With just a few simple instructions and an overall objective, you’ve got a system that is capable of again reasoning, planning, and executing a multi-step workflow on someone else’s behalf. - Brian Schoepfle
Aunt Mary can say, I want to go to Miami for three days, I want stay in a hotel by the beach, and I want to have dinner at this place. An agent can take everything that it knows about Aunt Mary, what she likes, plus the specific instructions, book the hotel room, make those dinner reservations and provide an itinerary for Mary to approve or make changes to. - Brian Schoepfle
Voice powered, objective based computing, not instruction based computing
In the past, we had to learn the tools, then design, integrate, build, and deliver all the bits and parts to try and deliver on the description of the objective in the spec.
In an agentic world, you start with the objective.
The machine assembles the tools, skills, and data to deliver on that objective.
Agent. Think like a travel agent. Acting on your behalf, based on your instrructons, preferences, and the objectives around this particular engagement.
Now imagine you have an agent for everything you need, not just travel.
What if everyone can build their own app for the objective at hand, using the worlds most powerful computers, and unimaginable amounts of context data?
We’re here. Welcome to the future. No C prompts required. Just talk to it.
Good news,
Brian and team are bringing these capability to public sector to the benefit of all constituents.
Full Length Interview
How Google and partners are scaling agentic AI in the public sector | Google Cloud Next 2026
with Brian Schoepfle, Head of Partner Enablement & Workspace Customer Engineering, Google, Hosted by Jeff Frick
#AgenticAI #AIAgents #FutureOfWork #ArtificialIntelligence #GenAI #AIProductivity #DigitalTransformation #WorkSmarter #TechForEveryone #AIWorkflow #LLM #Innovation #AITools #TechLeadership #FutureOfTech
