Why Reasoning Agents Change Everything in Defense | In The Arena

Reasoning agents have quietly crossed a threshold. In just the past few months, models gained the ability to choose and chain tools, turning them from text summarizers into systems that can actually reason, act, and adapt. For defense missions, that shift is enormous.Vannevar CEO Brett Granberg explains why the o1 –> o3 leap changed everything, how a task that once required 40 analysts a month now runs in 20 minutes, and why agents are more than “chatbots with tools.”We cover:Why tool use is the real step-changeConcrete defense use casesThe playbook for building with agents: model-agnostic, tool-first, and mission-drivenHow tech-enabled services could upend billions in prime contractsThe unsolved problems: model evaluation, UX, and hallucinationsWhat an “agent-native” team looks like, and how to build oneOutline00:00 The inflection point00:36 What are reasoning agents?03:38 Why now?04:46 New missions unlocked05:45 Vannevar case study09:17 The last 20% problem12:18 Ship fast, don’t break infra16:18 The agent product playbook20:13 Agents x sensing grid23:23 Be model-agnostic or lose26:19 3x Better, not 10% better28:24 Impact on tech-enabled services32:18 Paths forward for startups34:05 What’s real today36:10 Picking models that win38:22 Unsolved challenges39:15 Hiring an agent-native team