1405
✨️ Finish him. ✨️
(mander.xyz)
A place for majestic STEMLORD peacocking, as well as memes about the realities of working in a lab.
Rules
This is a science community. We use the Dawkins definition of meme.
Because science is about objective, provable fact following a known and public method. An organization can say their findings are reproducible, but reproducibility is more than just getting the same results every time the same lab runs the same PCR on the same machine. To be truly reproducible your results need to be able to be replicated by anyone with appropriate materials and equipment.
What you are describing is research, not science. It’s not that research is bad, but that science is a philosophical adherence to a method as much as it is that method itself.
The tobacco companies conducted research when they realized smoking caused cancer and hid those findings from others. That’s not science even if their internal researchers were consistent with each other.
Seems like the only difference is that if it's public or not ie published. I think it becomes a matter of opinion then, because independent teams within the same organisation can absolutely peer review eachother, use completely different methodology to prove the same hypothesis and publish papers internally so it can be reproduced internally.
Science should be made public, but just because it's not doesn't mean it's not science. When the organisation starts making public claims they should have to back that up along the official route, but they could just as well keep their findings a secret, use that secret to improve their working formula and make bank while doing that. Not calling their internal peer reviewed studies science just seems pretentious.
No, they can’t. Peer review is not the peers you determine - it’s the peers of your community. Science that is not public is not science, because it cannot be independently verified and reproduced. It is not a small point, it’s one of the foundations of the disciplines of science.
An organisation with fully independent teams tackling the same problems can absolutely be defined as peer review. Not in the traditional sense, but reviewing, confirming and replicating nonetheless. Following the scientific method is what makes something scientific, not the act of publishing.
You can argue of the merits of those papers, an organisation can never make public statements about private research. But saying that what their doing is not science, then you're just needlessly gatekeeping.
No it literally cannot be so defined. The last part of the scientific method is “report conclusions.” That means public scrutiny free of bias. Internal groups are not public.
This is akin to saying that a corporation doesn’t need to use the courts because it has internal judges. They might have trials, but by definition they are not doing justice.
Reporting your conclusions doesn't require being public. It means the larger group of people you release it to, the less bias you'll have. Meaning in a closed organisation you have added biases of companies and marginally less people to prove you wrong, decreasing the overal quality of the conducted science. But still science, which by definition isn't black and white.
I mean, Yann LeCunn disagrees with you but sure. Go on.
He's clearly taking the "but it's better for human kind" stand, which I support with all I can. But academics can be guilty of gatekeeping and being pretentious, which I've seen by many lmao
Gatekeeping on following the scientific method is pretty good gatekeeping if you ask me. Again, what you are arguing is anathema to centuries of scientific endeavors. You’re applying your own interpretation to something that has literally hundreds of years of meaning already, in a way that is just not right. It’s not gatekeeping any more than “a court of law” gatekeeps the concept of justice.