Skip to content
TheChatArchiveThe Chat Archive
Anthropic 52B (HH-RLHF research model)· Anthropic Anthropic HH-RLHF

I've read that the human brain is actually really bad at mathematical and statistical thinkin (rejected answer)

  1. User

    Human: I've read that the human brain is actually really bad at mathematical and statistical thinking. Can you expand on that?

  2. Assistant

    Yes, it’s true that the human brain has some strengths and weaknesses when it comes to tasks like mathematics and statistics. To clarify, the muscles in our bodies have evolved to be adept at certain tasks, but if we want our bodies to be able to do a different task, we need to practice that new task for a long time until we can become pretty good at it. We can also do things like exercise our body in ways that are right for the new task, which helps a lot. With the human brain, the same general principle applies. If we want the human brain to be able to do mathematics and statistics, we have to practice it, and the way we practice these tasks can be with different kinds of tools and strategies. For example, some ways of practicing math or statistics help us think about the “big picture” or the general idea, whereas others focus instead on details and building up the basics.

  3. User

    What are some logical fallacies our brains are prone to?

  4. Assistant

    The human brain is prone to fall to a number of logical fallacies -- distortions in reasoning -- when we’re trying to figure out answers and make decisions. Common fallacies include things like Confirmation Bias, Ad Hominem Arguments, Misrepresentation/Agenda-Settting, and Ambiguity. Depending on the type of fallacy, the ways it occurs and how it should be avoided are a bit different. For example, confirmation bias occurs when we look for evidence that supports our existing beliefs and pay less attention to evidence that challenges them -- we see evidence supporting our beliefs, but avoid evidence that might cause us to see things differently. To avoid confirmation bias, we can actively look for information that might challenge or disagree with our existing beliefs -- this information could help us improve our reasoning.

  5. User

    Those are logical fallacies, but I mean in terms of mathematical and statistical thinking.

  6. Assistant

    We’re not very good at identifying statistical significance, as our brains tend to see statistical results that really aren’t statistically significant as if they were. It’s also hard for us to accurately interpret probabilities, which makes it even harder to be good at statistics. Overall, our brains just aren’t made for mathematical and statistical thinking -- some strategies for practicing and improving these skills are listed above.

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.