Hacker News Reader

One way to do it is looking at what other experts say. I'm a long-time reader of statistician Andrew Gelman, who has a lot of good critiques of shoddy science. He responds to Guzey's critique here: https://statmodeling.stat.columbia.edu/2019/11/24/why-we-sle...

I second the recommendation of Gelman’s blog. He’s written a lot more about this particular case since then. Here’s a more recent article on the topic, for instance: https://statmodeling.stat.columbia.edu/2020/10/07/alexey-guz...

You don't know. And you never will.

Even well credentialed experts aren't strictly trustworthy, because they can often be wrong or have big gaps in knowledge. Credentials can be faked or bullshitted into. Even honest experts can fall into self corrupting games that force them to spread bad information to maintain their position in their field.

As individuals, we're going to constantly mess up and follow the wrong ideas. The best we can do is hold lightly to them and be willing to be proven wrong and always be weighing the risks of an idea being wrong. Always ask what the incentives of a person sharing an idea are. Profit incentives distort rational thinking. We all worship the thing that feeds us.

It's important in some cases to take leaps of faith so that you don't become paralyzed to inaction from the vast uncertainty of the world. Just be wary that you don't get stuck on a sinking island or take overly absurd leaps that don't have the reward to justify the risk.

Hopefully, the human colossus is marching closer to truth, even if we as individuals can never achieve it.

You shouldn't trust anybody other than yourself and your ability to recognise scientific consensus. This process is not easy and as goes for most things in life, diversification is key. Basing your research on a single article/book/source is bad, and your confidence that you understand/are knowledgeable about a topic should rise as you do more research.

You can also trust individual people and institutions if you follow them for a while and know how they do things.

That’s only slightly marginally better than randomly trusting any well respected source. In the end no single source should ever be viewed as definitive.

On the other hand you can’t spend all that much time researching every individual topic. Which is where the problem comes from, but trick is to hold options lightly.

> That’s only slightly marginally better than randomly trusting any well respected source.

Following someone's work for years is a much stronger signal than just checking their credentials imo.

It feels better, but in terms of actual evidence it’s not. One core issue is everyone makes mistakes, and nobody can research everything. So, the minimum error rate in all human content isn’t that great.

> You shouldn't trust anybody other than yourself

And probably not even that either.

Yourself is the last person you should trust. It’s much easier to detect logical flaws in other people and be critical about what they say.

> You shouldn't trust anybody other than yourself

Oh, definitely don't trust yourself. Everybody is wrong about almost everything, that includes yourself

Hide under a blanket?

it's the only way

So you’re saying trust nobody, not even yourself. Let me know how that’s working for ya

I'd probably rephrase "trust" to "maintain appropriate confidence intervals on your beliefs"

that being said, there's clearly something to trusting yourself beyond rational levels. How far would Kanye have gotten if not for boundless self belief?

> You shouldn't trust anybody other than yourself and your ability to recognise scientific consensus.

This is the anti-vaxxer motto, almost. I think I'll pass.

If you are highly intelligent, and recognize your own ignorance, then this works quite well. Anti-vaxxers get it wrong because they lack intellectual humility.

I thought we'd stop using "anti vaxx" as a slur now most of Europe and multiple world leaders have spent the last four months spreading or consuming anti-vaxx propaganda. It's not like that's a fringe position anymore.

> most of Europe and multiple world leaders have spent the last four months spreading or consuming anti-vaxx propaganda.

What? MOST of Europe? When did this happen?

I guess it depends how you define it, but France, Germany, Italy, Denmark etc all have been suspending the AZ vaccine as 'dangerous' although the actual reports of side effects are far less than the claimed danger from COVID. I admit I haven't made a formal list of every country and ticked them off to see if it's literally >50% of countries, but it's certainly >50% of Europe by GDP at this point.

That's still a far cry from "spreading propaganda".

You are correct in that there are fundamental problems with a statement as simple as the one I made. Don't take it at face value though. If you ignore your surroundings and formulate an opinion that is obviously different, I'd argue your ability to recongnise scientific consensus is severely lacking - Distilling it into something like "Don't trust anyone other than the scientific consensus" might be better.

> ou shouldn't trust anybody other than yourself

This is not good advice.

I’d say find experts that you trust and be aware that these experts may commit mistakes as any other human.

In short: find the experts that make the least mistakes.

Perhaps I should have been more precise (since you ommited a key part of my sentence) in pointing out that any person on their own is probably wrong about a lot of things. Only by taking into consideration a large enough sample of opinions can you begin to formulate your own with sufficient confidence of it being correct. Outsorcing your opinions to an 'expert' is giving up critical thinking.

Read the history of the field.

Read textbooks, oldest ones first.

Don't try to understand a field by looking at the new research.

Ignore anything you see in the popular press, newspapers, TED talks, or the internet generally.

Read opinions by experts in adjacent fields whose expertise you can more readily evaluate.

Good luck :-)

This is a good demonstration of the Lindy Effect. [0] Essentially, the longer something has been around, the more chance it has of still existing in the future.

This is exactly the same approach I take when I’m trying to learn a new field, especially one in which I have little to no background. I always begin with the texts that have survived the longest and are still recommended to this day.

[0] https://en.wikipedia.org/wiki/Lindy_effect

Can’t recommend this approach more. I’m a PhD student in cryptography and I often find that my understanding of a particular problem area is dramatically improved by reading stuff from 20-30 years ago (crypto is a young field) rather than recent publications.

While this could very well be valid in natural science fields, I’ve heard that in humanity fields as well as in something as controversial and largely unknown as neuroscience (different from traditional hard science fields such as maths and physics), the consensus gets overthrown so quickly/gets a makeover almost every decade, that old research might appear completely different from the newest consensus in the field. Of course you still have to draw a conclusion yourself in the end by synthesizing all available information, but I wouldn’t say that old information would necessarily be more trustworthy in this case. A lot of understanding in psychology/health from a few decades ago seem to be deemed nonsensical nowadays, and in many of the cases I agree with the newer conclusions.

> A lot of understanding in psychology/health from a few decades ago seem to be deemed nonsensical nowadays, and in many of the cases I agree with the newer conclusions.

This is why the Lindy heuristic mentioned by a sibling comment to yours is relevant.

If X% of the consensus in a field was overturned twenty years ago, then again ten years ago, your expectation should be that in ten years, X% of the current consensus will be thrown out as nonsense as well. So it's not necessarily that the old information was more accurate, but you can't interpret the new information well without it. What's X in your field? Maybe 5%, maybe 80%, then you should look at the field very differently.

Of course, you could be living through another annus mirabilis, where a new consensus is established that then stands for centuries. But this is by definition unlikely, so there must be strong evidence available.

The way I see it, something doesn’t need to be undeniably “True” to have utility. As humans we don’t actually need to be correct about everything we encounter, just correct enough. A working understanding can be perfectly sufficient because most of us aren’t neuroscientists who study sleep. It’s okay for us to be wrong and society will forget about this book in a short time. Someone will write a more “accurate” book about sleep that’ll get popular. But in the meantime if we start paying more attention to sleep hygiene that’s good enough for me.

It’s true that it’s hard to know for sure, but there are a bunch of things you can do. One of the most obvious is to see what other reputable sources say about the author. What’s their track record in the past? Then for any specific piece, you can look at the arguments they’re making, and then look at the sources they reference. Are those sources reputable, etc.

To put it another way, you need to have a bunch of sources you trust, and then see how strong the connections are from any new data back to those original trustworthy sources.

> how do I pick out good information when I don't know the field

Search HN threads on the topic, read people who agree and people who disagree then make your own opinion

> Search HN threads on the topic, read people who agree and people who disagree then make your own opinion

I honestly cant tell if this is parody or not.

Why would that be a parody?

I can't see how it's a bad idea to get some quick signals on how a community approaches a topic.

> unless you are lucky enough to stumble on a god source the first time.

But how will you even know that it is a “god source” as you so eloquently put it?

Just read a lot. All the time. On a wide variety of subjects. Or, resign yourself to being deceived by charlatans sometimes. No real shortcut here.

For one, don't go to pop sci books expecting the truth. You'll have a better shot with a peer-reviewed textbook. For sleep, try Principles and Practice of Sleep Medicine by Kryger.