Hacker News Reader

Comment by justizin | Hacker News Reader

I had the interesting experience of working in a startup under a VP of Engineering who was a Nuclear Scientist and was part of the team the US sent to Chernobyl's sister site - exact same plans, basically - to ensure its' safety.

He told me that overwhelmingly what went wrong at Chernobyl is that the head of the plant consistently ignored all of the alarms and monitoring in place, because he was overconfident at what the results of his test would be. He apparently died in the control room.

Interestingly, one thing the Russians did notably different was to have a big guy, I want to think at the sister site his name was Sergei, basically live in a hut next to the cooling lake, which was uphill from the plant. In the event of a disaster, he had a humongous mallet he would use to slam open the valve, draining the lake.

... or maybe this was all made up, I couldn't tell you tbh, but it sounded legit to me. ;)

View on HN

He told me that overwhelmingly what went wrong at Chernobyl is that the head of the plant consistently ignored all of the alarms

You may consider this an excuse of the kind "but that's not nuclear power's fault" and you'd be right-ish, but I simply don't care who exactly is at fault for whole regions to become uninhabitable in case of disaster.

That is connected to nuclear power in my risk assessment, even if you can make a rational argument that it shouldn't be. And I'm very risk-averse when the magnitude of consequences is that big.

Same with Fukushima: yes, everything was safe. Unconnected to the reactor per se they made huge planning errors, underestimating a tsunami. Again, I don't care.

It may be a bias or even a fallacy that Kahnemann has already described, but it's a useful and correct fallacy, I would argue.

> for whole regions to become uninhabitable

Chernobyl has a 30km exclusion zone, and that's unnecessarily large.

> Same with Fukushima

Literally nobody died in Fukushima. It was just a memorable piece of destruction caused by a natural disaster that killed 16,000 people.

And you can mathematically prove that no possible future disaster could possibly affect a larger area?

No? Okay, my point stands. I simply don't care about probability-weighted outcomes there.

That's not how this works. We're not dealing in absolutes because this is real life. We're dealing with probabilities and expected values. If nuclear power provides a net good, we should consider using it, not blindly reject it.

Saying you "don't care about probability-weighted outcomes" means you aren't willing to objectively value nuclear power. Why not? Why are you afraid of nuclear power?

The point this article makes is that nuclear power is an incredibly safe technology, and the concerns about it are exaggerated.

It would be good if these risks were objectively valued, for example by insuring the plant against meltdown costs.

Total economic loss estimates for the Fukushima meltdown are 250-500 bn USD[1]. And if the wind had been slightly different those numbers would be a lot higher, as Tokyo real estate would be affected.

And that insurance should also cover human error, terrorism, natural disasters. Won't be cheap is my guess.

Which brings us to the biggest myth of all, that nuclear power is cheap when you don't externalise so many costs.

[1] http://www.psr.org/environment-and-health/environmental-heal...

You can't really insure against environmental damage.

What exactly do you mean? It's too expensive? Legally impossible? No insurer would insure this?

It's a thought experiment to show what replacing wishful thinking with a tiny amount of objectivity would look like. OP was talking about expectation value of damages etc. Insurers are companies that deal with these sorts of things professionally.

The economic damages of Fukushima and Chernobyl are real and gigantic. It would be foolish to ignore them.

Yeah, it'll make sense to do that when you start insuring other power plants for the environmental damage they cause.

We're not charging a carbon tax on coal plants. We're not charging drivers for the air pollution they cause in cities. If you believe the WHO when they say 7 million people die prematurely from air pollution, are you going to start making air-polluters pay for that? (http://www.who.int/mediacentre/news/releases/2014/air-pollut...) Are you insuring dams against the floods they would cause if they break?

It'd be great to do those things, but it's unreasonable to unfairly punish nuclear when you're not willing to charge other power plant technology for their expected negative externalities.

If you're going to start enacting "reasonable taxes" on negative externalities, start by getting us off of dirty energy, not by preventing us from using reusable energy.

It would be great if externalities were insured by everybody, yes. But externalities are real, why would you ignore them "out of fairness"? The sane thing to do is to consider them for all kinds of power generation, and then make the best choice. Nuclear just has huge unconsidered negative externalities, we haven't even yet talked about cost of storing and guarding waste for centuries.

Thankfully there are only a few more years until wind & solar will have solved this once and for all by being cheaper, externalities considered or not.

You can't reject one form of power [as worse than existing power] because of externalities if you aren't considering the negative externalities of the existing power. It's not out of fairness, it's out of minimizing future damage.

If you aren't accurately considering the environmental damage of oil/coal/gas, then you can't say they're superior to nuclear just because you are considering its negatives.

I agree -- I look forward to a future of clean renewables, but even still wind and solar are inconsistent. Our storage systems aren't good enough to "just use batteries."[1]

We need a dependable, renewable form of energy that can bridge the gap for 80 years until our storage is good enough.

[1] http://worrydream.com/ClimateChange/#moving-storage

I'm fascinated by the idea of a "correct fallacy." How could such a thing ever make sense? Are you saying that you use this fallacy to reach a conclusion that you believe correct for other reasons? If so, why not state those other reasons and not use your fallacious reasoning? If not, how could your fallacy be correct?

Me too. :-)

I think that this fallacy is correct in the sense that averting extreme outcomes at the cost of a lower expected outcome is rational when it comes to survival.

I suspect you might be able to explain the existence of those "irrational" behaviours this way, although I haven't searched the literature.

Kind of the opposite of playing the lottery (which I also find rational, despite all the nerds decrying it): There you are giving up expected value, in order to gain variance.

Something is not rational simply because you think it is. A decision is rational if and only if the invested resources are proportional to the likelihood of success.

A fallacy is a fallacy because it violates this proportionality. Fallacious reasoning can never be rational.

Reply on news.ycombinator.com