During last week’s SD CTO Talks, Spence Lunderman shared insights that cut through the noise surrounding artificial intelligence. The presentation offered a refreshing perspective on large language models (LLMs), moving beyond marketing buzz to explore their practical potential in business settings.
The talk began by addressing a critical challenge: the gap between technological excitement and real-world application. Through a simple demonstration, the speaker illustrated how AI can produce generic, uninspired content when not carefully guided. This highlighted a fundamental truth about these technologies: they excel at predicting probable responses but struggle to generate truly meaningful insights without strategic direction.
The most powerful takeaway was a nuanced approach to working with AI. These systems are tools to be carefully managed, not magical solutions. They can generate content and attempt tasks but require careful framing and human oversight. The key is understanding their limitations while strategically leveraging their capabilities.
A compelling case study demonstrated how complex projects can be broken down into manageable steps. Instead of asking an AI to complete an entire task at once, the recommended approach involves creating a structured workflow. This might include separate agents responsible for research, content generation, data conversion, and final execution. Each step is carefully defined, ensuring more reliable and precise outcomes.

The real-world implications became clear through specific examples. In one instance, extracting precise information from emails—such as estimated arrival times—revealed the intricate challenges of AI implementation. Simple extraction is not enough; context, accuracy, and integration with existing systems are crucial.
This pragmatic approach stands in stark contrast to the breathless proclamations often surrounding AI technology. It's not about finding a silver bullet, but about thoughtful engineering. Successful AI integration requires breaking down complex problems, defining clear parameters, and creating robust, adaptable systems.
The underlying message was empowering: AI is a powerful tool, but one that requires human intelligence to truly unlock its potential. By approaching these technologies with strategic thinking, businesses can move beyond hype to create meaningful, transformative solutions.

As the evening ended, the room buzzed with energy. We extend our sincere gratitude to University Club San Diego for co-hosting the event. Keep the momentum going by joining us on March 20th at our Dev&Data Night in collaboration with Fresh Brewed Tech. This event features a top-tier panel with leaders from Tealium, Intuit and more, diving into data innovation and real-world tech applications. Don't miss this opportunity for insightful discussions, valuable connections, and forward-thinking perspectives.