Moving Past the Fear: How Publishers Can Actually Grow with AI

The publishing world is currently obsessed with AI, but mostly in a surface-level way. While many professionals are playing around with chatbots or basic automation tools, there is a massive gap between knowing how to use a tool and actually understanding the technology behind it. This lack of deep knowledge is a major blind spot that prevents publishers from truly leveraging AI to redefine their business models. Javier Celaya, who recently completed an intensive master’s program in AI, argues that we need to move beyond just 'prompting' and start understanding the mechanics of how these models are built and trained. By learning the language of AI, industry leaders can stop being passive users and start asking the right questions about data sovereignty, ownership, and the real limitations of these systems. This kind of informed skepticism is far more valuable than simply mastering a single software application. Right now, the industry is stuck in a defensive crouch, focusing almost exclusively on using AI to cut costs. While trimming expenses is fine, it is the absolute floor of what this technology can achieve. The real opportunity lies in using AI as a growth engine—think about scaling global distribution, automating complex translation workflows, and reaching entirely new audiences that were previously inaccessible due to language or logistical bottlenecks. When it comes to the relationship between publishers and AI companies, the answer isn't a simple 'friend or foe' dynamic. It is both. Publishers have a legitimate moral and legal case to fight for compensation regarding how their copyrighted works were used to train these models. However, that doesn't mean they shouldn't also be at the table negotiating partnerships to create new, innovative distribution channels for the future. Ultimately, AI education needs to be a non-negotiable requirement for anyone in a management position. You cannot steer a publishing house through this digital transformation if you don't understand the structural capabilities and risks of the technology you are deploying. The leaders who win in the next decade won't be the ones who just use AI; they will be the ones who understand it well enough to ask the right questions and build a strategy that goes beyond simple cost-cutting.