In this episode of the Utilizing AI Podcast, I joined Stephen Foskett and Dave Graham to dive deep into the current state of Agentic AI. We moved past the marketing hype to discuss what it looks like to deploy autonomous agents like OpenClaw today. While the promise of personal assistants that can manage our data and automate complex tasks is exciting, we explore why these “science experiments” aren’t quite ready for prime-time enterprise use, and the security risks you need to consider before handing over your API keys.
We shifted the discussion to the societal impact of digital actors, debating the future of ‘digital twins’ versus invisible assistants, from ‘universal basic compute’ to AI solving healthcare and government hurdles. If you’re curious about industry trends from model training to real-world inference, then make sure to give this episode a listen.
AI is more than just a Copilot license
Key Discussion Points:
- The “Normie” Gap: Why general users are being told to deploy agentic AI before the industry has solved for security and user experience.
- Security & “Roll Your Own” Agents: Our take on the risks of no-code installations and the “Amazon billing problem” of runaway token costs.
- Societal Shift: How agents might move from being hacky tools to becoming a seamless part of our social and professional fabric.
- The Future of Inference: Why the focus is shifting from “who has the best model” to “who has the best tool-calling and agentic experience.”