The AI conference circuit has a well-documented problem: too many panels about what AI could do, too few about what it's actually doing in production. The Silicon Slopes AI Summit on March 31, 2026 is designed around the opposite premise.
"The conversation around AI has shifted from what's possible to what's actually working," said Michael Mallin, Silicon Slopes Vice Chair and founder of Model Forge AI. "This summit is about bringing together the people who are building real systems, solving real problems, and turning AI from research into reality."
The event — free to attend and expecting over 500 participants at the Zions Bancorporation Technology Center in Midvale — features speakers from NVIDIA's education division, Google's AI agents team, and Databricks, alongside researchers from the University of Utah, BYU, and Utah State University.
The Keynote: What AI Can't Replace
Craig Clawson, who spent years directing NVIDIA's Deep Learning Institute before moving into AI education, delivers the opening keynote under a deliberately contrarian title: "What AI Can't Teach You: Lessons in Learning, Leadership, and Surviving the AI Shift."
The framing is intentional. After a period dominated by AI capability announcements — every week a new model that surpasses human performance on some benchmark — the conversation is shifting toward what humans bring to the table. Clawson's argument, based on years of working at the intersection of AI education and enterprise deployment, is that the human skills most under-discussed are also the most durable: adaptability to novel situations, judgment under ambiguity, and the ability to lead teams through rapid technological change.
It's a more nuanced message than either the "AI will replace everyone" narrative or the "nothing will change" dismissiveness, and it comes from someone who has watched thousands of engineers learn to work with AI systems at scale.
Agents in Production: The Google Perspective
Derek Egan and Aaron Davis from Google will take the stage together to discuss how AI agents are changing the nature of work at large organizations. This isn't a demo — the focus is on how enterprises are actually deploying agents in production workflows, what breaks in practice, and what's genuinely working.
The timing is significant. After 2024's wave of agent framework announcements and 2025's scramble to turn those frameworks into enterprise products, 2026 is shaping up as the year where production agentic deployments either prove or disprove the category's promise. Google, which has been embedding agents across Workspace products, is one of the few companies with enough enterprise deployments to speak from operational data rather than pilot results.
Financial Services as the AI Proving Ground
Antoine Amend, drawing on experience from both Databricks and Barclays, will present on AI's transformation of financial services — a sector that has been both an early adopter and a cautious one, given the regulatory requirements and fraud risks that come with financial data.
Financial services is a useful lens for AI's real-world maturity. The sector has genuine AI use cases — fraud detection, risk modeling, document processing, compliance monitoring — but also meaningful constraints. AI systems in finance need to be auditable, explainable, and robust under adversarial conditions. That's a much higher bar than most AI demos are held to, which makes successful financial deployments a reasonable signal of where the technology is genuinely reliable.
The Academic Thread
One of the more distinctive aspects of the Silicon Slopes summit is its explicit inclusion of university researchers alongside industry practitioners. A panel featuring Berton Earnshaw (University of Utah) and Emily Evans (BYU), moderated by Rebecca Brockbank of Utah State, will explore how academia and industry are collaborating to develop AI talent.
Utah's university system has been a quiet source of AI talent — the University of Utah's computing program in particular has produced researchers who went on to build foundational AI systems. The panel is a recognition that the talent pipeline connecting academic research to enterprise deployment runs directly through Utah's tech corridor, and that the corridor has an interest in keeping it healthy.
What the Summit Signals About AI's Maturity
The existence of an event like the Silicon Slopes AI Summit — practical, free, locally rooted, focused on demonstrated results rather than capability claims — is itself a signal about where AI stands in 2026.
Three years ago, AI conferences were dominated by researchers presenting breakthroughs and startups pitching vision. Today, there's enough deployed AI across enough enterprise environments that the more useful conversation is about implementation patterns, failure modes, and organizational change management. The Silicon Slopes format is built for that conversation.
For builders in Utah's tech ecosystem, the summit offers something that national conferences increasingly can't: direct access to practitioners who are navigating the same mid-market and enterprise dynamics that define the regional economy. For the speakers from NVIDIA, Google, and Databricks, it's a chance to hear directly from the builders who are taking their platforms and doing unexpected things with them.
The Silicon Slopes AI Summit takes place March 31, 2026 at the Zions Bancorporation Midvale Technology Center. Registration is free at siliconslopes.com.
