Miscalculation Invalidates LHC Safety Assurances

KentuckyFC writes "In a truly frightening study, physicists at the University of Oxford have identified a massive miscalculation that makes the LHC safety assurances more or less invalid (abstract). The focus of their work is not the safety of particle accelerators per se but the chances of any particular scientific argument being wrong. 'If the probability estimate given by an argument is dwarfed by the chance that the argument itself is flawed, then the estimate is suspect,' say the team. That has serious implications for the LHC, which some people worry could generate black holes that will swallow the planet. Nobody at CERN has put a figure on the chances of the LHC destroying the planet. One study simply said: 'there is no risk of any significance whatsoever from such black holes.' The danger is that this thinking could be entirely flawed, but what are the chances of this? The Oxford team say that roughly one in a thousand scientific papers have to be withdrawn because of errors but generously suppose that in particle physics, the rate is one in 10,000."

Comments

Are they good for anything?

[QuantumG]

Maybe I just like Romulans, but when I hear that the LHC will be making black holes I don't think about "woo, the earth is gunna get swallowed!" I wonder if there are any cool ways to use them for power generation.

Re:Are they good for anything?

[TheMeuge]

There is.

Matter being drawn into the black holes should be accelerated to damn close to the speed of light, and will emit massive amounts of gamma radiation, with a conversion rate that's higher than even fusion.

If we could harness the energy of the gamma emissions around artificial black holes, we'd be have vast energy generating capability, without the pesky fast neutrons that most fusion reactions generate.

My first thought from reading this

[Reapman]

My first thought from reading the summary is that essentially we're at a point in technology or whatever that we could, POSSIBLY, destroy the planet in a literal sense. That's a scary thought, especially if you think what we'll be capable in a hundred years from now.

I STILL don't think the LHC will kill us all but the fact we're debating it says something.

Re:My first thought from reading this

[gluefish]

SF Idea: Finding that ALL the black holes spawned by civilizations that were eaten up by them after experimenting with supercolliders ...that black holes are nature's cure to civilization

"That isn't right"

[Jeremiah Cornelius]

"It isn't even wrong..."

What if they are so far off, that not only do they not produce black holes, they do nothing, but dim the lights in Switzerland?

Re:Red Title?

[morgan_greywolf]

Editors: Would somebody PLEASE create a FAQ on this? A red title thread has been in several articles every day.

Answer: A red title is what appears on articles subscribers see in "The Mysterious Future!" previews. For some reason, as an article is taken out of "The Mysterious Future!", the flag that makes the article a subscriber-only preview seems to come off some period of time ahead of the flag that makes the title red, so what you are seeing is what subscribers see when the article is in subscriber preview mode.

Either they did this on purpose to indicate that the article is 'hot off the presses' or there's some sort of race condition in their new styling code.

Flawed, invalid, wrong, confused, or just nonsense

[DrVomact]

If the probability estimate given by an argument is dwarfed by the chance that the argument itself is flawed, then the estimate is suspect.

The headline says "Miscalculation Invalidates LHC Safety Assurances", yet the quote from the abstract seems to say that because arguments are sometimes "flawed" (terribly squishy word, that), it follows that for crucially important calculations we have to...well, the abstract doesn't say what we should do, and there's no link to the actual article. (Maybe there's a good reason for the latter.)

This amounts to the assertion that if an estimate is about something very important, then we can't trust the estimate, because some estimates are mistaken. In other words, we can't make estimates about important things—just trivial ones.

Unless someone produces the article in question, and unless it actually makes a more substantial argument than I quoted, I vote this a waste of my time on the part of whoever submitted it. May the rats eat your mail.

Of course

[Nerull]

The the safety of the LHC does not depend on a single calculation.

For a black hole created by the LHC to destroy the earth essentially requires everything we know about physics to be wrong.

First, can it even create them? The Standard Model says no - not even close. A certain category of String Theory models say maybe. This same models predict that these black holes are everywhere, being created all the time, even here on Earth.

Will black holes evaporate? They certainly should. If we are wrong about this than in all probability we are wrong about being able to create them at all as well - and we should hope we are, since they'd have swallowed up the universe by now if they were dangerous.

Is a stable micro black hole even dangerous? The numbers I've seen show a black hole like this would behave more or less like a neutrino. Maybe hitting an atom every few thousand or million years. The sun will enter its red giant stage, destroy Earth, and shrink down to a white dwarf before the black hole gains any significant mass. I don't think we will care much at that point.

Re:Clarifications

[kwikrick]

Okay, I've read the paper. I do not agree with it.

You claim that when a probability estimate for some event is very small, much smaller than the chance of the estimate being flawed, then we should consider the actual probability of the event to be larger.

In the papers terminology:

P(X) = P(X|A)P(A) + P(X|not A)P(not A)

Where given are:

- P(X|A) the probability estimate based on some argument and

- P(A) the chance that the argument is flawed

You then argue that if P(X|A) is very small, then P(X|not A), an arbitrary number, is probably much larger, and therefore P(X|not A) P(not A) may be significant compared to P(X|A)P(A). Thus, you argue, P(X) is probably higher than P(X|A).

This is where you go wrong.

You basically insert an arbitrary number, P(X|not A), based on an inexpert opinion, in the equation. Sure, the maths add up, you get a higher value for probability of the event, but your new probability estimate is no longer based on expert knowledge. You use some small number (1/1000) for P(X|not A) in your examples, but this makes no sense. Why not assume 50/50 if you simply don't know how likely it is that some event will happen? The main component of your new probability estimate is now only the statistical probability of an expert being wrong, which is independent of the probability of some event taking place.

In other words: you simply replace the estimate of an expert with the estimate of a lay person, multiplied by the change that the expert is wrong. However, even if the expert is wrong, that does not mean the lay person is right. The number you end up with is junk, it is meaningless. It should definitely not be used for risk assessment.

This is very dangerous use of probability theory. The argument in your paper is easy to follow, but it is false, and seeing that is not so simple. I sure hope no policy maker will read your paper and base important decisions on it.