What about misuse?

[sdack]

I have a new question. Are all results archived or does the Pande Group destroy all bad data/findings?

Maybe I should ask if Folding@Home could be misused for creating biological weapons and I am sure the answer here to will be yes. I am also certain that all basic research can be used for creating biological weapons but still need to be made public in order to find cures. Therefore I am wondering if anything is being done to avoid misuse, if some careful filtering is in place to destroy potentially dangerous findings and if only the good and useful results are being kept and archived.

[John Naylor]

That's a good point...

However, I don't think any results are stored once papers have passed per review... the amount of storage required would be so enormous as to be unfeasible, so I think (and this is only speculation) that results would be deleted once papers are published.

[shatteredsilicon]

Considering that writing papers and papers getting peer reviewed can take months, that implies that the amount of data stored has to be enormous anyway, so the point seems rather moot.

I think the simple fact is that any science can be used for good and bad. It's the same argument as nuclear power plant vs. nuclear bomb. And let's not forget that a vast amount of knowledge that modern medicine is based on came from highly unethical research. I'm not going to elaborate for fear of falling foul of Goodwin's law, but you can probably make your own conclusions on what I'm referring to.

[anandhanju]

Wait a minute... Results deleted after a paper? I hope not. What if the same data can be used for a related but different analysis? I'd be a waste having to wade through 2 years and a few hundred thou WUs of 1487s to look up something related to DPPC_DOPC_CHOL again if it is already being captured in the current simulation/ result set.

Storage@home, anyone?

[shatteredsilicon]

Storage@home, anyone?

Hmm... I'm thinking encrypted distributed hash table network, a-la Freenet/Entropy.
Nigh-on impossible to ensure the data doesn't fall out of the network, though...

[Ivoshiee]

The project data storing has come up before and I recall as (almost?) all the FAH generated data is available for third parties (universities, researchers, ....) to tinker with. I recall as some major data centers in US are getting the FAH data for archival.

[Ivoshiee]

Storage@home is not here, yet. But if you read public documents then it is designed to help running of the FAH DCS. I doubt that it will be used for storing historical data of the FAH project.

[John Naylor]

It'd be a waste having to wade through 2 years and a few hundred thou WUs of 1487s to look up something related to DPPC_DOPC_CHOL again if it is already being captured in the current simulation/ result set.

*shudder* one core of my main comp keeps getting those God I hate them. You make a good point

[sdack]

The project data storing has come up before and I recall as (almost?) all the FAH generated data is available for third parties (universities, researchers, ....) to tinker with. I recall as some major data centers in US are getting the FAH data for archival.

This shall not turn into a "The Great US Evil"-thread*), but I believe that many nations continue doing research in biological warfare even when they have signed agreements not to build any such weapons. May be that they even agreed on not doing any research but I would not bet on it. Just like agreeing on not doing any live nuclear tests does not stop them from simulating such tests with super computers. So basically everyone who is folding may contribute unwillingly to such research, including the Pande Group itself, without any control unless some is put into place?

*) I bet it would be an amusing thread

[Ivoshiee]

I have a new question. Are all results archived or does the Pande Group destroy all bad data/findings?

Maybe I should ask if Folding@Home could be misused for creating biological weapons and I am sure the answer here to will be yes. I am also certain that all basic research can be used for creating biological weapons but still need to be made public in order to find cures. Therefore I am wondering if anything is being done to avoid misuse, if some careful filtering is in place to destroy potentially dangerous findings and if only the good and useful results are being kept and archived.

There is no obvious criteria to say this piece of data is "bad" and that one is "good".

[7im]

I suppose the simulation research could be used to simulate just about anything biological, but the disease research is very targeted. I very much doubt the Alzheimer's treatment could be corrupted to actually cause it.

But then you have a higher chance of getting hit by an asteroid than to get Alzheimers from a corrupted DC project. Have you installed your asteroid proof roofing yet? No?! Then you don't need to worry about the fah data getting corrupted either.

There are bigger things you should be worrying about, and let Stanford worry about how their data might be used.

[sdack]

There are bigger things you should be worrying about, and let Stanford worry about how their data might be used.

As long as Stanford uses hundred of thousands of computers around the world that are not their own does this make it a serious question. You cannot answer it by telling people to shut up and keep folding, 7im.

[Ivoshiee]

There are bigger things you should be worrying about, and let Stanford worry about how their data might be used.

As long as Stanford uses hundred of thousands of computers around the world that are not their own does this make it a serious question. You cannot answer it by telling people to shut up and keep folding, 7im.

You either trust PG or not - that simple it is. As much I've seen they does not have any projects started or published containing phrases like "biological warfare" or "that project will help us kill more people".
PG results are peer reviewed, PG data can be accessed by other interested parties etc. Do you want them to destroy all data just after the scientific paper has been published and no 3rd party access to their data?
I do not understand what you are after or what do you want to hear from us or PG.

[anandhanju]

I guess what sdack is trying to ask here is if FAH has any policy to mask information in results that they feel can be used for destructive purposes before releasing it to the public domain. Given that FAH is doing basic research at understanding how proteins work and possibly find a way to stop them for doing the bad things they do, there may be inferred information on how to cause those bad things to occur if one wished/ needed to.

I suspect the answer is no as shatteredsilicon opines and appreciate the thought provoking topics that sdack brings up, much to some people's chargin!

[sdack]

I will try to give an example. You choose to examine Alzheimer related proteins. This in itself is already a means of controlling misuse of the computational power. One could as well draw a random name out of a hat and decide to examine aimlessly any protein. This is not the case. Now your research comes to a point where you have several options to continue with and one option includes an early, dangerous finding. Naturally all researcher will want to explore all options and theoretically leaves no choice if they want to succeed. Let us say that this early, dangerous finding shows how the Alzheimer disease can be triggered in a human 500 times more often than it occurs naturally. What do you do?

a) You make the finding public because it allows you to infect mice, rats and monkeys with Alzheimer more easily, and greatly help in lab research.
b) Because you think it could be misused you do not make it public. Instead you just keep the data to possibly continue research into even the dangerous areas because your goal is to cure the disease.
c) You destroy the finding and continue with your other options making it your very last resort. Chances are your methods have improved and you want to repeat your research in any way, or perhaps just to verify on the danger involved.