What Do Java Developers Think of the Rise of GenAI?

Python is known as the go-to programming language for generative AI, so naturally it has been hyper-focused on Python frameworks and libraries. But what does this mean for Java developers?

It’s an important question to ask, as a lot of the world’s largest IT systems are powered by Java. Java has proven itself to be the language of choice when it comes to system scalability and robustness. With Java’s six-month release cadence and new features such as those introduced by Project Loom and Project Panama, we can expect Java to keep getting even more powerful.

We decided to ask Java developers their thoughts about the rising importance of GenAI with respect to the higher adoption rate of Python. Java developers have traditionally been identified as enterprise application developers, who tend to be specialists in charge of the design and implementations of a corporation’s backend production systems. With ChatGPT taking the world by storm, has that been raising the eyebrows of Java developers — or are they not paying any attention at all?

Survey Methodology

We asked Java developers how they felt about GenAI being hyper-focused on Python. We had the survey up for a week on three different social media platforms: LinkedIn, X (formerly Twitter) and Mastodon.

The specific question asked was “GenAI is hyper-focused on Python libraries and frameworks. How does this leave you feeling?”

With LinkedIn and X being the mainstream platforms, we saw a higher number of engagements both in the number of views and votes received. The LinkedIn poll had close to 2,900 of viewers , of which 84 (3%) voted on the question. X had almost 2,800 viewers, of which 134 (5%) voted. Mastodon did not display the number of viewers, but it registered 15 people who had cast their votes — for a grand total of 233 responses to the question.

Due to the limitations of the sample, the results should not be extrapolated to the larger population. That being said, the data does allow the community to have a better understanding of the subject.

The Observations

Overall, 48% of respondents believe “Java will catch up” with Python. In contrast, 18% said that Python’s GenAI edge means they are more likely to “switch to Python.” Among the remaining respondents, 21% are not sure what to think and 13% plan to start researching the subject. Voters on LinkedIn were more likely to believe Java will catch up and less likely to plan to do additional research.

👁 Image

Here are a few conclusions we can draw:

• The surveys show that a considerable number of developers feel surprised, concerned or threatened by Python’s dominance in generative AI. There is a sense that Python has “taken over” this emerging field.

• At the same time, some Java developers see opportunities to leverage Java’s strengths (performance, static typing, enterprise capabilities) to complement Python’s traction in AI/ML. There is interest in bridging the gap between the two languages.
• Opinions differ on whether Java can “catch up” to Python in AI or if the momentum is too far on Python’s side now. Some think Java needs to rapidly evolve to be competitive, while others think it’s too late.
• There are calls for Oracle and the Java community to be more proactive in supporting AI, machine learning and data science use cases. Many feel Python’s ecosystems and libraries for ML/AI are far ahead.

So What Does the Future Hold?

Results from the brief survey have given us some level of confidence that Java developers are more ready to start learning about GenAI. Since it was initially developed using Python, it makes sense that it currently has better Python support and integration. Moving forward, however, expanding language support is likely a priority.

Java is still extremely popular, especially in large enterprises, despite the occasional naysayer’s comment of “Java is dead.” So there’s definitely incentive for GenAI to expand its Java capabilities to open up more use cases. Java and Python have different strengths that make them suitable for different tasks; they each have their place in the computing world. Java is known for performance, scalability, and concurrency support, among other things. The upcoming Java 21 will have virtual threads, as part of Project Loom, integrated into the release; this will bring concurrency computing to a new level. Python’s ease of use and faster development cycles will continue to be an advantage.

But Java isn’t standing still when it comes to ease of use. There are ongoing efforts to make the language more concise and developer-friendly. Libraries such as PyTorch and TensorFlow now have Java APIs for AI/ML development.

Ultimately, we think GenAI will provide a great experience for developers in both Java and Python. Supporting developer freedom of choice in languages and frameworks will be important — flexibility will always be advantageous for any wave of technological advances. Of course, how quickly GenAI expands its Java support depends on demand and technical considerations. The expectation is that this will improve over time; as noted above, the technical piece of the puzzle is already being solved.

In summary, while GenAI is more Python-focused today, Java is still very relevant and supporting it well should be a priority going forward. The languages can complement each other with their respective strengths. Who knows? The 18% of respondents who indicated that they’d switch to Python might eventually find themselves convinced to stick with Java.

Learn more about DataStax by subscribing to DataStax Developers on YouTube and following DataStax on LinkedIn; learn how DataStax enables generative AI here.