JetBrains

What happens when one programming language quietly influences Java, Kotlin, Swift, and the broader JVM ecosystem? In the latest Marco Show episode, Martin Odersky (creator of Scala and one of the architects behind Java generics) talks about functional programming on the JVM, the evolution of Java, AI-generated code, and why future programming languages may need entirely new safety models for AI agents.

📺 Full episode: https://youtu.be/Xn_YpUtXWT4
Also available on your favorite podcast platforms.

Mellum started as a model for code completion.
Mellum2 expands far beyond that.

Now open source on Hugging Face, Mellum2 is built for modern AI workflows in software engineering, including routing, RAG, sub-agents, summarization, and real-time inference tasks.

Its new mixture-of-experts (MoE) architecture is optimized for ultra-low-latency and high-performance inference, making it practical for production deployments at scale.

Why teams are exploring Mellum2:

• Efficient inference for real-time workloads
• High-throughput production performance
• Local and self-hosted deployment options
• Full control over infrastructure and data
Available now under the Apache 2.0 license.

Learn more: https://jb.gg/t9z9qd