JVM in the Age of AI

Artificial intelligence is here to stay, whether we like it or not. Its continuous development brings both opportunities and challenges, requiring us to take greater responsibility for the world around us. AI can also have a darker side—and it doesn’t have to involve a rogue superintelligence deciding to wipe out humanity like in the movies. In this talk, I’d like to highlight a few (perhaps) less obvious aspects of AI. And one final question: can you tell whether this abstract was generated by AI or not?

Are you already tired of the hype around yet another “LLM-as-a-Service”?

So instead of introducing yet another tool for prompt engineers, I want to talk about something genuinely interesting—what actually needs to happen inside the virtual machine itself for the JVM to become a solid platform for Machine Learning and AI.

While the entire industry argues about models, we’ll look under the hood: how hardware is evolving and what challenges this evolution creates for the JVM; what projects like Valhalla and Babylon will bring; what standardization efforts such as JSR381 – Visual Recognition API look like; and where TornadoVM fits into all of this.

This will be a high-level overview of how the JVM can catch up with—and perhaps even surpass—the demands of modern artificial intelligence and machine learning.