The Technology Perspective

Emergency vehicles outdoors at the Presidio — ambulance and fire apparatus, provided in-kind by a Bay Area department as a community outreach activity. ATEC and EMERGENZ staff conduct structured walkthroughs for groups of 8–10, explaining the operational environment that AI tools will be deployed into: the ambulance interior, the communication equipment, the documentation workflow, the conditions under which decisions get made. This is the most direct exposure the technology community can get to the operational reality — before the first session starts.

Delivered by an ATEC-connected East Coast practitioner — the continuity link between both events. Opens with what the East Coast found: the specific operational realities, concerns, and unanswered questions that emerged from public safety practitioners across all disciplines. Moves through a structured tour of what working as a first responder actually looks like — decisions made in seconds, documentation happening in moving vehicles, supervision happening through radio — and what that means for any AI system intended for deployment in this environment. Closes by framing the day's central question: what would responsible deployment actually require?

A panel of operational voices — multiple disciplines, multiple roles — speaking directly to the technology community about what would actually be useful, what constraints make most demos non-deployable in their current form, and what the field would pay for if it existed and was validated. Moderated to keep conversation constructive and forward-looking rather than a critique session. Intended to give builders, researchers, and product leaders direct access to a practitioner brief they cannot get from market research.

Second round of apparatus walkthroughs for those who didn't catch the pre-session. Informal engagement between practitioners and technologists encouraged — this is where unscripted conversations happen.

The technical and governance panel. Opens with where AI safety research currently sits and where public safety intersects — and notably where it doesn't. Examines what responsible deployment would require from an AI safety research and governance perspective: pre-deployment validation protocols, automation bias mitigation, adversarial robustness for voice synthesis and spoofed calls, shadow use governance frameworks. Intended to give practitioners access to the research community and give researchers access to a deployment context they have not fully engaged with.

Three parallel tracks. Mixed composition by design — each group ideally includes both practitioners and technologists, but the format adapts to whatever ratio the room presents on the day. Facilitators adjust framing based on actual composition: if a group is heavily technical, the session becomes a practitioner-briefed design challenge; if it is heavily practitioner, it becomes a structured gap and requirements documentation session. No minimum headcount per breakout. All formats produce the same output: 5–7 documented findings or design constraints.

Rapporteurs from each breakout present findings — approximately 12 minutes each. The moderator synthesizes across West Coast presentations and holds them against the East Coast Summary of Findings, identifying confirmations, contradictions, and new material specific to the technology community perspective. No live voting or intensive audience participation is required. Attendees contribute via a structured web form (QR code distributed at session open and emailed post-event) with 48-hour collection window.

The moderator presents a draft framework of the West Coast's key themes alongside the East Coast summary. The draft is shared with all registered attendees via email post-event alongside the 48-hour web form. Final West Coast Summary and cross-series synthesis document are published within 5–7 days under EMERGENZ's name as educational and research documents — not advocacy. The series synthesis gives the AI safety research and policy community a structured, practitioner-grounded account of where the field stands and what responsible deployment would require.