This episode analyzes Thinking Machines' Inkling (975B MoE), Meta’s cost-efficient Muse Spark 1.1, and Anthropic’s J-space paper, which reveals internal model processing for enhanced AI safety.
Read article
Odevo shares its AI-native transformation journey, detailing how adopting agentic coding requires fundamental process changes—not just licenses—to achieve massive gains in software delivery speed.
Read article
Android Studio Quail 2 streamlines development by integrating parallel agent tasks (Gemini), built-in Leak Canary memory leak detection, and App Quality Insights for automated crash fixing.
Read article
Lila Sciences is building scientific superintelligence by treating the wet lab as an RL verifier. The goal is to create a general reasoning model trained on 10 trillion experimentally-verified tokens across biology, chemistry, and materials.
Read article
CQ (Stack Overflow for agents) standardizes knowledge sharing across AI agents. It allows lessons learned from solving novel problems to be captured as Knowledge Units (KUs), preventing repeated mistakes in automated workflows.
Read article
Inkling is a massive, open-weight multimodal model with 1T parameters and 1M context window. It supports BF16/NVFP4 formats and integrates seamlessly with vLLM, SGLang, and llama.cpp for robust deployment.
Read article
Omnigent introduces the concept of a 'meta-harness' layer, allowing developers to compose AI agents by abstracting away model changes. The discussion covers agent debate, tokenomics as DevOps, and the critical need for stateful memory.
Read article
Managing AI coding skills requires treating them as first-class software assets. The talk outlines best practices to prevent 'skill sprawl,' emphasizing centralized registries, automated reviews, and dedicated lifecycle management.
Read article
Google DeepMind used Gemini Omni and Veo to reconstruct Pelé's legendary 1959 goal. The project showcases advanced AI techniques for cultural preservation, combining motion modeling with deep historical research.
Read article
Patrick Debois maps AI-native development maturity, outlining a progression of patterns (Prompting $\rightarrow$ Context $\rightarrow$ Harness $\rightarrow$ Loop) and emphasizing that organizational success requires platform enablement and measuring shared component contributions.
Read article
Pi Agent offers superior flexibility over competitors like Cloud Code and CodeX. Its core strength is its extension system, allowing programmatic modification of tools, contexts, and hooks to build highly customized, autonomous AI systems.
Read article
Engram discusses why long context and RAG break down when dealing with trillions of proprietary tokens. Solutions involve 'cartridges' for knowledge compression and 'test-time training' to build true AI intuition.
Read article
Patrick Debois discusses scaling autonomous coding agents, arguing that success requires shifting focus from individual code fixes to improving the underlying system and platform architecture.
Read article
Forget loop engineering. Agentic Engineering is about building AI developer workflows inside your software factory, orchestrating engineers, agents, and deterministic code.
Read article
Learn how to build self-improving AI loops that automate complex tasks. We cover the Loop Contract, advanced triggers (Combo/Workflow), and multi-agent verification systems for CI.
Read article
Joseph Katsioloudes demonstrates how AI can scale scarce security expertise by integrating advanced tools like MCPs and agentic workflows into the CI/CD pipeline. He details best practices for secure coding at scale.
Read article
Learn how to apply AI-Native Engineering practices to brownfield codebases, moving from maintenance mode to evolution. Discover the Strangler Fig Pattern and Branch by Extraction.
Read article
AGI must emulate collective human intelligence, moving beyond narrow tasks like chatbots. The goal is to build 'perception agents' that model user intent and achieve cognitive alignment.
Read article
Building large systems with AI requires more than 100% test coverage. Observability and using existing 'test articles' (like S3) are key to enforcing correctness in distributed environments.
Read article
Build Small revealed amazing hackathon winners, showcasing how small models (SLMs) can power massive utility. Learn about local-first AI, 64% offline capability, and advanced techniques like GGUF quantization.
Read article