Hacker News new | ask | show | jobs
by anon1253 3439 days ago
Well yeah, but rigor is part-of the scientific mindset. One that we now try to enforce with statistics, peer-review, and reproducibility. But those are rather new ideas. Proper based evidence medicine, for example, is something we only came up with in the 60s or so. Rigorous experimental science might produce better insights in some way, because the data is better. But there is still value in other methods. An interesting example here is the advent in genetics for personalized medicine: did it deliver on the promises? No, it's still utter hype and vaporware. Part of the reason, I believe, is that it was merely stamp collecting. Sequence enough, go fishing for correlations, and hope that something useful comes of it. The current hype around CRISPR and related technology? Not because of rigorous scientific testing on large peer-reviewed data sets. Someone, was "merely" curios about bacterial immunological defense mechanisms. Curious enough to try and understand it. It was the explanation that made it worth it. Ideally they go hand in hand, but anyone who has ever been in a scientific institute will agree that this "idealized" view is rarely, if ever, true. An interesting paper about the subject is "you can't play 20 questions with nature and win" by Alan Newell

http://repository.cmu.edu/cgi/viewcontent.cgi?article=3032&c...

1 comments

A little off-topic, but part of the problem with personalized medicine I think is that the target market for any drug developed for a particular personalization is necessarily small. In other words, right now it is a bad business decision. Perhaps it can be used as a delta- to a mass-market drug in the future, like seasonings to a basic recipe, but that's a bit off.
Perhaps, but not sure if it's the correct explanation for https://en.wikipedia.org/wiki/Eroom's_law

I think the focus on data-driven pipelines for discovery, rather than focus on fundamental understanding of the biology and chemistry of organisms might also play a role. But I have no "proof" to back that up, other than my anecdotal experience as a researcher in a genetics institute :p