2
$\begingroup$

Premise

The Dunning-Kruger effect is a sociological / psychological phenomenon whereby low-skilled individuals mistakenly think they are good at something. In other words, individuals lack the meta-cognition needed to give an unbiased evaluation of their competence, which leads to an unwarranted level of confidence -- or at least that is the case as measured by empirical studies of the present.

enter image description here

Analogy: The Internet

Following the recent advent of the internet, we live in a world with very accessible information, but not everyone gets the same utility out of this information. Just because the information is available, that doesn't change the facts on the ground: we have pluralistic cultures, many voices/interpretations, not to mention different innate cognitive abilities. In other words, just because we have easy-to-get information in the internet, doesn't mean everyone accepts it and/or understands it, or has time to try. It's an interesting facet of the Dunning-Kruger situation, but developments are still ongoing; it's hard to make definitive claims. I'm certainly not trying to make definitive claims either, I'm just trying to give my question more insightful context. Nonetheless, the internet analogy will provide an antertaining baseline for this thought experiment.

World

Instead of providing easy access to information like the internet, let us imagine that in the trans-humans hardwire the information to their brains. This is done by implanting neural computer chips in their brains. Consider the chips to be your stereotypical neural chip implant, capable of downloading a wealth of information and allowing the trans-humans to learn calculus, foreign languages as well as a multitude of other knowledge bases instantly. The idea is that with a neural chip, we no longer have data or knowledge in the nominal sense, but also the ability to implement. Suffice to say the neural chip will be the equivalent of having the software and user manual all on-board (as the comments are right to point out: a table of engineering formulae are not the same as an engineer) Let's also stipulate it is a distributed network that updates itself via some fancy algorithm when/if the hard-wired knowledge is outdated (something new is learned).

Aftermath

Now that we have distributed the pinnacle of mankind's knowledge in a uniform fashion, I'm curious as to what becomes of the Dunning-Kruger effect. Here are my observations:

What works in neural chips favor:

  • Uniformity of knowledge: everyone has the same level of knowledge, so presumably no one's confidence will be higher or lower than anyone elses

Skepticism:

  • Specialization: there are so many skills, knowing everything there is to know in the universe might just be hubris

Question

Is there reason to believe that having the entire corpus of mankind's knowledge stored on a neural-chip would ameliorate the Dunning-Kruger effect, or would it just make it worse? Why?

Asumptions:

  • Absolute benevolence: there are no bad actors installing neural implants to brain-wash people or otherwise control their minds
  • Universal hardware/software: there are no "VIP" trans-humans who have superior neural-chips or better knowledge software. Everyone has equal equipment installed
  • Muscle-memory: fine motor skills also can be downloaded, so we can have Matrix kung-fu and piano playing hands
  • near-future: we already have chips in animal brains and can transfer basic data, let's assume it's not too far off
$\endgroup$
  • 1
    $\begingroup$ Can the trans humans check versions or determine if they have a software? For instance, would you know that you have downloaded "jiu jitsu.exe" version 1.337? $\endgroup$ – Crettig Jul 16 '18 at 23:43
  • $\begingroup$ @Crettig Well, we're assuming everyone has the same version since it's updated automatically. But yea, in principle, you could check the version. $\endgroup$ – Arash Howaida Jul 16 '18 at 23:45
  • 2
    $\begingroup$ As a many years ago I used to work in education, I must say that this question makes a rather fundamental confusion between the ability to have access to data and learning. Being able to summon up instantly any piece of data about calculus is not at all the same as knowing calculus. Being able to look up instantly any word in a dictionary is not at all the same as knowing a language. People have had immediate access to vast amounts of knowledge since the very first edition of the Encyclopedia Britannica; this did not made learning useless. A book of engineering tables is not an engineer. $\endgroup$ – AlexP Jul 17 '18 at 12:05
  • 2
    $\begingroup$ You cannot "download the software". Learning makes physical changes in the brain. The brain is not a computer, the "software" is not independent of the "hardware". $\endgroup$ – AlexP Jul 17 '18 at 12:51
  • 1
    $\begingroup$ Primary consideration: Conscious vs unconscious competence: having knowledge available vs enough experience applying said knowledge to develop and optimise your application per circumstance, improving performance. Secondary consideration: Muscle-memory appears to reside in local neural bundles near to the site of the muscles in question - it's not in central brain storage. This means updating those datasets would be both difficult and highly confusing - I had an arm badly injured and it took me 6 months to relearn how to use my hand - huge mental and emotional impact. $\endgroup$ – GerardFalla Jul 17 '18 at 16:04
2
$\begingroup$

Neural-chips would have no effect on the Dunning-Kruger effect.

What most people don't realize is that the Dunning-Kruger effect, when projected backward or forward is a saw-tooth pattern.

All the chips would do is make people very confident that they know how the world works.

They will be confident until something comes up and shows them that they are wrong. Then they will realize that they were climbing Mt. Stupid, not Mt. Guru. Then their confidence that they know how the world works will go down until they figure out how the world really works. Wash, rinse, repeat.

If you ever have the chance, watch James Burke's The Day the Universe Changed. The basic premise of the series is that, "At every point in history, we have been certain how the world works and every single time, we have been dead wrong." I highly recommend this series (and Connections) for everyone.

The only things that the neural chips will do is make the cycle faster and make everyone wrong in the same way. I expect that people will get use to that.

There are two other outcomes, however:

  1. If everyone thinks the same, there may be no one to ask the questions that let them know that they are wrong and they forever climb Mt. Stupid.
  2. They may realize that there is no Mt. Guru and try to level Mt. Stupid as soon as possible.
$\endgroup$
  • $\begingroup$ The Dunning-Kruger effect is a comparison between individuals not societies so projecting it is meaningless. . $\endgroup$ – John Oct 10 '18 at 4:57
4
$\begingroup$

Pure Knowledge != Knowing What To Do With the Knowledge

The effect will be in some ways better and in some ways worse, but it certainly won't go away. Knowledge of a domain is not the same as knowing what to do with that knowledge. Some examples:

  • Math

If you have a website with tables for trigonometric functions then there is no difference between knowing how to calculate those values and simply reading them off of the database. Similarly, if you have "infinite" multiplication tables effectively memorized then everyone is able to be a math wizard and answer 514,139,613 * 415,978,910 or whatever is asked.

But if you ask someone to calculate the area of a polygon (excluding the simple ones like rectangles and triangles) then simple knowledge of formulas and tables of values will not help, or will only help for some parts of the problem and Mt. Stupid will still be a big hill.

  • Business/Economics

Anyone will be able to calculate compound interest or amortization tables because the data will be instantly available. But that won't translate into everyone becoming a wizard of Wall Street.

  • Medicine

Anyone will be able to compare an X-ray or MRI against thousands of samples to determine healthy vs. cancer instantly. But that won't give people the insight to properly treat patients with multiple problems or provide a good bedside manner.

Actually, the more I think about it, the more this gets to the issue of Expert Systems vs. Artificial Intelligence. Essentially, every human could become as "expert" as the best Expert System - raw knowledge in HUGE quantities + fairly simplistic algorithms for filtering/using that knowledge. (I know Expert Systems are already past that stage in many cases, but I am picturing how the typical low-skipped individual would make use of the knowledge.) But if they are lacking in intelligence then they will not magically become artificially intelligent. The androids will have a better chance at that.

$\endgroup$
1
$\begingroup$

I don't see how this would do anything other than annihilate the Dunning-Kruger effect. Assuming that by "learn" you actually mean "learn" and not "have fast, inherent access to endless lookup tables" then the only reason for concern would be that humanity, collectively would be wrong about something. That's definitely possible, but:

Everyone immediately gains the maximum available degree of knowledge in any area of even passing interest, and that knowledge is always as complete as the bleeding edge of what knowledge-generation can provide. And with the time and effort required to learn something at zero each, why would anyone not learn something? This should have a few effects:

  • Everyone can be completely confident in their knowledge relative to what anyone knows. There is no chance that you've misunderstood something or otherwise have an inferior understanding relative to the greatest thinkers on that topic, ever.
  • Research and other methods of testing/validating knowledge should speed up dramatically. You learn 100% of what's known about some field of active scientific research, for example, along with 100% of the knowledge and technical skill needed to operate the most advanced experimental or analytical equipment and software in existence. This would radically lower the cost of entry into participating in the most advanced research environment on any topic at any time but also be like a subscription to every research journal in existence with zero time investment in reading and perfect comprehension.
  • You always know the precise bounds of what is currently understood, along with how and why the state of knowledge is what it is.
  • The value of research stays the same as it is now, but also becomes much easier to do-- expert information on how to structure and execute experiments and simulations would be available to you. And as new information is produced it is immediately disseminated to everyone else, so everyone is always up to date with zero effort, so the realm of what is known should expand wildly and quickly.
  • While just having information isn't the same as processing it (a great scientist might still be better at thinking through new problems than someone else, even with the same knowledge base) that is irrelevant to Dunning-Kruger.
  • Along with every other field, I don't see why epistemology wouldn't advance as well, especially in fields that can be technically and precisely defined (the ones where "right and wrong" answers are meaningful). So in addition to gaining great knowledge in the subjects themselves, everyone would have access to the most advanced ideas regarding what can be known and what "knowledge" as a concept even is. I would expect this to really dissuade people from overconfidence in knowledge.
  • Finally, depending on what you're thinking of including as "knowledge", information on how knowledge can be applied to various situations could also be disseminated to everyone, all the time. You don't need great problem solving skills to follow step-by-step instructions that you already perfectly understand, and don't even have to expend any mental effort to choose which instructions to follow-- as long as you can perceive a situation, you'll know what to do!

The only chance I see for everyday people to be "wrong" about something they encounter is if it's totally novel. And how confident would someone be in the face of an experience where they don't already know everything there is to know about it?

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.