<p><a href="https://www.linkedin.com/in/mattmcelvogue/"><u>Matt McElvogue</u></a>, VP of Design at <a href="https://teague.com/"><u>Teague</u></a>, doesn&#39;t talk about design as a polish layer. He talks about it as a mission-critical failure point. His clearest example: a $5,000 tactical device used by JTAC operators for calling in 9-line bombing runs, where zeroizing the device was buried so deep in the menu that soldiers in the field resorted to shooting it or blowing it up. That failure isn&#39;t a UX anecdote; it&#39;s the operational cost of ignoring the experience layer. </p><p>Matt lays out how AI and autonomous systems are now forcing a fundamental rethink of that layer, specifically the shift from in-the-loop to on-the-loop decision-making, and the three trust requirements every autonomous system interface must satisfy: legibility, predictability, and recoverability. He also describes a coming design challenge that has no precedent, which is building interface components for AI systems that will dynamically assemble the UX themselves, in real time, based on individual context. </p><p><strong>Topics discussed:</strong></p><ul><li><p>How AI and autonomy are shifting military operators from in-the-loop to on-the-loop as threat volumes scale </p></li><li><p>The three trust requirements for autonomous system UI design: legibility, predictability, and recoverability</p></li><li><p>Working with early-stage defense companies before contracts arrive and how early design involvement shapes technical requirements </p></li><li><p>AI systems that dynamically assemble their own UX, requiring designers to build components for experiences they can’t fully predict</p></li><li><p>How procurement decision-makers who grew up with iPhones are raising the bar for defense technology usability</p></li><li><p>Why trust erosion from poor interface design is effectively irreversible so the military ends up with expensive equipment operators work around </p></li></ul>