It's a good historical retelling, and it has some unique ideas, but they seem rather milquetoast compared to the scale of the problem. Asking journals to split

It's a good historical retelling, and it has some unique ideas, but they seem rather milquetoast compared to the scale of the problem. Asking journals to split themselves into exploratory and authoritative isn't going to work in a prestige system where everyone wants to be seen as authoritative. And there was nothing about the more concrete issues like widespread fraud. Your conclusions about what needs to be changed will depend on how deep you go in your analysis. A lot of writing about the problems in science take a surface level view and so try to come up with ideas like different archiving strategies for data, different ways to communicate. It's also possible to conclude that these are only treatments for the symptoms not causes, and so won't work. You'd just replace one set of symptoms with another. Evidence for this view can be found in the outcomes of prior attempts to reform science in the modern era. For example a frequent criticism of academic work is that they don't make their data available, thus papers have to be taken at face value as a claim by authority rather than the original scientific ideal of Nullius In Verba that the article mentions. Distributing data is easy but academics complained it is too hard for them, so some journals compromised and said OK, you just have to sign a document saying that you'll send your data to anyone who asks. The result was predictable. Eventually someone tested how often scientists would send data when asked like they'd agreed to in writing, and discovered it was pathetically low, like 5% or lower. There had been the appearance of reform without reform actually taking place. This happens because the incentives are bad. Scientists earn more money and gain more power when they are accepted as authorities. Being accepted as such means never showing doubt, never admitting to mistakes, never admitting to uncertainty. All the things scientists are meant to do freely, in practice, they are incentivized not to do. And there are no penalties anywhere for manipulating people through pseudo-scientific claims. Lots to win, nothing to lose = lots of fraud and collapsing trust, except of course amongst that subset of the population that wishes to live by blindly following authority (perhaps because they themselves are a part of the prestige credentials game). Science is structurally creating division and polarization. The fix for this is as brutal as it is simple: eliminate all charitable or taxpayer funding for science. All science should be corporate. This moves science out of the prestige economy into the market economy, which has many mechanisms and evolved cultures designed to tackle bad incentives. For example, it's hard to directly monetize scientific claims in a market economy, you're usually monetizing results (minor exception: self help books, ted talks, etc). You can try to create fake results but then there are many laws, law enforcement bodies and even journalists trying to catch you. There are truth in advertising laws in some places, your employer can be sued for fraud under many statutes and there's a culture of not blindly trusting companies so those mechanisms actually work. See the huge outcome gap between fraud at Theranos and the endless frauds at universities.