wal.sh
home > events > emtech-ai > 2026 · Jul 18, 2026
emtechmit technology reviewmit media labcambridgeaiagentschatgptopenaiwalmartascendionhuttenlocherintegrationhallucinationsapril 2026

EmTech AI 2026
MIT Media Lab, Cambridge, April 21-23, 2026

Table of Contents

Logistics

Overview

MIT TR's applied-AI event. Framing this year is "The Great Integration" — the move from pilots to production, and the reality check on hallucinations, autonomy, and reasoning limits.

Tuesday is a half-day opening; the Great Integration opening chapter runs 4-6pm with five back-to-back talks and a networking reception.

Schedule

Day 1 — Tuesday, April 21

  • 2:00 pm — Registration Opens
  • 2:30 pm — MIT Designs the Future (AI Leadership Roundtables)
  • 3:30 pm — Networking Break

The Great Integration (4:00 - 6:00 pm)

Opening chapter. Framing: AI is moving from experimentation to execution — from organizations already scaling AI across operations, to emerging platforms enabling unified intelligence, to the realities leaders must confront as hype meets hard limits.

  • 4:00 — Welcome Remarks
  • 4:10 — From Pilots to Production
    • Donna Morris, EVP, Chief People Officer, Walmart Inc.
    • Framing: moving AI from isolated wins to enterprise-wide impact — disciplined focus, cross-functional alignment, challenging entrenched processes; breaking out of pilot purgatory.
  • 4:35 — Exclusive First Look: The 10 Things That Matter in AI Right Now
    • Amy Nordrum, Executive Editor, MIT Technology Review
    • Niall Firth, Executive Editor, Newsroom, MIT Technology Review
    • MIT TR's newsroom presents on stage for the first time an overview of 10 key technologies, emerging trends, and movements in AI for
  • 5:00 — AI Engineering for the Last Mile (Ascendion-presented)
    • Karthik Krishnamurthy, CEO, Ascendion
    • Brian Bryson, Principal Technology Analyst, MIT Technology Review
    • Framing: the hardest challenge in AI is not building models — it's making them work in production. Embedding AI into workflows, decision processes, and systems.
  • 5:10 — The Rise of the AI Platform
    • Sulman Choudhry, Head of Engineering, ChatGPT, OpenAI
    • Framing: the next generation functions less like a tool and more like a platform — models, applications, human workflows converge. Conversational systems as integrated operating environments.
  • 5:35 — Hallucinations, Hype, and the Realities of AI
    • Daniel Huttenlocher, Dean, Schwarzman College of Computing, MIT
    • Framing: the gap between AI's perceived intelligence and its actual reliability, autonomy, and reasoning. Where systems mislead us; risks underestimated; realistic planning for breakthroughs and bottlenecks.
  • 6:00 - 7:30 pm — EmTech AI Reception (in-person only)

Day 2 — Wednesday, April 22

Full day. Four chapters: Agents at Work, Revenue Machines, The New AI Stack, Imagination to Implementation. Plus a hands-on afternoon workshop (in-person only).

  • 7:45 am — MIT Campus Tour
  • 8:00 am — Registration and Breakfast

Agents at Work (9:00 - 10:30 am)

Framing: AI agents are moving far beyond chatbots and simple automations — realities of deploying agentic systems at scale, the architectures required to coordinate them, and the frontier innovations defining the next generation of agentic intelligence.

  • 9:00 — Welcome Remarks
  • 9:15 — How to Build an Agentic Workforce
    • Kellie Romack, Chief Digital Information Officer, ServiceNow
    • The lived playbook: scaling from isolated agents to a fully agentic workforce across IT, security, and operations. What worked, what didn't, how value was measured.
  • 9:40 — The Signals That Matter — MIT Insider's Panel
    • Mat Honan, Editor in Chief, MIT Technology Review
    • Will Douglas Heaven, Senior Editor, AI, MIT Technology Review
    • James O'Donnell, Senior Reporter, AI, MIT Technology Review
    • The MIT TR editorial team on trends, tensions, technological shifts. Contrarian takes and early signals not yet in headlines.
  • 10:05 — The Agent Engine
    • Ash Edwards, Head of Forward Deployed Research Engineering, Poolside
    • Multi-agent pipelines, evaluation, interoperable systems for enterprise adoption. Poolside is a coding-agent company — the framing overlaps the site's CLI-coding-agents thread.
  • 10:30 - 11:00 am — Networking Break

Revenue Machines (11:00 - 12:25 pm)

Framing: AI now shapes the entire path to revenue — from how teams build products to how brands attract customers and sales organizations convert demand.

  • 11:00 — From Messaging to Conversion with AI
    • Andrew Bialecki, Cofounder & Co-CEO, Klaviyo
    • Autonomous agents handling marketing work formerly owned by humans: interpreting data, adapting messaging, acting inside real systems while preserving brand safety.
  • 11:30 — Refactoring How We Code
    • Ziniu Wu, Cofounder, G5 Labs; PhD '26, MIT CSAIL
    • Michael Abbott, Lead, Global Banking and Capital Markets, Accenture
    • Co-development patterns, guardrails, architectural shifts for AI- driven developer workflows. Direct hit on the site's AI-assisted coding thread.
  • 12:00 — Rewiring the Sales Engine for AI
    • Amit Bendov, CEO & Cofounder, Gong.io
    • Lead qualification, forecasting, real-time buyer insights.
  • 12:25 - 2:00 pm — Lunch and Networking Break

Afternoon Workshop (in-person only)

  • 1:05 pm — From Prompt to Agent: Build an Autonomous AI in 45 Minutes
    • Santanu Bhattacharya, Scientist, MIT Media Lab
    • Maria Gorskikh, Researcher, Project NANDA
    • Design, deploy, test a working agent — stop using AI as a chatbot. Adjacent to the site's tools tier (message-format, mutation, gc-viz, fair-queue as small hands-on artifacts).

The New AI Stack (2:00 - 3:50 pm)

Framing: A nontechnical guide to the core layers of the modern AI stack — trusted data, orchestrated workflows, secure systems.

  • 2:00 — Operationalizing AI for Scale and Sovereignty (HPE-presented)
    • Chris Davidson, VP HPC & AI Customer Solutions, HPE
    • Arjun Shankar, Division Director, National Center for Computational Science, Oak Ridge National Laboratory
    • Elizabeth Bramson, CEO and Publisher, MIT Technology Review
    • AI factories — data ownership vs safe/trusted data flow. Data control as a strategic imperative.
  • 2:30 — Building the Data Fabric for AI
    • Anahita Tafvizi, Chief Data and Analytics Officer, Snowflake
    • Governed data fabric: accessible, trustworthy, AI-ready enterprise data.
  • 2:55 — Building the Orchestration Layer for AI
    • Viren Baraiya, CTO & Founder, Orkes
    • Orchestration for multi-step operations across teams and systems.
  • 3:25 — Securing the New AI Stack
    • Michael Spisak, Managing Director Cybersecurity R&D, Unit 42, Palo Alto Networks
    • Data leakage, supply chain vulnerabilities, shadow AI, insecure code, model drift.
  • 3:50 - 4:15 pm — Networking Break

Imagination to Implementation (4:15 - 6:00 pm)

Framing: Where creative spark meets systems that scale. Visual effects, music, product design.

  • 4:15 — Cyber-Insecurity in the AI Era (GC Cybersecurity-presented)
    • Tarique Mustafa, Cofounder, CEO & CTO, GC Cybersecurity
    • Brian Bryson, Principal Technology Analyst, MIT Technology Review
    • Why security must be rethought with AI at its core, not layered on after the fact.
  • 4:30 — The Art of AI-Assisted Design
    • Zach Lieberman, Associate Professor, MIT Media Lab
    • AI as a creative partner in exploratory, human-centered design.
  • 4:55 — Rebuilding Creative Pipelines
    • Scott Mann, Cofounder & Co-CEO, Flawless
    • Human direction + AI-assisted production for scaled personalized content.
  • 5:20 — The Future of Human Expression
    • Duncan Crabtree-Ireland, National Executive Director and Chief Negotiator, SAG-AFTRA
    • Authorship, ownership, consent, control over creative work and digital likenesses in an AI-first world.

Day 3 — Thursday, April 23

Half day. Two chapters: Talent, Team, Transformation and Exploring AI's New Frontier. Closes with lunch and depart.

  • 7:45 am — MIT Campus Tour (in-person only)
  • 8:00 am — Editorial Breakfast

Talent, Team, Transformation (9:00 - 10:30 am)

Framing: How to reskill teams, rethink talent pipelines, and lead in a world where human and machine intelligence are inextricably linked.

  • 9:00 — Welcome Remarks
  • 9:10 — Leading the AI Organization
    • Kevin Delaney, Cofounder & Editor in Chief, Charter
    • Emerging leadership models and team designs; people and intelligent systems working together.
  • 9:35 — The Trust Recession
    • Peter Smart, CXO & Managing Partner, Fantasy
    • Framing: Consumers are turning to AI agents, not brands, to decide what to trust. Discovery, evaluation, loyalty in an agent-mediated economy — what to build so products remain findable and credible.
    • Direct AEO adjacency: this is the same shift UNBOUND 2026's Deep Dives frame as Answer Engine Optimization ("What 10,000 Hours of AEO Taught Us", "Inside HubSpot's AI Search Lab: 7 AEO Experiments"). Two conferences, same claim, different framing. See UNBOUND 2026 for the marketer's side of the story.
  • 10:05 — Reskilling the Frontline for the AI Era
    • Melissa Krchma, Director of Workplace & Digital Empowerment, Bluedog Design
    • Sabrina Mansur, Director, Massachusetts Artificial Intelligence (AI) Hub
    • Redesigning roles, reskilling distributed teams, balancing automation with empowerment.
  • 10:30 - 11:00 am — Networking Break

Exploring AI's New Frontier (11:00 am - 12:30 pm)

Framing: A tour of groundbreaking research from MIT labs — the realistic path toward advanced intelligence.

  • 11:00 — From Models to Medicines
    • Gevorg Grigoryan, Cofounder & CTO, Generate:Biomedicines
    • Generative AI turning biology into a design discipline — large models on protein data generating therapeutic candidates before lab testing.
  • 11:35 — Embodied AI Comes to Life
    • Mikell Taylor, Director, Robotics Strategy, General Motors
    • Aaron Prather, Director, ASTM International
    • Adaptable general-purpose robots — human assistance, collaboration, safety and alignment for embodied intelligence.
  • 12:00 — The Shape of Tomorrow's AI
    • David Cox, IBM Director, MIT-IBM Watson AI Lab
    • Insider view of early experiments and emerging capabilities. Closes EmTech AI.
  • 12:30 - 1:30 pm — Lunch and depart

Cross-references to current work

Sessions map cleanly onto ongoing research threads. Where a session lands adjacent to a note, the link is inline so the two can be read together once the recording drops.

  • From Pilots to Production (Walmart, Tue 4:10) — the enterprise-scale view of the same transition the site tracks at the individual/team level. Adjacent to agent-context-thread and agent-memory-institutional-knowledge — both frame "pilot → production" as a memory-and-context problem, not a model-capability one.
  • The 10 Things That Matter in AI Right Now (MIT TR, Tue 4:35) — this is a curated 2026 landscape read; expect it to overlap the site's quarterly reads. Cross-check against 2026-q2-cli-coding-agents, 2026-q2-claude-code-features, and 2026-agent-memory-systems once slides land — diff the 10 against what the site already tracks; the delta is worth writing up.
  • AI Engineering for the Last Mile (Ascendion + MIT TR, Tue 5:00) — the "last mile" framing is the site's operational thread. Direct adjacency to crowsnest (local agent-telemetry receiver) and experiments (feature-gate SDK) — both are last-mile plumbing that most vendor talks skip.
  • The Rise of the AI Platform (Sulman Choudhry, OpenAI, Tue 5:10) — ChatGPT-as-platform is the shape being copied everywhere; watch this one against 2026-q2-claude-code-features (the Claude Code quarterly), which tracks the parallel Anthropic answer.
  • Hallucinations, Hype, and the Realities of AI (Huttenlocher, MIT Schwarzman, Tue 5:35) — this is the most cited thread on the site. Direct adjacency to:
    • annotation-systems/spec — property-drawer verdict system (:VERDICT: corrected, disputed, needs-citation, verified, reproduced, speculative) built exactly for the reliability problem Huttenlocher will describe.
    • marginalia — the annotation workspace SPA that renders those drawers as editable cards.
    • The tamper-evident validation ledger .verify/chain.jsonl is the site's answer to "how do you know a claim was actually checked?"
  • How to Build an Agentic Workforce (Romack, ServiceNow, Wed 9:15) — the "playbook from experience" framing is what the site's agent-memory-institutional-knowledge note argues for at the individual/team scale. ServiceNow's enterprise view is the missing size-tier.
  • The Agent Engine (Poolside, Wed 10:05) — Poolside is a coding-agent company; the framing (multi-agent pipelines, evaluation, interoperable systems) is the direct enterprise counterpart to 2026-q2-cli-coding-agents.
  • Refactoring How We Code (Wu + Abbott, Wed 11:30) — co-development patterns, guardrails, architectural shifts for AI-assisted dev workflow. Diff against 2026-q2-claude-code-features to see how Claude Code's approach maps to the taxonomy.
  • Securing the New AI Stack (Spisak, Palo Alto Networks Unit 42, Wed 3:25) — data leakage, supply chain vulnerabilities, shadow AI, insecure code, model drift. Adjacent to the site's bot-detection and compliance work under bots/compliance-spec.
  • The Trust Recession (Peter Smart, Fantasy, Thu 9:35) — the single strongest cross-conference cable in this note. The framing ("consumers turn to AI agents, not brands, to decide what to trust") is the same shift UNBOUND 2026 frames as Answer Engine Optimization ("What 10,000 Hours of AEO Taught Us", "Inside HubSpot's AI Search Lab: 7 AEO Experiments"). Two conferences (MIT Tech Review, HubSpot), three months apart, same claim from different angles. Read the transcripts side-by-side. See UNBOUND 2026 AEO thread.
  • Leading the AI Organization (Kevin Delaney, Charter, Thu 9:10) — the "team designs for human + intelligent systems" framing pairs with agent-context-thread.
  • The Shape of Tomorrow's AI (David Cox, MIT-IBM Watson AI Lab, Thu 12:00) — closing MIT-Insider read. Diff the 10-Things (Tue 4:35) with Cox's forward-looking set: what MIT+IBM's research pipeline says is arriving next. Cross-check against 2026-agent-memory-systems.

Overlap alert

RubyConf 2026's Who are we Now? (Jessica Kerr, Tue Jul 14 5pm) picks up the same identity-under-agentic-tooling thread that EmTech's opening chapter frames as The Great Integration. Worth reading the two together once transcripts drop.

Sponsors (Tue slot annotations)

  • Ascendion — Presenting Partner (sponsors the Tue 5:00 session).
  • Additional partner tiers listed on the source page but not enumerated here (Event Partner, Experience Collaborator, Strategic Partner) — refresh when named.

Adjacency to earlier notes

  • Same Cambridge venue used by GitHub Copilot Dev Days Boston (Apr 29) and AI Tinkerers Boston (Jun 29) — plan a single Cambridge pass in Apr-May-Jun if attending.
  • MIT TR's other 2026 event (Future Compute) is not yet on the tracker; worth adding if the AI overlap is meaningful.

TODO

  • [ ] Refresh once Wed/Thu agendas hydrate.
  • [ ] Diff the 10 Things That Matter in AI list against the site's quarterly reads once slides drop; write up the delta.
  • [ ] Cross-link Huttenlocher's talk into docs/annotation-methodology-review.org if the "realities of AI" framing offers named refutation criteria we can adopt.
  • [ ] Pull the Sulman Choudhry (OpenAI) framing of "AI Platform" into the Claude-Code quarterly for direct comparison.

Sources