{"version":"1.0","type":"rich","provider_name":"Acast","provider_url":"https://acast.com","height":250,"width":700,"html":"<iframe src=\"https://embed.acast.com/$/6a108d0055b99c7f8915c7a7/6a4beb035fd88ad0059ef4e6?\" frameBorder=\"0\" width=\"700\" height=\"250\"></iframe>","title":"The Human in the Loop Isn't Watching","thumbnail_width":200,"thumbnail_height":200,"thumbnail_url":"https://open-images.acast.com/shows/6a108d0055b99c7f8915c7a7/1783360201119-048e5464-624d-4456-9e63-354327e7d6da.jpeg?height=200","description":"<p>Somebody in your building is approving four hundred agent decisions an hour, and every governance deck in the company is calling that a \"human in the loop.\" It isn't. This week, the practitioner's read on the most reassuring phrase in enterprise AI — where it came from, what it actually costs to do it right, and why the reviewer you're counting on is, in the exact moments that matter, statistically indistinguishable from a rubber stamp.</p><p>This episode of Above the Noise — the unbiased AI brief for enterprise leaders:</p><p><br></p><ul><li><strong>News Brief:</strong> The Digital Omnibus is now law — the Aug 2 deadline moved, but not the way vendors will pitch it · a new U.S. Executive Order made \"voluntary\" the frame, in the same three-week window Europe made oversight slower · Gartner sized the vendor panic at $234B — and what that means for the wave of agent bolt-ons headed at your stack.</li><li><strong>Deep Dive:</strong> Where \"human in the loop\" actually came from (Fitts, radiology, aviation), the four preconditions nobody wrote down — time, expertise, authority, exposure — and why the enterprise version threw away every one of them. Five structural failure modes and the four-lever honest reboot.</li><li><strong>Expose a Lie:</strong> \"We have a human in the loop.\" Three receipts — Amazon's Kiro / Cost Explorer 13-hour outage, the CSA/Token Security 65% incident rate, and Deloitte's 53-point governance gap — and the one question that turns the phrase honest in your next review.</li><li><strong>The question to sit with:</strong> If the agent had recommended the opposite outcome on every item last month, how many would your reviewer have approved differently?</li></ul><p>An honest owning-up on the Omnibus prediction from Episode 3 (I called it a rumor with good PR; six days later it became law), then why the underlying read still holds.</p><p>No vendors. No hype. Just the signal.</p><p>👉 <strong>Follow Above the Noise</strong> wherever you listen — new episode every two weeks. Send it to one leader who keeps saying \"we have a human in the loop,\" and leave a rating so more people find it. Between episodes, find Shaun on LinkedIn — come argue with him there.</p>","author_name":"Shaun Gehring"}