The Hallway Track
Every signal tracked from the world's AI conferences — 4,665 talks, quotes & announcements, newest first.
Microsoft's shrinking approval time from open-sourcing TypeScript to Chakra reflects a major internal culture shift toward open development.
“Developer products really amass a fan base and really actually become better products when they are built in the open.”
Apple hosts a beginner Q&A lab on SwiftUI, its declarative UI framework for building apps across Apple platforms.
“SwiftUI is Apple's declarative user interface framework used to build apps across all of Apple's platforms.”
RAG isn't dead; hybrid tool-rich retrieval is becoming the default for serious agentic search.
“rag is dead, how hybrid tool tool rich retrieval is becoming a default for serious agentic search.”
Benchling makes everyone on the team review AI traces through structured rotations, user feedback, and feature ownership.
“we have a weekly "fire chief" rotation who's addressing issues”
AWS demos hands-free insurance claims intake using Strands Agents and Bedrock AgentCore Browser Tool for automated FNOL processing.
“The result is automation of manual screen work while preserving human oversight and auditability.”
NVIDIA introduces Auto-FL, using AI agents to automate federated learning experiment design and analysis in NVIDIA FLARE.
“Federated learning (FL) research often begins with a deceptively simple question: What should we try next?”
Google DeepMind launched Gemini 3.1 Flash Live, a full-duplex real-time sound-to-sound conversational model, plus Gemma 4 with on-device audio understanding.
“we recently launched Gemini 3.1 flash life, which is our kind of full duplex, you know, sound to sound real-time conversational model”
LangChain launches LangSmith Fleet, letting anyone build and manage no-code agent fleets via chat.
“With Fleet, we skip the "if this, then that" canvas entirely and go straight to a fully agentic system that combines significant power with the simplest creation experience: chat.”
AWS shows how to build an agentic incident triage assistant using Amazon Quick, New Relic MCP Server, and Asana.
“From a single prompt, the Amazon Quick agent investigates the incident, assembles a root cause analysis (RCA) brief with evidence links, and creates a tracked Asana task ready for handoff.”
Cohere launched North Mini Code, its first model built specifically for developers.
Google DeepMind launches Gemini 3.5 Live Translate for near real-time natural voice translation.
Rick Rubin's success stems from genuine taste and being a 'professional listener' rather than waiting to respond.
“You're a professional listener.”
Google launches Gemini 3.5 Live Translate for developers via API, enabling real-time speech-to-speech translation across 70+ languages.
“The model is now available on the Gemini API and in AI Studio.”
Microsoft's Quantum Development Kit March 2026 release adds cross-language program composability across Q#, Python, and OpenQASM.
“We're making it possible for people to vibe code their quantum.”
Microsoft's Quantum Development Kit integrates GitHub Copilot in VS Code to write, test, and run quantum code from natural language.
“In a few minutes, I had Copilot write some quantum code starting from the simple image of a circuit, refactor it, test it, explain some of the algorithm to me, explain some other quantum concepts, fix a bug, run a simulation, connect to the service, submit the job, and fetch the results. All without me writing a single line of code.”
DeepSeek's Liang Wenfeng built a competitive AI lab by stockpiling Nvidia A100s and reworking training to reduce GPU reliance after US export controls.
“Look, the way this works is we're going to tell you it's totally hopeless to compete with us on training foundation models. You shouldn't try, and it's your job to like try anyway. And I believe both of those things.”
NVIDIA shows how Agent Skills and Nemotron Speech models speed up evaluation of clinical ASR systems on hard medical terminology.
“Off-the-shelf speech systems can sound fluent and still miss the words”
Apple's Foundation Models framework now offers frictionless frontier-model access via Private Cloud Compute, plus agentic coding in Xcode.
“the ability to now just access one of these frontier models in Private Cloud Compute with basically no setup or friction at all”
Google DeepMind announces Gemma 4 12B, a unified encoder-free multimodal open model.
Google DeepMind outlines efforts to power the future of robotics in Europe.
Automotive engineering firm AVL uses Databricks Lakehouse to manage testing data, with generative AI as a supervised co-pilot for engineers.
“Generative AI, they need a supervisor. The agent is the co-pilot.”
Transformers.js uses ONNX as an open standard to run models portably across devices and backends.
“ONNX says what to compute. ONNX runtime decides how to compute it.”
Nextdoor engineers use OpenAI Codex with GPT-5.5 to debug, build cross-platform, and focus on product outcomes.
An AI agent autonomously built a 3D Paris gallery by chaining two Hugging Face Spaces together.