Twenty-eight seconds is a small interval to put in a headline. It becomes misleading when detached from the place and event it measures. In its December 7, 2023 bulletin, timed 14:03:54, Mexico’s CIRES reported an earthquake detected initially by 40 SASMEX sensor stations and an alert for Mexico City. The bulletin gave that city a preliminary warning interval of 28 seconds. It was a record of a particular detection-and-alert sequence, not a promise to every resident or a prediction made before the earthquake began.
The National Seismological Service at UNAM explicitly separates those jobs: it says earthquakes cannot be predicted and that the SSN does not operate the alert. Detection can still create useful time because a transmitted warning may arrive before damaging waves. Conflating the institutions makes it harder for the public to know who measured an event and who issued a warning.
The available interval depends on where a rupture starts, how waves travel, and where a recipient stands. For some people it may be enough to stop a dangerous task or move away from a falling object; for others it may be very short or absent. A sensor network cannot promise a fixed number of seconds to every address. Nor should the public interpret a missing alert as evidence that no earthquake occurred. Alert thresholds and location matter. The system must decide when a signal merits broad interruption, and that decision creates both value and limits.
Mexico’s experience also shows why technical accuracy is part of public trust. If news copy says the system “predicted” a quake, it encourages users to expect a capability nobody has demonstrated. When a later tremor arrives without the imagined advance warning, confidence in a functioning detection system can erode. The opposite mistake is to dismiss an alert because it did not arrive minutes ahead. It was never a calendar forecast. It was an attempt to turn sensor data into a message faster than ground motion could travel to some places.
A useful technology story therefore follows the complete chain: seismic stations, decision rules, transmission, audible alert, and practiced human response. It also asks whether the signal reaches people with hearing disabilities, workers in noisy spaces, and those without connected devices. The December 2023 bulletin records a particular alert, not universal accessibility. The UNAM warning against prediction should stay beside any explanation of the system. The achievement is neither prophecy nor total safety. It is a Mexican public-warning infrastructure that can make a small interval usable, provided institutions explain the interval honestly and communities can act on it. Schools and workplaces need practice before a siren sounds, because a warning that must first be explained is already spending the time it was meant to create.