Google failed to warn 10M of Turkey earthquake

>"Google says the system is supposed to be supplementary and is not a replacement for national systems."

Before immediately going into the classic "but SOME people say otherwise!"

That said, I think the more basic problem I have with this article in terms of education of the reading public is the failure to touch on the false-postive vs false-negative tradeoff that is so core to tuning warning algorithms. It's easy to cherry pick with 20/20 hindsight after the fact and it feels clear when a warning that might have saved lives isn't sent. But in reality, constant warnings can be as bad as no warning because they have the same basic effect of people "not getting the message". The mechanism (people tuning warnings out, boy-who-cried-wolf syndrome) of course differs, but the end result is the same. And outcomes should be the focus when it comes to public health and security. I think this is a super important thing for reporting to acknowledge, because without at least a somewhat educated public understanding there is a systemic incentive by a lot of governments and organizations to lean into false-positives due to being better at ass covering and blame reversal. They can say "well, we warned you, and if we were wrong other times that was just being cautious, and it's YOUR fault that YOU turned off the warnings!" If Google had set off the alert for that quake, but had also set off the alert for lots of smaller quakes previously, even if the end result was the same number of deaths I suspect this article would not have been written. Or, as you say, simply not trying in the first place and keeping their virtual heads down.

Walking the fine line of balance in mass scale public safety is just very, very hard. And to achieve a systemic approach to safety that really maximizes good outcomes takes a good faith, no fault sort of environment (short of actual malice) vs finger pointing. We can see that play out in areas of life like commercial aviation which has been incredibly successful at reducing fatal accidents over the decades.

Like the internet is the merest speck on Human history and yet it is considered so ridiculously critical that people will make death threats when they have minor issues accessing it.

Median price of phone went from couple of hundred in Nokia era to a thousand or more today, justified by all the features including this one , meaning they certainly made money of providing all these value added services.

It’s no different from a fire or burglar alarm . Yes they are not substitutes for the police and fire departments .

However the alarm company cannot turn around and say you shouldn’t depend on us. The government shouldn’t depend/expect them to do anything for emergencies yes but buyers of their product can.

IF a company sold a product and advertised features for it, then it is a problem when said features don’t work especially when said features are designed around emergencies.

> After the earthquake Google's researchers changed the algorithm, and simulated the first earthquake again.

> This time, the system generated 10 million Take Action alerts to those at most risk – and a further 67 million Be Aware alerts to those living further away from the epicentre.

So, basically, Google learned from the data, improved the algorithm, and they are blamed for not having done that before the earthquake.

The headline makes it sound like Google had the good model before, and something went wrong in the notification system so nobody got the alert.

Countries and the general public shouldn't be relying on private technology firms to provide this service, nor should they be defunding the public agencies that provide it.