L.A. Times' 'Quakebot' Erroneously Reports Massive Earthquake From 1925

At about 4:30 p.m. on Wednesday, there came news from the L.A. Times about a massive earthquake. The report said that a 6.8 temblor had just struck the Santa Barbara area. To give you a sense of scope, the 1994 Northridge Quake was of a magnitude 6.7— this one was bigger.

The event spelled great danger, but...no one felt anything. Outside, the birds kept chirping, and motorists sat idly in gridlock traffic. Everything was as usual. What happened? It turned out that "Quakebot," the Times' automated earthquake reporter (it actually gets its own byline), had relayed a bit of false information from the U.S. Geological Survey. That 6.8 quake had actually happened in 1925.

Wait, what?

Let's backtrack. Quakebot is actually an algorithm that was designed by a journalist and programmer for the L.A. Times, according to Slate. Whenever the USGS sends out a report of a quake that's above a certain size, Quakebot takes that data and plugs it into a pre-written template, and voila, we have an article about an earthquake.

Except something went wrong this time around. The Times retracted the story about the 6.8 quake almost as soon as it'd gone up. "The quake appears to have been from 1925," said the Times.

The problem seems to have started when the USGS sent out a false alert:

And here's the report itself:

The Times later explained what had happened: it turns out that a Caltech staffer, while trying to correct the location of the 1925 quake, had accidentally sent out an alert for that quake in the process. Those with a keen eye (or those who bothered to read the report) would have seen that the quake was described as happening in 2025, which would have aroused suspicions immediately. After the report was sent out, Quakebot did its job and relayed it back to us feeble humans. So that's how it all went down.

Anyway, there's no massive earthquake! We'll all having an OK day, though it may be a rough one for a certain non-sentient reporter.

LAist reached out to the USGS, but no one was immediately available for comment.