← Back to Videos

AI Engineer World's Fair 2026

Three days of mainstage keynotes in San Francisco — per-speaker notes and a daily summary for each day. Transcript-grounded.

Day 1 · Jul 1, 2026 — Software Factories & opening keynotes. Nine mainstage talks from Swix, Microsoft, OpenAI, OpenClaw, Z.ai, Hugging Face × MiniMax, HumanLayer, Linnet's Labs, and Cursor.

Daily Summary — Day 1
AIE World's Fair 2026 — Day 1 Recap: Software Factories & Loopcraft

Day 1 of AI Engineer World's Fair 2026 gathered ~7,000 attendees under the banner of Software Factories & Loopcraft. Swix opened by framing AI engineering as the craft of stacking loops; the mainstage then moved through knowledge and retrieval (Microsoft), open agentic stacks and new frontier models (OpenAI Codex, Z.ai GLM, Hugging Face × MiniMax), and orchestration at scale (OpenClaw). It closed on a sharp debate about the limits of automation — HumanLayer and Linnet's Labs pushing back on slop and unsafe agents, and Cursor showing where recursive model self-improvement is heading.

Read the full Day 1 recap →

Keynote Talks

AIE World's Fair 2026 — Opening Keynote: Software Factories & Loopcraft

AI ENGINEERDay 1 · Keynote

Swix (Shawn Wang) opens the AI Engineer World's Fair 2026 by framing the day's theme around software factories and his essay Loopcraft — the idea that AI engineering is fundamentally about stacking loops and knowing when to move up a loop for scale or down a loop for reliability. He extends the loop metaphor from code to human life and civilization, argues agents are generalizing beyond coding into vertical domains, and celebrates the event's growth to roughly 7,000 attendees. He closes by calling the World's Fair itself the highest loop — "the loop that makes loops," where humans gather to figure out what the next loop is.

AIE World's Fair 2026 — Microsoft: On AI and Knowledge

AI ENGINEERDay 1 · Keynote

Pablo Castro (CVP & Distinguished Engineer, Microsoft) frames the AI-and-knowledge problem through three lenses: intrinsic knowledge baked into models, extrinsic knowledge that agents retrieve to ground themselves, and learned knowledge captured by continuously improving agents. He traces the exponential arc from IntelliSense (1996) through ML ranking and GitHub Copilot to today's agents, then demos how Microsoft Foundry, Microsoft IQ, and Azure AI Search deliver company-wide grounding and agentic retrieval. He closes on a real learning loop via Foundry's agent optimizer, which hill-climbs an agent's instructions against an evaluation.

AIE World's Fair 2026 — OpenAI: The Codex Keynote

AI ENGINEERDay 1 · Keynote

Alexander Embiricos and Roman Huitt of OpenAI argue that far from replacing engineers, AI is triggering a return to engineering's roots — "AI engineers are eating the world." They frame Codex as a layered, open stack (model, responses API, open-source harness, apps server, app-layer plugins) that OpenAI uses internally exactly as it ships to developers, and preview the GPT 5.6 series across cost-efficient (Terra, Luna), fast (Codex Spark, Cerebras at 750 tok/s), and frontier variants. The talk closes on "value maxing" — extracting real value from agents through cost, speed, and parallel cloud execution rather than "token maxing."

AIE World's Fair 2026 — OpenClaw: Better Loops, Not More Terminals

AI ENGINEERDay 1 · Keynote

Peter Steinberger ("the claw father," now at OpenAI) argues that the era of juggling 10+ terminal windows was a false peak of productivity where the human was really acting as scheduler, router, and memory. He describes a shift toward managing a long-lived manager agent that delegates to a team of workers, enabled by three changes: server-side compaction, coordination, and automated triggers. As agents get better, the human bottleneck moves from tokens to compute to attention — so the real work is deciding where to spend it. His closing thesis: "The future is not 20 terminals, it's better loops."

AIE World's Fair 2026 — Z.ai: GLM 4.2 / 5.2

AI ENGINEERDay 1 · Keynote

Zishan Lee of Z.ai dialed in remotely to introduce GLM 5.2 (and GLM 4.2), the company's latest open-weight models. He explained that GLM stands for General Language Model — a name rooted in a 2021 paper on autoregressive blank-filling — and positioned GLM 5.2's coding and agentic capability as sitting between Opus 4.7 and 4.8, with a new "high" thinking level for better token efficiency. Lee argued the case for open weights (security, on-prem control, fine-tuning, co-designing the future) and closed with a "one more thing": Zcode, Z.ai's own Codex-like coding harness.

AIE World's Fair 2026 — Hugging Face × MiniMax: Open Models & the M3 Fireside

AI ENGINEERDay 1 · Keynote

Thomas Wolf, co-founder and Chief Science Officer of Hugging Face, opens by mapping the open-model landscape — GLM at #2 on Artificial Analysis and the field of Chinese "AI dragons" (DeepSeek, Moonshot/Kimi, Z/GLM, MiniMax) — then hosts a fireside with MiniMax's Olive on M3. Released in June as the top open-source model at the time, M3 is a compact ~400B total / ~20B active (cited as 428B/23B) model with native multimodality (text, image, video) and a functional 1M-token context powered by MiniMax Sparse Attention (MSA). The conversation traces MiniMax's long-context lineage (M1, MiniMax-01), its from-the-first-step multimodal training, its intern-driven research culture, and its commitment to open source.

AIE World's Fair 2026 — HumanLayer: Harness Engineering Is Not Enough

AI ENGINEERDay 1 · Keynote

Dex Horthy (HumanLayer) argues that the race to build fully automated "software factories" is producing slop because the limits are not in the harness but in how coding models are trained. He shows that reinforcement learning rewards making tests pass, not maintaining codebase quality, so no amount of loop or harness engineering can fix what is fundamentally a model-training problem. His answer is to "turn the lights back on": use AI-assisted product review, architecture, program design, and vertical slices up front so humans can still read every line and move fast safely.

AIE World's Fair 2026 — Linnet's Labs: Eric Meyer: Provably Safe Agents: Using Type Systems and Proof-Carrying Code to Tame AI

AI ENGINEERDay 1 · Keynote

Eric Meyer, research scholar at Linnet's Labs, delivers a tongue-in-cheek but technically serious tutorial arguing that AI agents are intrinsically dangerous the moment you give them tool calls, and that alignment baked into model weights is not a real safety guarantee. His thesis: treat safety as a programming-language problem. Instead of letting an agent execute its side-effecting loop directly, have the model emit a program (an expression of type IO) that you can statically analyze, taint-check, and formally prove safe before running. He frames this as a repackaging of 1990s proof-carrying code.

AIE World's Fair 2026 — Cursor: Model Behavior: Recursive Model Improvement: How Cursor Trains Composer (and Teaches It to Train Itself)

AI ENGINEERDay 1 · Keynote

Lee Robinson, an ML engineer on model behavior at Cursor, walks through how Cursor trains its Composer models via a two-part flywheel: an outer loop of real-world feedback and A/B tests, and an inner loop of high-quality evals and hard, verifiable training problems. He details reward-hacking discoveries, a novel textual feedback RL method, and a SpaceX/Colossus compute partnership. The thesis: as the top model gets smarter, it can distill derivative models that speed up every part of training — a path toward recursive self-improvement.