KBall and returning guest Tejas Kumar dive into the topic of building LLM agents using JavaScript. What they are, how they can be useful (including how Tejas used home-built agents to double his podcasting productivity) & how to get started building and running your own agents, even all on your own device with local models. Join the discussionChangelog++ members save 7 minutes on this episode because they made the ads disappear. Join today!Sponsors:Socket – Secure your supply chain and ship with confidence. Install the GitHub app, book a demo or learn more Neon – Fleets of Postgres! Enterprises use Neon to operate hundreds of thousands of Postgres databases: Automated, instant provisioning of the world’s most popular database. Featuring:Tejas Kumar – Twitter, GitHubKevin Ball – Twitter, GitHub, LinkedIn, WebsiteShow Notes: Bringing the vibe Andrew Ng on Agentic workflows Ollama Mixtral 8x 22b Stately Agent Vercel AI SDK Tejas’s talk on rebuilding React Hugging Face Jupyter Notebooks Datastax Langflow Something missing or broken? PRs welcome!

Jesús Espino, a tech expert from Mattermost, dives into his enlightening journey through the Go source code, revealing six of his memorable 'aha moments.' He explores the secrets of goroutines and their efficiency, and sheds light on Go's compiler optimizations like inlining. Jesús also discusses intriguing concepts such as slices, maps, and channels, and compares Go's memory management to Python's. Expect fascinating insights into programming tools and the evolving landscape of software development.

Kyutai, the creator of Moshi, the first real-time AI voice assistant, shares groundbreaking insights on his innovative technology. He discusses the potential of open-sourced voice assistants and their implications for future AI integration. The conversation dives into Gartner's hype cycle for generative AI, addressing organizations' current disillusionment despite heavy investments. The hosts also ponder the transformative power of AI in business efficiency and creativity, likening it to the internet's disruptive journey.

Data and tech expert Benn Stancil discusses the modern data stack, challenges in data analytics, and the effectiveness of analytics teams. He also delves into making engaging conference talks and the benefits of the Neon database platform for developers.

Computer security expert Marcus J. Ranum discusses the six dumbest ideas in computer security. They also cover why story points are useless in software development, the efficiency of the Posting HTTP client, instant branching for Postgres databases, ES Toolkit upgrade, and reflections on ten years of Varnish project work by Poul-Henning Kamp (phk).

Shawn “swyx” Wang discusses the state of DevRel, AI Engineering, AGI speculation, and the next innovation round of AI. Topics include DevRel's challenges, the importance of providing value, efficient secrets management, defining DevRel roles, successful AI engineering companies, generative AI, and the revolution in AI engineering with Devin.

Discussion includes decentralized Git design, challenges in self-hosting Git servers, backend of gitea.com, use of Sentry for error monitoring, quirky fasting stories, GitT platform exploration, managing large binary files, S3 implementation in CI/CD, frontend development with Retool, changing ownership models in products, enhancing Git workflow with Git Extras

Paul Copplestone, CEO of Supabase, discusses Supabase's focus on Postgres as the base-layer substrate for their platform. The podcast covers topics like PG vector integration, product development challenges, incremental software changes, and the evolution of Superbase within the Postgres ecosystem.

Members of the Ember core team discuss reactivity, build systems, and the longevity of the Ember community. They delve into topics like web development build steps, EmberData, framework evolution, and the importance of upgradability for app longevity.

Roie Schwaber-Cohen discusses the advantages of vector databases in machine learning pipelines. Topics include Pinecone's efficient storage and retrieval of vector data, the importance of vector databases, and the evolution of Pinecone Serverless for powerful AI applications.