Episode Summary:
In this episode, Maribel Lopez speaks with Dell’s Chief Technology Officer and Chief AI Officer John Roese about Dell’s enterprise AI technologies from their development to their future. Roese explains how the initial magical thinking around artificial intelligence has shifted into a more practical approach that aims to maximize the benefits of each implementation. He also discusses the emerging trends and ideas that he is seeing in the AI space.
Key Themes:
Maribel and John start by delving into enterprise AI and AI markets in general. John explains the types of AI markets and how enterprise AI differs from other applications. The conversation then moves into the challenges of AI and the steps that Dell is taking to address them.
Next, John and Maribel reflect on the near-universal reactive approach that companies took to AI two years ago and how parts of that approach backfired or fizzled out. While this technology could have been approached better, its widespread use has provided companies with a real world understanding of LLMs and their applications. Now, companies take a more practical approach to AI while continuing to innovate.
The key innovation that Roese highlights is agentic architecture. This technology differs from previous generative AI applications because it can operate autonomously and is highly specialized. Individual “agents” can have job descriptions that they are trained for much like a human being, and they can interact with each other as a human team would.
For detailed show notes, navigate the episode using the time stamps below:
[1:26] Maribel introduces the guest of the episode, John Roese. Roese is the CTO and Chief AI Officer at Dell Technologies.
[1:59] The AI market is not a singular market – there is a traditional market, a training market, and an enterprise market. The enterprise market is very pragmatic in its applications.
[4:10] Maribel asks about the challenges businesses see in enterprise adoption. Early discussions of new AI technology treated it like magic. Now that we have real world use cases and a better understanding of the technology, Dell is able to have grounded conversations about AI applications with real impacts.
[7:48] Roesch explains the challenges that Dell is facing with AI. One of the challenges was determining where to prioritize as a company. Another is the process by which you develop application ideas. Dell had this issue when bringing ideas that were not fully formed to their legal team.
[11:44] No one got AI perfectly right. Almost universally, companies reacted at the technical level before looking at business priorities. Roese encourages companies to move toward a more thoughtful approach to AI technology.
[13:36] Dell learned that its goal was to add in the minimum sufficient AI structure to address the maximum use cases. In Dell’s case, half of their use cases were related to converting proprietary data into generative outcomes. Creating one model to handle all of these cases is the most efficient approach.
[15:05] Maribel asks Roese about the trends Dell is seeing in AI. Roese points to the emergence of agentic architecture. The idea behind agentic architecture is that they are autonomously performing agents with very specialized purposes. They can be combined much like a team of human beings.
Follow John Roese: https://www.linkedin.com/in/johnroese/
Learn more about Dell’s AI solutions: https://www.dell.com/en-us/shop/scc/sc/artificial-intelligence
Follow Maribel Lopez on X/Twitter: https://x.com/maribellopez
Subscribe to Maribel Lopez on YouTube: https://www.youtube.com/c/MaribelLopezResearch
Follow Maribel Lopez on LinkedIn: https://linkedin.com/in/maribellopez/
