> The good thing about science is that it doesn’t depend on trusting any one person or institution, it depends on a process designed to catch bias and error.
In the long term, yes. In the short term it is extremely hard to know what is good science, and what is noise produced by the system that incentivizes publishing and not actual science.
> Scientists don’t just publish opinions. Well, they can, but we generally call these people crackpots. However, in modern times they do financially well on YouTube and podcasts. Scientists test ideas through predictions and experiments, share their data and methods, and other scientists try to reproduce the results. If findings can’t be repeated, they’re usually rejected. Over time, only the most reliable results hold up.
I think we do not have to go further than Covid times to see that it is absolutely not the case. Mass hysteria of the pandemic absolutely ostracized anyone who dared to suggest that (a) covid maybe is coming for a Wuhan lab, or (2) that lockdowns maybe not the best way to handle things especially if we consider impact on children. So... The evidence suggests that scientists are people like everyone else, make the same mistakes, may use their status to push their own biases and their own agendas, and pursue their own interests.
> Yes, funding and politics can influence science, on all sides. But the scientific system has: peer review, conflict-of-interest disclosures, reproducibility, and open data. These are not perfect, but they make science far more reliable than all other known systems.
All these things work only if people act with honor. The moment the incentives are misaligned nothing will work. Peer review is absolutely gamed. I do not know if you remember, but someone took his own life because of this -> https://news.ycombinator.com/item?id=27296433
> What I believe is a continuation of how we built our modern civilization, since the time of Newton, and earlier. I cannot personally audit all of science, so instead I rely on the scientific method, which is the best system (warts and all) that we have yet found to discover the base truth.
Totally agree. Me too (as a scientist myself). However, I have absolutely no belief in the system itself. Today, I only believe those who I think do good science, either because I read their papers and I see that they are absolutely honest and upfront about issues with their work and ideas, or because I know from personal experience that these people will never falsify data just to put the paper out. My advisor is like this: we never publish a paper until we know that all the data tracks, and all the environment, scripts, data, etc are archived to ensure that if needed we can track the issue. However, some advisors are not like that, and I know it because students talk between themselves, and you just hear those stories, and you know.
There is no incentive today for a scientist to be honest (unless they grew up this way).