Oracle, Java’s top defender, sees Java as having a “triple play advantage” in AI, leveraging the integration of AI services with cloud infrastructure, frameworks, and business logic. It also plans to “make Java even better” for native AI, along with integration with enterprise data and cloud services.
“It’s a sign of success that there are already rich frameworks and tools for Java developers to leverage AI services,” said Donald Smith, vice president of product management for Oracle’s Java platform. “Java developers can take advantage of all the benefits of Java—strong typing, memory safety, excellent core libraries, and more—when they use these frameworks, not to mention that most enterprise business logic already exists in Java.”
Java technology vendor Azul also sees a bright future for Java in AI. “As AI becomes more integrated into traditional business logic and what needs to be done at the application level, Java’s sweet spot and popularity will only grow,” Azul CEO Scott Sellers predicted. “Python is very limited in terms of performance and scalability,” he added.
Analyst and research vice president of IDC Software Development, Analyst Anna Dayaratna, said Java’s immense popularity has led to its growing prominence in AI. “Java remains the world’s most popular programming language, which is why it’s so important for AI development,” Dayaratna said. “Furthermore, Java is the most widely used language within enterprises, especially for production-grade and mission-critical applications.”
Dayaratna said that while Java currently lags behind Python in popularity for machine learning development, he expects it to be increasingly used for AI and generative AI development as applications move from proof-of-concept (POC) phase to production grade.
The native Java AI framework mentioned by Oracle’s Smith is Tribuo, LangChain4j, CoreNLPThere is. Tribuo is a machine learning library written in Java, providing tools for classification, regression, clustering, model development, and other functions. Langchain4j is a Java version of the Langchain framework for building applications powered by large language models, and aims to simplify the integration of LLM into Java applications. And CoreNLP provides a set of tools for performing natural language processing in Java.
Oracle’s ambitious plans for Java AI are part of the OpenJDK project, which aims to interconnect the JVM and native code. Project Panamaan embeddable, high-performance Python 3 runtime for Java. GraalPyIt’s about integrating AI services with business logic. “Just as Java has expanded with new technologies over the last 30 years, we expect to see more integration support over time,” Smith says. “Innovations in Java projects like Valhalla, Babylon, and Panama are helping Java to run closer to native computing, which has become synonymous with generative AI.”
IDC’s Dayaratna believes Java has a “high probability” of replacing Python for machine learning development. “Java is widely recognized as being faster and more performant than Python,” he said. “As enterprises begin to leverage generative AI, especially for more production-grade use cases, Java will increasingly gain traction due to its advantages around resource consumption, application performance, execution speed, and security.”
Additionally, Dayaratna added, “The fact that the Java community is investing heavily in improving Java syntax and making it easier to learn will also be another driver for increasing Java adoption for generative AI development.”
editor@itworld.co.kr
Source: www.itworld.co.kr
