The reason stems from my thinking that all of the following are important and true (and too often overlooked) regarding science:

1) A lot of science is real AND valuable to society.

2) A lot of “science” is actually fake – see, for instance, a decent percentage of papers in psychology 15 years ago.

3) “Science” (as an approach to knowledge discovery) is one of humanity’s greatest inventions – but in practice, it is reasonably often misapplied, or the process is distorted due to bad incentives or poor training. Unfortunately, not all fields of science have done a good job of being self-correcting either, so sometimes, fields go in bad directions for quite a while and need reform. There are different kinds of bad science:

(i) Sometimes, science is “bad” because it uses unsound methods for figuring out the truth (such as when p-hacking is rampant).

(ii) Sometimes it is “bad” because it overclaims (e.g., “Importance Hacking” where scientists claim they found something important/valuable when they didn’t actually demonstrate what they claim in their study. Or cases where science is used to “prove” questions that can’t be proven by science – such as which policy is better in a particular context when it’s actually a tradeoff between different values).

(iii) Other times science is bad because it is biased (e.g., when people are only willing to run or publish studies that show X but not that show the opposite of X).

(iv) And sometimes science is bad because it’s simply fraudulent.

4) Promoting broad “trust the science” is misguided (and actually harmful) because a bunch of science is fake. If you tell people to always just “trust the science,” then you are going to cause them to be tricked by a bunch of bad science, or you are going to contribute to their disillusionment and loss of trust when they discover (correctly) that some of the science you’re saying is good is actually garbage.

5) The “distrust all science” view is probably an even worse take than “trust the science.” If you distrust all science, you are likely to miss out on incredible things (such as highly effective treatments), and you set yourself up to fall for tons of things that don’t work (e.g., widely used unscientific treatments). Those who tell people to always just “trust the science” sometimes accidentally push people into the “distrust all science” view when those people realize that some of what they are being told to trust is crap.

6) So, hard as it is, rather than promoting either “trust all science” or “distrust all science,” the course of action I believe in with regard to science education is to teach people that “Science” (as a method) is an incredibly powerful and useful invention, but that “science” (as actually practiced) is much like every other field: some of it is good, some of it is crap. There are good hairdressers and bad hairdressers, and there is good science and bad science (and unfortunately, some bad science ends up in the very top journals – while journals and peer review absolutely do block some bad science, they unfortunately still let through quite a lot of it).

Since some science is well done, and some of it is poorly done, it’s very valuable to learn to tell the difference to make the best use of scientific results – both with regard to applying it in your own life and using it to form your beliefs about the world.

If we pretend science is all good or all bad, we do a lot of harm. We need nuance to see through the bad stuff while maintaining the tremendous benefits.

This piece was first written on November 20, 2024, and first appeared on my website on January 14, 2025.

(1) Two groups have a small (but meaningful) difference in their average value of some trait, with heavily overlapping distributions.

(2) Some people (“Dichotomizers” or “Oversimplifiers”) observe this difference (in their everyday life or media reports) and turn this small average difference into a (sometimes very harmful) oversimplification: “A’s are like this, B’s are like that.” They also fairly often make it seem like this difference is large (or applies to almost everyone in the group), important, and fundamental (e.g., inherent and unchangeable).

(3) Other people (“Difference Deniers”), often acting with good intentions, criticize this oversimplification, which they correctly perceive as harmful. But instead of saying some combination of:

• “the difference in averages is small”
• “the distributions are heavily overlapping”
• “judging an individual based on a small difference in group averages is a poor way to make predictions, as well as unjust”
• “we shouldn’t judge people for differing on that trait” (if it’s not a trait one should be judged on)
• “if we want to remove the difference in averages, we should consider implementing policy XYZ”

they say “the difference in averages does not exist.” After denying the difference and seeing those they respect deny it, some of them become convinced anyone who believes in the existence of this (actually existing) small average difference is nefarious (and lump such people in with those who harmfully oversimplify people into “A’s are like this, B’s are like that.”) Others among them know the average difference exists but pretend not to because they want to fit into the group that adamantly denies the difference, or because they feel guilty about believing it (even though they are right about it existing).

(4) Oversimiplifiers from (2), who remain totally convinced an average difference exists (and are correct about its existence but exaggerate its magnitude), assume that the Difference Deniers from (3) must be either stupid (for not realizing there is a difference), or untrustworthy liars (for denying what they must see is true), or cruel lunatics (for getting enraged at people for believing in “the truth”).

Queue endless fights between the Dichotomizers and the Difference Deniers, both of which are misrepresenting the actual reality of the situation and demonizing each other.

The problem with turning small averages into “A’s are like this, B’s are like that” is that it is an inaccurate oversimplification and often unfair to A’s or B’s or both.

The problem with denying the existence of average differences that, while small, really do exist is that you end up believing falsehoods, or you end up lying, or both, and you may end up unfairly misjudging people who are (without malice) reporting on real average differences.

To avoid the weaknesses of both the Dichotomizers and the Difference Deniers, I think the best way to handle these cases is to:

1) Avoid pre-judging people based on their membership in broad groups – learn about people as individuals before coming to judgments about them.

2) Avoid language like “A’s are like this, B’s are like that” so that you aren’t a Dichotomizer.

3) Avoid denying that an average difference exists when it really does exist, so that way, you aren’t a Difference Denier.

4) When relevant, remind people that small average differences are not a good basis for judging individuals (epistemically and morally), and point out that the distributions between the two groups are heavily overlapping (when they are) to combat people using differences in the average as a justification for stereotyping.

5) Point to (when relevant, helpful, and accurate) policies that may help close the gap between the two groups (keeping in mind that some gaps in averages are fine if the trait in question is merely a difference and not something “good” or “bad”)

6) Point out (when the difference in question is not something people should be judged for) that this attribute should not be a basis for judging people, i.e., that having different values of that trait is completely okay.

Another approach that can be taken when the group differences in the average are small but meaningful is well described by Guy Srinivasan in the comments on an earlier draft of this post: “Can we agree to make decisions as if there were no average difference, since usually all such decisions would turn out the same, and usually when they wouldn’t it’s perpetuating systemic problems to make the decision differently?”

Of course, as with any binary categories, some people will only be partial Difference Deniers or Dichotomizers – people are absurdly complex, and this model I present here is purposely simplified in order to help communicate this dynamic clearly.

Okay, but are there cases where the Dichotomizers or Difference Deniers are actually just right?

Absolutely, there are some.

When a group difference is SO huge that the distributions are nearly non-overlapping, then it’s reasonable to say, “A’s are like this, and B’s are like that.” For instance, it makes sense to say that “blue whales are big, mice are small.” In such cases, the Dichotomizers aren’t really oversimplifying. But when we’re talking about human groups, this kind of situation is very rare.

And in situations when the difference in averages between groups is so small as to be essentially insignificant for all purposes, the Difference Deniers aren’t actually denying reality. For instance, if it turns out that right-handed people are 0.001% better at school than left-handed people, that difference is so small as to not be meaningfully different from zero for all purposes, and so saying there is “no difference” is an extremely reasonable thing to do. There are, in fact, many attributes along which human groups differ so little that “no difference” is an accurate way to describe it (even though the difference is not literally zero to the 10th decimal point).

Note (March 26, 2023): I first wrote this list on December 3, 2022. Since then, GPT-4 has come out, and several of the points in this list are closer to happening. For example, point #2 is partly true already, thanks to Bing Chat (which runs on GPT-4).

Here are nine ways I think that AIs that generate text (like GPT-3) will have a >50% chance of changing the world for the better and worse in the next ten years:

#1: The internet will get flooded with AI-written articles, and you often won’t know if you’re reading something written by a human.

#2: Search engines will generate answers to your questions on the fly (from scratch) instead of just showing a list of websites to you and instead of using pre-extracted answers. Google will have to adapt, or it may finally lose its dominance.

#3: Cheating on school essays will become rampant, as AIs will be able to get students good grades in many classes (at negligible cost), and it will be very hard to detect such cheating since each essay will be unique.

#4: You will be able to train an AI on samples of your own writing, give it a new essay title and a bulleted list of points you want to make in the essay, and it will write a pretty high-quality essay covering all the points you listed, in a style that matches your own writing.

#5: Spam messages (and text-based phishing attacks) will become unique. Rather than sending the same message to each person, spam will be unique for each recipient. And it may even have its style adapted to what is known about each recipient (e.g., demographics).

#6: Propaganda on social media will start to become automated. Rather than bad actors having hundreds of people on their payroll to promote a viewpoint, they’ll replace them with larger swarms of human-seeming bots that each act uniquely.

#7: AIs will be fine-tuned on our own personal email corpus, and then (much of the time) you’ll be able to start with an automatically generated first draft of email replies rather than having to write emails from scratch or receiving mere sentence-level suggestions.

#8: The text of ads will get automatically edited/rewritten by AI to be fine-tuned to different audiences to help maximize clicks.

#9: AIs will start being used in education as digital private tutors to explain concepts to students, re-explain things and simplify explanations when a student is confused, point out mistakes made by students, etc.

We’re entering a wild time when it comes to AI. Its effects on our lives will be felt more and more, to say the least.

If you read this line, please do us a favor and click here to answer one quick question.

Increasingly, people in the U.S. are concluding that society is really F’d up. But the reasons are subtle, hidden, and complex, so each group picks their favorite simple enemy to blame. Problems go unfixed.

Points in favor of the claim that people are increasingly coming to this realization:

1. This helps explain Trump’s election in 2016. Many rural Republicans and disaffected independents believed society was going down the tubes. They didn’t think that electing the same people (who would use the same old solutions) would fix it.

2. COVID-19 exposed malfunctions of various authorities (the FDA, Surgeon General, CDC, etc.) that are pretty hard to ignore. Liberals blame Trump for these failures, which may be fair, but another possibility is that these American institutions have been rotting for a lot longer. I don’t know enough to say with confidence.

3. Costs for health care, education, housing, infrastructure projects, and energy ($ per kWh) have (as far as I can tell?) been rising for a long time, with debates raging over the (surely complex) causes. People feel the sting of these rising costs. Meanwhile, many argue that wages have stagnated (which has been vigorously debated). What’s more, inequality seems to have risen, though this, too, is debated.

4. An appalling number of people in the U.S. end up in jails and prisons. Shockingly, we have one of the highest incarceration rates in the world. This is obvious to the communities that are most impacted. Though still awful, the numbers have fortunately been improving somewhat in recent years [H/T Alyssa Vance].

5. Student debt growth is out of control, and it’s hard for most students to ignore this.

6. People don’t have adequate savings, the saving rate seems to be relatively low, and a disturbing number say they don’t have enough cash to deal with emergencies. “About four out of 10 Americans said they had enough in savings to cover a surprise $500 expense. Another 21 percent said they would rely on a credit card, while 20 percent said they’d cut back on other expenses. Another 11 percent said they’d turn to family or friends for the money” (source). That being said, the wording of the survey that generated those results was rather confusing, so the results should be interpreted with caution. Note that “a 2016 study [found] that 76% of households had $400 in liquid assets, after…monthly expenses” (source).

7. There is a huge racial wealth gap (even more so than an income gap, I believe), contributing significantly to poorer quality of life for Black Americans.

8. Eric Weinstein’s “the Portal” podcast, which seeks to give explanations about what the hell has gone wrong in America, rose extremely rapidly to popularity. Its theories are quite interesting, in my opinion. It is nuanced (and hard to summarize), but it blames a significant portion of the problem on broken institutions that were predicated on rapid economic/productivity growth that we didn’t end up having, the collective behaviors of the Baby Boomer generation, misleading justifications given to us about immigration, a broken scientific establishment, and a propped up system of harmful deceptions.

So what’s really to blame for the U.S.’s problems?

Trump blames immigration, China, and the left.

The left blames Trump, Wall Street, the right, and (sometimes) capitalism/wealthy people/big tech.

Yang points to automation and globalization.

The BLM movement points to racism and police violence.

The right blames the left, declining moral values, and political correctness.

Libertarians (and the right, too) blame regulation and bloated government.

What seems to me to be the case: many of the explanations by different groups have some truth in them. But they are each only pointing at a part of a much greater problem without really getting to the core of it.

I think that many of the biggest issues in society are simply not being addressed, and their root causes go largely undiscussed. We are spoon-fed simple, misleading explanations for very complex things.

To solve complex problems, we have to know what’s true. To know what’s true, we need to resist political bias, wishful thinking, scapegoating, and overly simplistic (yet appealing) explanations.

To fix society, we must unearth the (hard to spot) root causes and cooperate in order to implement reasonable solutions.

What’s at stake here is way bigger than left, right, democrat, republican, or libertarian. What’s at stake here is the flourishing or decay of our society.

Is society actually getting worse, though? I think the perception is that it is. Is it actually worse than 50 years ago? I really doubt that. Is it worse than 15 years ago? It’s hard to say, and depends on what you care about, but it seems to be worse on at least some metrics, like: the cost of healthcare, the cost of education, student debt, inequality, political polarization, and so on. It’s also improved on some metrics, like GDP per capita. Any way you slice it, society has serious problems. Whether they’ve been getting worse on net is harder to say.

If you read this line, please do us a favor and click here to answer one quick question.

This piece was first written on September 1, 2020, and first appeared on this site on March 19, 2023.