Making Data Simple

Making Data Simple treats enterprise AI as something that has to survive contact with real organizations. The show is built around conversations with people close to the work: AI researchers, security leaders, IBM executives, product builders, privacy officers, and founders. It does not stay at the slogan level for long. A discussion of agents quickly becomes a question about inferencing cost, policy, observability, and whether a company should start smaller before betting a workflow on the technology. A conversation about modern AI turns into a look at reasoning, model limits, and why impressive peaks do not erase the strange valleys in system behavior. The show has a practical center. It wants the value case, not just the demo. Recent episodes move through AGI predictions, small and efficient models, enterprise MCP, AI-native IP infrastructure, sovereign AI, security, governance, ethics, and the messy data behind training systems. The format is conversational and sometimes loose. That looseness helps. Guests can explain difficult ideas at length, and then the conversation often snaps back to a blunt clarifying question. The show is especially strong when technical ambition meets operational doubt. It asks how models are trained, how data rights are tracked, how agents should be controlled, and how governance can keep up when nation states and companies are racing. There is also an inclusive streak in the framing. Listeners are not expected to know every acronym before the episode starts. When the topic gets dense, the show usually pulls it back toward a concrete example or a business consequence. It is not anti-AI. It is not breathless either. Making Data Simple is for people who want to understand how AI, data, trust, infrastructure, and regulation are being argued over by the people trying to make them usable.