If funding scarcity combined with stack-ranking of grant proposals leads to funding fewer groundbreaking ideas, we should expect lottery-based methods (where if you meet a minimum quality bar you get put into the funding lottery) to do better than stack-ranking methods, right? I know there's been some recent use of funding lotteries, but has there been any research yet on how groundbreaking the resulting science is?
Nicely written, better than many published and peer reviewed papers I've seen (though haven't checked provenance of data and graphs of course) - however it nowhere mentions replication? Can you provide higher resolution images of graphs pls?
The biggest game changer that promises to redress the imbalance between old ideas and new ideas, at least in biopharma research, is the Cloud Lab Revolution.
Cloud Labs, of which there is only one that is fully functional at the current moment (www.emeraldcloudlab.com), will give young researchers the ability to write grant proposals to do novel research that are ten times more cost effective than they’ve been in the past. This is because there will be no need for grant(s) to finance the capital expense of a young researcher setting up a fully equipped lab. The prior funding of fully equipped labs as a result of a string of previous grants is a huge competitive advantage for old scientists doing incremental research vs. young scientists doing novel research. Strings of prior grants also allow old scientists to finance the build up of a pipeline of grad student and post doc lab slaves who do all the wet lab grunt work in legacy PI plantations, mostly pursuing themes and variations on old scientists’ stale ideas. Young scientists using a Cloud Lab won’t need lab slaves. This means that the over-production of Ph.D.’s for whom no future faculty positions exist will moderate.
Cloud Labs will allow universities to poach top scientific talent without having to pay the capital expense of building them a replacement lab to lure them away from stale old institutions. This should break the hegemony of the incumbent grant-hogging MIT/Harvard/Stanford cabal, further stimulating novel research. Carnegie Mellon University (CMU) figured this out, which is why they are building the world’s first university-wide cloud lab. Startups are figuring this out too.
As usual, the old guard and the NIH are ten steps behind. But just wait until the new generation of super-scientists bred in cloud labs start layering machine learning and AI on top of their experimental protocols, a process that has already begun at CMU. (Oh, and did I mention push-button reproducibility?). Fun stuff.
This reminds me of the saying that science advances one funeral at a time. More seriously, the 2000 American Academy of Arts and Sciences report, ARISE also observed that major US funding agencies, NIH and NSF in particular, were conservative in their funding choices.
If funding scarcity combined with stack-ranking of grant proposals leads to funding fewer groundbreaking ideas, we should expect lottery-based methods (where if you meet a minimum quality bar you get put into the funding lottery) to do better than stack-ranking methods, right? I know there's been some recent use of funding lotteries, but has there been any research yet on how groundbreaking the resulting science is?
I agree. Haven't seen anything on lotteries yet though. But I suspect people are working on it as we write.
Nicely written, better than many published and peer reviewed papers I've seen (though haven't checked provenance of data and graphs of course) - however it nowhere mentions replication? Can you provide higher resolution images of graphs pls?
The biggest game changer that promises to redress the imbalance between old ideas and new ideas, at least in biopharma research, is the Cloud Lab Revolution.
https://20visioneers15.com/blog/f/the-cloud-lab-revolution
Cloud Labs, of which there is only one that is fully functional at the current moment (www.emeraldcloudlab.com), will give young researchers the ability to write grant proposals to do novel research that are ten times more cost effective than they’ve been in the past. This is because there will be no need for grant(s) to finance the capital expense of a young researcher setting up a fully equipped lab. The prior funding of fully equipped labs as a result of a string of previous grants is a huge competitive advantage for old scientists doing incremental research vs. young scientists doing novel research. Strings of prior grants also allow old scientists to finance the build up of a pipeline of grad student and post doc lab slaves who do all the wet lab grunt work in legacy PI plantations, mostly pursuing themes and variations on old scientists’ stale ideas. Young scientists using a Cloud Lab won’t need lab slaves. This means that the over-production of Ph.D.’s for whom no future faculty positions exist will moderate.
Cloud Labs will allow universities to poach top scientific talent without having to pay the capital expense of building them a replacement lab to lure them away from stale old institutions. This should break the hegemony of the incumbent grant-hogging MIT/Harvard/Stanford cabal, further stimulating novel research. Carnegie Mellon University (CMU) figured this out, which is why they are building the world’s first university-wide cloud lab. Startups are figuring this out too.
https://makepossible.cmu.edu/cmu-cloud-lab/
As usual, the old guard and the NIH are ten steps behind. But just wait until the new generation of super-scientists bred in cloud labs start layering machine learning and AI on top of their experimental protocols, a process that has already begun at CMU. (Oh, and did I mention push-button reproducibility?). Fun stuff.
This reminds me of the saying that science advances one funeral at a time. More seriously, the 2000 American Academy of Arts and Sciences report, ARISE also observed that major US funding agencies, NIH and NSF in particular, were conservative in their funding choices.