First Stop AI - AI Academy October 14th Webinar Recap

Video recording: A NYSTIA team member is currently reviewing the AI-generated captions to ensure accuracy. Stay tuned!

During last week’s Wednesday webinar, Matt Clement, Strategic Advisor & EVP at KHM, kicked off the NYSTIA AI Academy series with “First Stop AI,” a human-first guide to getting started with AI. 

The session showed how DMOs and small tourism businesses can move from individual AI experiments to a shared, well-governed way of working. From choosing a workspace and setting the rules to sketching a first pilot, Matt provided a confident path forward with AI from day one. 

AI Is the Next Step, Not the End

• Today’s AI fears can be compared to past worries about mechanized agriculture and the Industrial Revolution. AI represents the next evolutionary step in how people work, rather than an endpoint.
• For DMOs and small businesses with limited resources, AI as a way to level the playing field and expand capacity rather than replace people.

Agents and Workspaces

• An AI tool has two parts: the model (the engine) and the harness (the platform you use, such as Claude or ChatGPT).
• An agent is an AI platform pointed at a folder containing written instructions that tell it how to behave.
• Your workspace should live on a shared drive like Google Drive or OneDrive, not inside an AI platform. This lets one agent work across platforms and devices, and keeps work from leaving when staff do.

Rules, Roles, and Data

• Most DMOs and small businesses don’t need enterprise accounts.
• Classify information as green (approved public), yellow (internal, ask the owner first), or red (personal or visitor data, kept out of the agent’s reach).
• Assign a sponsor, an administrator, workflow owners, and an independent reviewer for anything going outside the organization.
• Plan for security, billing oversight, connectors to tools like Gmail and Calendar, and the cost of training.

Starting Your First Pilot

• Start with the smallest useful workspace and a job your team already does.
• Follow this workflow: define the outcome, share your context and content, ask AI to identify gaps, review its plan, draft, review, and save.
• Update the agent’s instructions whenever it repeats a mistake.

Key Takeaways

• Choose a workspace and rules before choosing a tool.
• Keep one shared operating system so everyone’s AI follows the same rules.
• Always keep a human in the loop to review outputs.

Thank you to Matt Clement for presenting, as well as everyone who attended! Join us next week for “Guardrails Before Guidebooks” with Janette Roush from Brand USA, at 11:00 a.m.