None of these roles, developers or artists, will disappear in the foreseeable future, but parts of their work will. We brought in new talent, trained them in AI, and returned them to their teams as change agents. That builds trust. Talk to employees openly about efficiency and opportunity.”