Hawaii official who sent false missile alert has been fired | Hacker News Reader

Hawaii official who sent false missile alert has been fired

theguardian.com

30 pointsdboreham34 comments

They got rid of the one employee least likely to make a mistake again.

It was hardly their fault either. "This is not a drill" was part of the phone message they received. Why the hell would you say that when it is a drill?

The whole thing was a fiasco from the top to the bottom and shows they haven't even remotely approached operations from the very simple premise of "Humans make mistakes". They didn't even have a runbook for handling the scenario of a message being accidentally sent out.

While I agree with you in principle, there's more to it than that.

There was another report that claimed that the employee suddenly refused to cooperate with any investigation to figure out what happened. It's likely their lawyers told them to stop talking, and that we don't know the full story here.

My take on "there's more to it" is that the employee had allegedly already caused two false alarms through his own actions.

The fact that he made a written statement then refused any other cooperation suggests to me that, rather than shirking his duty by this act, perhaps he knows better than to dig himself a hole by trying to be "helpful".

> It's likely their lawyers told them to stop talking, and that we don't know the full story here

Any defense attorney will give him this advice. My guess is the investigation into this is not looking to discover the truth, it's looking for someone to blame. And the lower they are on the totem pole, the better.

If the investigation is an impartial arbiter only concerned with truth and justice, then the guy should cooperate. In the real world, it's probably a kangaroo court that will forget all exculpatory statements and amplify all damaging ones. So if he tells the investigators:

"Yeah, I clicked on it, but the message told me it was not a drill, and my manager never advised me there was an ongoing exercise"

Then the prosecutor will tell the court: "During questioning he confessed to clicking the button. It's an open-and-shut case, send him to jail!"

It states in the article the employee made the mistake several times previously.

https://en.wikipedia.org/wiki/5_Whys

No it doesn't. Besides, that mostly means (we've all seen the interface) that that mistake was made before ... and nobody saw any reason to fix that UI.

Mistakes have a cause. Not fixing it will cause repeats.

Actually, we haven't seen the interface.

What we've seen is a general mockup of the interface, because they wouldn't release an image of it publicly.

This isn't an argument with your overall point. But your second sentence isn't quite right: "This is not a drill" is part of the alert message itself. If it is a drill, the instruction to issue the alert is prefaced with "EXERCISE EXERCISE EXERCISE." This is from a thing I read yesterday that I can't find now, so I can't back it up I'm afraid. But you see that it isn't quite as you think.

(The UX of the system we all saw was truly terrible, but this employee had a track record of making similar mistakes.)

Yes. Or similar. In Fire / EMS, we do training exercises and for anything involving detailed communications any "block" of radio interactions is both started and finished with "This is a training exercise". It's not done with every communication because that would be onerous but each block (typically managed by Dispatch, rather than field personnel).

This avoids War of the Worlds style issues too, as our radio traffic is not encrypted and available on scanners.

That's exactly what I thought, but then the recent FCC report (quoted at https://arstechnica.com/tech-policy/2018/01/hawaii-ema-fires... with the source scanned and available at https://cdn.arstechnica.net/wp-content/uploads/2018/01/repor...) says:

>Employee 1 has been a source of concern for the same SWP staff for over 10 years. Employee 1’s poor performance has been counseled and documented and the SWP members have stated that they are “not comfortable with Employee 1 as a supervisor, two-man team, or as a part of the SWP in general. He does not take initiative and has to be directed before he takes action. He is unable to comprehend the situation at hand and has confused real life events and drills on at least two separate occasions.”

Was that sentence written before or after the event ?

Answer: after.

So I give it at least 50% that this is scapegoating.

They should blame the [1] software and issue a redesign

[1] https://hackernoon.com/redesigning-hawaiis-emergy-alert-inte...

It has since came out that the alert was not sent in error due to bad design (in addition, that image has since been turned out to not be real), but the alert was sent on purpose due to the operator believing that an attack was imminent (due to a confusing drill message).

https://www.washingtonpost.com/news/the-switch/wp/2018/01/30...

>Then a Facebook and Twitter message was put up on the emergency management agency’s accounts. It was not until 8.24am that the corrective message was retweeted by the Hawaii governor, David Ige.

>The FCC report notes drily: “The governor has stated that he was unable to do this earlier because he did not know his Twitter password.”

Moral of the story: Memorize your Twitter passwords if you're going to take office. Or just use something easy, like '1234' or 'password'.

Having an simple / easy-to-remember password on your Twitter account is a bad idea if you're a state governor. Or really, anyone.

The right answer here is, use a complex password (that you likely won't remember), and/or 2FA, and use a password safe that's available on your mobile device.

My suggestion of a '1234' password wasn't serious. Although they should at least memorize their passwords and not offload that onto assistants or whatever, although what you suggested is better.

Previous discussions

https://hn.algolia.com/?query=hawaii&sort=byDate&prefix&page...

Little that they know that Hawaii itself is ours early warning system.

So the "bad UI" thing was fake news, it seems.

I really hate the term "fake news". It implies that someone created the whole story on purpose to deceive people. I think the "bad UI" thing was just someone trying to figure out what really happened and noticed that the webpage for the alerts was atrocious. They incorrectly thought that this could have been the reason for the mistake. By no means was it fake, rather it was simply a mistake. There's nothing malicious about that.

The screenshot was not a real screenshot. As a result, all news about "bad UI" could be considered to some degree fake, since it was based on fake screenshots.

After looking into it some more, it appears as though that screenshot was an attempt to simulate what the user may have seen because the agency did not release any sort of official images. Anecdotal evidence led everyone to speculate on what the "bad UI" may have looked like.

I still wouldn't call it "fake news", just a half-assed attempt to try to mock-up what the operator saw with little information. Its bad journalism (and should be called out as such!) but its not malicious. And that's why I hate the term "fake news" so much: it carries a negative overtone but it can applied to ineptitude and/or stupidity. Seeing as ineptitude and stupidity are rampant in life, assigning malice to them could be used as a rhetorical technique by a bad actor. For example, a politician wanting to bring down news groups that don't support him/her could say that all news is "fake news" and offer up examples such as the "bad ui".

I think it is worse than that because the screen shot originated with the Hawaii state government, but it turns out to be not genuine, AND they took no steps to debunk the resulting widely circulated stories that bad UI had been the root cause.

http://www.ajc.com/news/national/photo-hawaii-missile-warnin...

>I really hate the term "fake news". It implies that someone created the whole story on purpose to deceive people

No it just means someone published the story without proper fact checking.

Back in 2016 it meant fabricated news, usually about politics. Donald Trump started abusing the term not long after it was revealed that the fabricated news at the time was mostly successful in targeting his followers, presumably to dilute the term.

See https://m.youtube.com/watch?v=XoLHF3Z9eMg or Google "Jestin Coler"

I stand corrected.

Maybe, maybe not. For some reason the guardian thought it'd be okay to go to publication without interviewing the officer who actually clicked the button, so they only reported the agencies version of events. Feels like they're just making him a scapegoat, and the guardian is helping.

Either way, you shouldn't say "this is not a drill," when, you know, it's a drill.

That officer refused to talk to reporters (which is his right), but also refused to be interviewed by the Federal Communications Commission as well, as part of their investigation. He was fired for that reason.

Arguably, if you don't include the words "this is not a drill" in your exercise then your exercise is not sufficiently realistic. The real question is: why include those words in a real announcement?

Root cause analysis is not mandated to finding one flaw and fix it. It’s mandated to find at least one flaw.

The problem identified earlier still needs to be addressed because the same outcome can arise from two causes.

I can't help but feel betrayed clicking the link "guy ... has been fired" and then immediately reading "guy has resigned". That's not exactly the same thing.

Not the same person.

> The head of Hawaii’s emergency management agency has resigned and a state employee who sent out an false alarm of an imminent missile attack has been fired

Okay, sorry for the noise, then. I'll try to write my comments after I read the article next time.

Reply on news.ycombinator.com