this reminds me of a distinction I made a while ago between the "entrepreneurial" mindset (risk-taking, willing to distort reality within bounds) vs the "scientific" mindset (truth-oriented, cautious, skeptical). but this post made me realize that what I call the "entrepreneurial" mindset can be useful in science too!
Amazing. Yes! Bad projects. Good projects. Science. Pseudo science. Success. Failure. Lies. Fraud... all of that you wrote ...and I could not NOT think of Rupert Sheldrake's theories on "Resonance" - on the "Field" ... Once an idea or a practice is out there... it will be used....one way or the other!
thank you for this fascinating piece - again!
Sage
PS By the way what did you think of the movie Oppenheimer... ? I want to see it and I dont want to see it.... 🤔
are we (collectively) maximizing the upsides (let's be crazy / pursue what seems impossible-endeavors) or optimizing the downsides (let's play it safe / be cautious)? it always feels like this pendulum is switching from one side to the other, creating tension / mishaps like in between.
Interesting post! I agree with you and with some prodding may even endorse the spicy take. If you haven't already read this essay, I will point to https://www.experimental-history.com/p/science-is-a-strong-link-problem by Adam Mastroianni, which very persuasively argues that in science the only thing that matters is getting the best stuff and we should have minimal to no gatekeeping and increase variance. Bad ideas, whether fraudulent or simply incorrect, will eventually be exposed, and the upside to having truly transformative discoveries is worth the wasted time/effort.
Yes, Adam's posts on science are really great! The whole LK-99 saga vindicates his arguments, too: no pre-publication peer review (it was published on arXiv), but lots of *post*-publication peer review, for a really broad definition of "peer"... I really do hope that it doesn't end up to be a hoax, since that would have the potential to convince lots of people that we need more carefulness and gatekeeping in science.
Agree on all fronts. Also glad to see you write something, even if tangentially, about LK-99. I have been looking around for essays about it here and haven't really found much.
This was great! Reminds me a tad of the how much should you lie post I wrote a while back, though that was on tech companies and not science. Well done!
This is just a special case of the evergreen temptation of "lying for a good cause". Sure, it's not actually *true*, one says, but maybe it can convince someone else about my great ideology?
I think you already know the answer to that in your heart. Deceiving someone to push them towards a given direction, when you have no idea if that direction is the right one, is ultimately self-destructive. If you knew whether it actually was the right direction, you wouldn't need to lie.
About "the optimal amount of fraud is not zero", that's a misapplied maxim. The point of that maxim is that other people will inevitably attempt to fraud you, and that your response shouldn't be to apply maximum defensiveness against that, burning all of your own resources in the process. It's not a maxim that tries to argue that fraud is actually good.
I mostly agree with what you're saying. I think my post can be considered an exploration of an edge case: sometimes we *really* don't know what direction is the right one, and as a result we may not be moving fast enough because of hesitation. Or we think we know the right direction, but are mistaken. In those cases *something* needs to come and shake us a bit so that we start out in a new direction, and that something can involve deception (and it has, in some historical cases). If you're familiar with optimization terminology, it's about escaping local maxima.
Also, I *am* voluntarily applying "the optimal amount of X is not zero" to a broader domain than its usual application! I think it's a phrase that holds a lot of interesting truth in many interesting ways.
this reminds me of a distinction I made a while ago between the "entrepreneurial" mindset (risk-taking, willing to distort reality within bounds) vs the "scientific" mindset (truth-oriented, cautious, skeptical). but this post made me realize that what I call the "entrepreneurial" mindset can be useful in science too!
Hmm it seems likely that these two mindsets converge in people we consider "geniuses" 🤔
Amazing. Yes! Bad projects. Good projects. Science. Pseudo science. Success. Failure. Lies. Fraud... all of that you wrote ...and I could not NOT think of Rupert Sheldrake's theories on "Resonance" - on the "Field" ... Once an idea or a practice is out there... it will be used....one way or the other!
thank you for this fascinating piece - again!
Sage
PS By the way what did you think of the movie Oppenheimer... ? I want to see it and I dont want to see it.... 🤔
I liked Oppenheimer! Perhaps more political than I might have expected, but I like politics-themed movies.
are we (collectively) maximizing the upsides (let's be crazy / pursue what seems impossible-endeavors) or optimizing the downsides (let's play it safe / be cautious)? it always feels like this pendulum is switching from one side to the other, creating tension / mishaps like in between.
great post!
Interesting post! I agree with you and with some prodding may even endorse the spicy take. If you haven't already read this essay, I will point to https://www.experimental-history.com/p/science-is-a-strong-link-problem by Adam Mastroianni, which very persuasively argues that in science the only thing that matters is getting the best stuff and we should have minimal to no gatekeeping and increase variance. Bad ideas, whether fraudulent or simply incorrect, will eventually be exposed, and the upside to having truly transformative discoveries is worth the wasted time/effort.
Yes, Adam's posts on science are really great! The whole LK-99 saga vindicates his arguments, too: no pre-publication peer review (it was published on arXiv), but lots of *post*-publication peer review, for a really broad definition of "peer"... I really do hope that it doesn't end up to be a hoax, since that would have the potential to convince lots of people that we need more carefulness and gatekeeping in science.
Agree on all fronts. Also glad to see you write something, even if tangentially, about LK-99. I have been looking around for essays about it here and haven't really found much.
Yeah, I'm glad I got to write about it too—it's been a pretty messy event to follow. It seems like the hype has already begun to slow down, though.
This was great! Reminds me a tad of the how much should you lie post I wrote a while back, though that was on tech companies and not science. Well done!
Thanks Rohit! Mind posting a link to that post? Sounds vaguely familiar but I couldn't find it.
Sure! Here you go - https://www.strangeloopcanon.com/p/how-much-should-you-lie
This is just a special case of the evergreen temptation of "lying for a good cause". Sure, it's not actually *true*, one says, but maybe it can convince someone else about my great ideology?
I think you already know the answer to that in your heart. Deceiving someone to push them towards a given direction, when you have no idea if that direction is the right one, is ultimately self-destructive. If you knew whether it actually was the right direction, you wouldn't need to lie.
About "the optimal amount of fraud is not zero", that's a misapplied maxim. The point of that maxim is that other people will inevitably attempt to fraud you, and that your response shouldn't be to apply maximum defensiveness against that, burning all of your own resources in the process. It's not a maxim that tries to argue that fraud is actually good.
I mostly agree with what you're saying. I think my post can be considered an exploration of an edge case: sometimes we *really* don't know what direction is the right one, and as a result we may not be moving fast enough because of hesitation. Or we think we know the right direction, but are mistaken. In those cases *something* needs to come and shake us a bit so that we start out in a new direction, and that something can involve deception (and it has, in some historical cases). If you're familiar with optimization terminology, it's about escaping local maxima.
Also, I *am* voluntarily applying "the optimal amount of X is not zero" to a broader domain than its usual application! I think it's a phrase that holds a lot of interesting truth in many interesting ways.