WF2026: Software Factories & Keynotes ft. Microsoft, OpenAI, OpenClaw, Z.ai (GLM), MiniMax, HF
summary
Skip the fluff unless the talk names what was built, how it’s measured, and the actual results. This stream delivers that: Microsoft demos Foundry IQ plus an agent optimizer that hill‑climbs instruction sets and swaps configs into production; Zoo labs show OpenClaw: no handwritten code creating deployable apps; Z.ai / Minimax touts 1M token context and multimodal M3; OpenAI reveals Codex app and an open Codex harness and cites 7,000 attendees — watch if you want those tools, configs, and numbers, skip otherwise.
AI-written summary · about
chapters (58)
- 0:00Introduction and context
- 1:08The AI engineer's role
- 2:58Understanding loops
- 4:44The human loop
- 6:19Conference logistics and vision
- 8:47The software factory era
- 11:27AI and knowledge integration
- 24:23Learning loops in practice
- 28:37Empowering engineers
- 34:40Product modalities and codex
- 38:25Open ecosystem layers
- 44:00Value maxing and performance
- 48:18The agentic future
- 54:26Future generations and Z.ai
- 58:55Model performance and open source
- 1:10:17Multimodal agents
- 1:14:34Architectural research and native multimodality
- 1:21:17The research culture
- 1:32:23Security in the agentic era
- 1:37:10Reproducibility and observability
- 1:44:05Browser agents and workflows
- 1:48:59Practical automation principles
- 1:53:38Knowledge management and automation
- 1:58:40Agentic team coordination
- 2:03:50The real world of agent loops
- 2:10:06The software factory paradigm
- 2:15:33Orchestrating agent workflows
- 2:20:43Context management and readiness
- 2:26:18Human roles in the factory
- 2:31:02Browser agent benchmarks
- 2:33:59Browser automation demonstration
- 2:36:45The agentic AI engineer workflow
- 2:40:37Engineering principles for building
- 2:44:17Collaboration and automation
- 2:47:32Principles for high-velocity teams
- 2:50:14Building a brain for the factory
- 2:53:58Memory, layering, and agents
- 2:57:43Agentic coding in enterprise
- 3:00:34Statistical evaluation of agentic code
- 3:03:35Autonomous code validation
- 3:05:40Thinking like an agent
- 3:06:50Reimagining mobile development
- 3:09:00A computer-controlled workflow
- 3:11:07The story of machine craft
- 3:11:44Interview with a technical founder
- 3:13:44Infrastructure and VC philosophy
- 3:15:21Production evaluation and reliability
- 3:16:26Hierarchy of failure modes
- 3:17:27Anti-gravity and the agentic CLI
- 3:20:53Primitives for the 2026 agent era
- 3:23:46Building an autonomous kernel
- 3:27:17A durable execution platform
- 3:30:15Agentic design and simulation
- 3:33:35The vision for autonomous software
- 3:35:36Combatting development slop
- 3:36:53Building agentic maintainability
- 3:38:30New primitives for agentic code
- 3:40:47Closing thoughts
related talks
09h 11mAIE Europe Keynotes & OpenClaw ft Deepmind, OpenAI, Vercel, @pragmaticengineer , @mattpocockuk
33mKeynote: Fine-tuning our way towards openness in AI
07h 54mAIE Miami Keynote & Talks ft. OpenCode. Google Deepmind, OpenAI, and more!
58mAI in Reality Fireside Chat: Enterprise AI & Open‑Source Innovation
59mKeynote: AI without the BS, for humans
17mKeynote Analysis
06mSponsored Keynote: Building Your AI Factory: Wh...
27mQ&A #2
