Comment by angry_octet
4 years ago
Research without ethics is research without value.
Unbelievable that this could have passed ethics review, so I'd bet it was never reviewed. Big black eye for University of Minnesota. Imagine if you are another doctoral student is CS/EE and this tool has ruined your ability to participate in Linux.
> Research without ethics is research without value.
didn't we learn a lot from nazi/japanese experiments from ww2?
From my understanding - no, actually. We learnt a bit, on the very extreme scale of things, but most of the "experiments" were not conducted in any kind of way that would yield usable data.
Yes and no. It’s my understanding that the Germans pioneered the field of implanted medical prostheses (like titanium pins to stabilize broken bones). A lot of that research was done on prisoners, and they were even kind enough to extend the benefits of the medical treatments that they developed to prisoners of war (no sarcasm intended).
We did. Often we wish they could have got more decimal points in a measurement, or had known how to check for some factor. Despite all the gains and potential breakthroughs lost nobody is willing to repeat them or anything like them. I know just enough medically people given 2 weeks to live who were still around 10 years latter that I can't think of any situation where I'd make an exception.
Though what a lot is is also open to question. Much of what we learned isn't that useful to real world problems. However some has been important.
It doesn't really sound like that's the case: https://en.wikipedia.org/wiki/Nazi_human_experimentation#Aft...
are there more impactful benefits that aren't listed on the wikipedia page? It sounds like the main research contribution is philosophical discussions over whether or not it would be okay to use the data if someone had a good reason for it.
1 reply →
Learn how to torture? Maybe. Learn real knowledge? No. Most of those info are not just sick but also impractical.
The goal of military is to protect or conquer. The goal of science is to find the truth, and the goal of the engineering is to offer solutions. Any of the true leaders in either fields knows there're more efficient means/systems to get those goals, even in ww2 era.
Experiments producing lots of data doesn't necessarily mean they were useful. If the experiment was run improperly the data is untrustworthy, and if the experiment was designed to achieve things that aren't useful they may not have controlled for the right variables.
And ultimately, we know what their priorities were and what kind of worldview they were operating under, so the odds are bad that any given experiment they ran would have been rigorous enough to produce results that could be reproduced in other studies and applied elsewhere. I'm not personally aware of any major breakthroughs that would have been impossible without the "aid" of eugenicist war criminals, though it's possible there's some major example I'm missing.
We certainly did bring over lots of German scientists to work on nukes and rockets, so your question is not entirely off-base - but I suspect almost everyone involved in those choices would argue that rocketry research isn't unethical.
By in large no. The Nazi experiments were based on faulty race science and were indistinguishable from brutal torture and what remains is either useless or impossible to reproduce for ethical reasons.
I'm a total neophyte when it comes to the Linux kernel development process, but couldn't they just, y'know, use a Gmail address or something? Couldn't the original researchers have done the same?
Yes, they could. This is actually addressed in the original email thread:
> But they can't then use that type of "hiding" to get away with claiming it was done for a University research project as that's even more unethical than what they are doing now.
I was also thinking that commits from e-mails ending in ".edu" are probably more likely to be assumed to be good-faith; they are from real students/professors/researchers at real universities using their real identities. There's probably going to be way more scrutiny on a commits from some random gmail address.
1 reply →
Some CS labs at UMN take ethics very seriously. Their UXR lab for example.
Other CS labs at UMN, well... apparently not so much.
Ethics are highly subjective on the margins. In this case they completely missed this issue. However the opposite is more often the case.
A good example is challenge testing Covid vaccines. This was widely deemed to be unethical despite large numbers of volunteers. Perhaps a million lives could have been saved if we had vaccines a few months sooner.
Research without ethics (as currently practiced) can have value.
I can't agree that widespread challenge testing would have been ethical. It's a larger topic than HN can accommodate, but some factors I consider important: (1) NPIs are effective at reducing transmission, (2) the consequences of an outcome with side effects could include global and long-lived anti-vax sentiment -- COVID19 is unlikely to be our last pandemic.
Issue (2) arose with the EU response to rare AZ/J+J side effects, where I believe the EU is more deserving of criticism. They will undoubtedly cause more deaths in their own populations and throughout the world than would occur from clotting complications, but no one will hold them to account. But they weighed their equities as more important than global benefit.
To save some folks an acronym lookup: NPI stands for Non-Pharmaceutical Intervention, and refers to things like wearing a mask, washing hands, physical isolation, etc.
As you agree challenge testing is unethical and it clearly would have had value (saving lots of lives) are you conceding that unethical research can have value?
1 reply →
Life support machinery was developed with methods like cutting dog heads, plugging them in and see how long it shows signs of life.
If only we could have taught dogs to review kernel patch, ... we would probably be all out of work
Individuals should still be able to contribute, just not under the name University of Minnesota.