← Head back to all of our AI Engineer World's Fair recaps
GitHub Next Explorations
Rahul Pandita @pandita_rahul / GitHub
Watch it on YouTube | AI.Engineer Talk Details
This talk by Rahul Pandita from GitHub Next discusses their approach to exploring the future of software engineering with AI, focusing on two key experiments: Copilot next edit suggestions and Copilot workspace. He explains how these tools aim to enhance code completion, task management, and overall developer productivity, while emphasizing the importance of exploration and user feedback in shaping AI-assisted development tools.
Embeddings are Stunting Agents: How Codeium Breaks Through the Ceiling for Retrieval
Kevin Hou @kevinhou22 / Codeium
Watch it on YouTube | AI.Engineer Talk Details
This talk by Kevin from Codeium discusses how traditional embedding-based retrieval methods are limiting AI agents, particularly in code generation. He introduces Codeium's approach called Mquery, which uses parallel LLM calls for more accurate and context-aware code generation, made possible by their vertically integrated infrastructure and custom models.
Cursor: Building the Human-AI Hybrid Engineer
Michael Truell @mntruell / Cursor
Sualeh Asif @sualehasif996 / Cursor
Watch it on YouTube | AI.Engineer Talk Details
This talk by the founders of Cursor discusses their vision for the future of programming in the age of AI, focusing on their code editor that integrates AI to enhance programmer productivity. They explain features like Copilot++ for predicting next edits and Command K for instruction-based code changes, while also discussing their technical approaches and future plans for transforming software engineering.
⭐ The AI emperor has no DAUs: why most devs still don't use code AI
Quinn Slack @sqs / Sourcegraph
Watch it on YouTube | AI.Engineer Talk Details
This talk by Quinn Slack, CEO of Sourcegraph, addresses the surprisingly low adoption rate of Code AI tools among developers, with only about 5% of professional developers using them regularly. He discusses the challenges in the AI market, shares lessons from building Cody (Sourcegraph's Code AI tool), and emphasizes the importance of creating practical, frequently used features while avoiding hype to increase adoption and build successful AI products.
Read our Deep Dive on this talk as well.
Code Generation and Maintenance at Scale
Morgante Pell @morgantepell / Grit
Watch it on YouTube | AI.Engineer Talk Details
This talk by Ragante, founder of GRIT, discusses how their platform enables large-scale code generation and maintenance using AI agents, focusing on automating complex code changes across thousands of repositories. He explains the technical challenges and solutions involved in making AI-driven code modifications reliable and efficient, including custom query engines, parallel processing, and optimized edit formats.
Self-Evolving Code with AI: Enhancing Quality and Security in CI
Gunjan Patel @gunjan_patel1 / Palo Alto Networks
Watch it on YouTube | AI.Engineer Talk Details
This talk is about using AI to help improve code after developers write it. The speaker describes a system that can automatically enhance code comments, add unit tests, find security issues, and suggest fixes - all without the developer having to do that boring work themselves.