Hacker News Reader

Comment by dimitrios1 | Hacker News Reader

So what is the argument here? We shouldn't have nice things that could greatly benefit society because terrorists or war-mongering dictators might come after them?

What if a warring nation destroyed a hydroelectric dam and it flooded an entire town? Does this mean we should give up on hydro power?

What would happen if a terrorist gained access to your city's water treatment facility and contaminated it with a bioweapon resistant to chlorine? Does this mean we shouldn't have water delivery for our cities?

What would happen if a terrorist or warring nation gained access to your country's institute of virology where you are performing gain-of-function research, and they unleash a superbug you didn't intend into the population? Should we not perform medical science then?

What would happen if you built a really tall building and a terrorist decided to fly an airplane into it? Should we not build tall buildings anymore?

View on HN

The argument is that nuclear pollution scorches the earth for centuries if not longer. Any of the examples you've mentioned has - relatively speaking - as short term impact. A nuclear plant shedding a plume of radioactive elements? All you can do is set up an exclusion zone.

That doesn't mean nuclear technology doesn't have advantages. It very much has. Arguably, it's both the cleanest and most consistent way of energy production at the same time.

The trade-off is security, as mentioned by the commenter, is security. If society decides to invest in nuclear energy, it has to be willing to accept the consequences of that choice.

The technology isn't "bad" in and of itself. It's context and intentions leveraging said technology which shifts perceptions towards "good" or "bad". If the Russians would clearly recognize the challenges associated with securing a captured nuclear power plant, and take them to heart, the conversation would steer into a different direction.

None of your analogies are applicable. Nuclear power stations require extraordinary stability in a world that cannot provide it. Earthquakes, tsunamis, war, or another failure mode nobody’s even thought of yet can lead to entire states becoming uninhabitable. We should desire that the consequences of failure draw from the domain of the normal distribution. Nuclear power draws its failure consequences from the exponential distribution. And failures will always happen eventually.

Nuclear power stations on 1970s or older technology*

The newer technologies, with the much smaller reactor cores can greatly limit the destructive capabilities even to the point where a meltdown caused by some unforseen event will be entirely containable.

> desire that the consequences of failure draw from the domain of the normal distribution

Almost all interesting tech we're playing with lately does not fall in this category... and there's nothing we can do about it, that's the evolution of tech.

When interstellar-spaceships will be built, they'll practically be equivalent to "planet destroying nukes" and pretty sure accidents will happen (eg. fully scortching a planet inhabited by a few billions). AI if something ~human-level gets capable of a stretch of exponential self-improvement will be similar, but that will probably be the end of bio-humans and nothing we can do about it so no need to worry (just focus on making sure important human knowledge gets passed on to our non-human descendants).

Only way to mitigate this is to DISPERSE (establish colonies on multiple planets) and MULTIPLY (so that an accident with eg. a 99% lethality thing still leaves enough humans around to maintain a technological civilization preserving the knowledge of our ancestors). We're a bloody virus, so better embrace our nature and start to act like an effective one. There's genocide in our pasts, and there's likely exponentially more genocide in our future, we need to be resilient and accept (mass) death as part of our techno-evolution...

I think the argument is more focused around we shouldn't have nice things that could greatly benefit society because nuke plants haven't displayed a great success rate with securing their premises. The NNSA (nuke nsa) seems to be pretty strong in cyber, so I'll limit this to just physical security.

But the difference between hydro, tall buildings, etc is the critical physical security aspect nuke plants have. I agree with the point that the nuke plants seem to be capable of pretty safe operations.

But, using the 40 years of activism from the Plowshares movement[0] as one example, doesn't seem like we're capable of securing the plants from "terrorists" with the no-fail rate that I'd somewhat want for nuke physical security. As recently as '12, Oak Ridge of all places was broken into via Plowshares activists.

So if we can't track a known activist group that does this, what are our long term odds of tracking an unknown terror group with good opsec (signal, decentralized org, etc), with 100% success rate? Your comparison to virology fits here, and the two (nuke vs. dangerous medicine) seem to be treated differently in this regard, and that scares me, for one.

[0] https://en.wikipedia.org/wiki/Plowshares_movementhttps://en....

Reply on news.ycombinator.com