Creativity vs Universal Explainers

Bruce Nielson
6 min readAug 15, 2019

I’m carrying on a Twitter conversation with multiple Popperians about the difference between “Creativity” and “Universal Explainers.” Unfortunately, due to the limitations of Twitter, the conversation is hard to follow, so I’m writing it up here so that it’s more clear.

First, KS and I were discussing if Creativity required Universality or not. He was arguing Creativity required Universality and I was arguing that it did not if by “creativity” you mean “knowledge creation.”

As with a lot of conversations where you are thinking on your feet, it’s easy to talk past each other and I believe that was what was going on here. I eventually realized this, but didn’t at first.

Given that KS was arguing that creativity is always universal, it’s not surprising that he believes dogs to be automatons. I am not convinced dogs are automatons and it is well known that they have have the ability to learn (i.e. create knowledge via error correction) through classical conditioning. But they are not universal explainers and may have no explanatory knowledge at all for all I know.

So I made an argument to support my position that computers aren’t necessarily universal, but that at some point they make a leap to universality. (See this post for more info, but if you’re a Deutschian, you don’t need me to explain universal computers.) So maybe creativity is the same way. Maybe there are forms of creativity that aren’t universal and then a sudden leap to universality. (If by “creativity” you mean only “knowledge creation” then this is not merely a hypothesis as per my example of classical conditioning.)

Here are my words:

So I see no reason (yet) that creativity can’t be similar [i.e. to computers where they aren’t at first universal but then there is a sudden jump]. I’m not saying it is like that (I don’t know).

I realized later that actually, I did know. But I’m getting to that.

Brett Hall Joins the Fray

So at this point Brett Hall jumped in and responded to me:

*Because* it’s universal..so one can’t be “more universal”. It’s like wondering if one can be even *more* zero, or something. (There are levels of infinity…but what levels of universality of explanation could be, isn’t clear).

Now trying to read each other’s words required one to conjecture their meaning and then refute the bad interpretations. This is why we often talk past each other or misunderstand each other.

So the question is “what is Brett referring to when he says ‘it’s’ univseral?”

Was Brett Saying Creativity Was Always Universal?

Well, he’s responding to my post about “creativity,” so naturally I thought he meant:

*Because* [creativity is] universal..so one can’t be “more universal”. It’s like wondering if one can be even *more* zero, or something. (There are levels of infinity…but what levels of universality of explanation could be, isn’t clear).

In other words, I understood him as saying creativity is universal, just like KS is arguing.

At this point it finally occurred to me that we might be talking past each other. I have noticed in the past that sometimes Deutschians do use the word “creativity” to be the same as a “universal explainer.” And, this is what KS seems to be doing (more examples here and here).

And, if the “it’s” in Brett’s statement refers to creativity (if it doesn’t, what does it refer to? I’m not sure) then it must be the case that Brett is also, in that Tweet anyhow, tautologically understanding “creativity” as the same as a “universal explainer.”

However, I was not using “creativity” to be equivalent to “universal explainer,” so I suggested we might just be using different definitions and that we’re talking past each other. See here and here.

Brett Claims I Misrepresented His Words

At this point, Brett claimed that I was misrepresenting what he said:

You’ve made 3 assertions about my view (and David’s there) - all of which are false. (1) it’s not a tautology because (2) we’re not arguing it’s *equivalent* (3) creativity of the human kind merely depends upon universal computation which is *part of* a universal explainer.

Okay, so something is wrong here. If Brett’s “it’s” meant “creativity” (what else could it mean and have his argument still make sense? I’m not sure.) then he was tautologically using “creativity” to refer to the “universal” kind.

But having pointed that out (or so I thought), he’s now saying that I’m falsely representing his view and that he is not arguing “creativity” is the same as “being a universal explainer.”

I’m Confused By Brett’s Argument

So I’m really not clear what Brett’s argument above was actually about. It seems to me that there are only two possibilities here: 1) He was, in the first tweet, referring to “creativity” as always being “universal” (in which case I was right that he was tautologically equating the terms) in that Tweet and then later he stopped equating them and didn’t realize he had switched definitions, or 2) I initially misunderstood him because his “it’s” didn’t refer to “creativity.”

I’m taking time on this point because nearly everything that followed was confusion over this point. I’m not trying to put words into Brett’s mouth here, I’m asking for clarification of his meaning in his first Tweet.

Brett Hall Clarifies “Creative” Doesn’t Require “Universality”

Okay, so Brett goes on to make it very clear that he does accept that the word “creativity” can and does refer to things other than “universal explainers.” See here and here and here.

Biological Evolution is Creative, But Not Universal

Meanwhile, I continue my conversation with KS and I finally ask him if he considered biological evolution to be creative. If he is (at that moment) equating creativity with being a universal explainer, then I’d expect him to answer that evolution is not creative. Here is KSs answer:

Not biological evolution because I thought creativity requires a running computer program.

I responded:

So most people do think of evolution as ‘creative’ and with good reason. Because it creates knowledge.

At this point, KS agreed with me.

Defining Creative and Human-Creativity (i.e. Universal Explainer)

I now better understood my own argument, so I proposed that we refine how we’re using our words to be the following:

Proposal to @ks445599 and @ToKTeacher. Let’s avoid word confusion by using “creative” to mean anything that creates knowledge (including evolution) and “human-creativity” for explanatory knowledge created by a UE.

Brett, in the meantime, on a separate thread ended up agreeing with me on my definition of creativity:

Evolution is creative. But it’s not creativity of the explanatory sort.

My Position: Dogs are Creative, but not Universal

And my final post, I explained my own argument using my refined definitions:

Let me restate my position now that we’ve defined terms better and I’ve been able to get my own thoughts more straight: I believe a dog is not an “automaton” because it has a brain that creates knowledge via an evolutionary algorithm of some sort that is NOT a universal explainer

Summary

My point of view is that dogs are not automatons because I believe their brains are actually running a real genetic algorithm of some sort and that it is a creative process because it uses evolution to create knowledge. They are not universal explainers.

I believe KS is simply using “creativity” as equivalent to “universal explainer” and since dogs are not universal explainers, they are not “creative” in the sense that he is using the term. So we’re both right. (There is still an open disagreement between us if dogs are automatons or not.)

I believe Brett initially started using “creative” as equivalent to “universal explainer” (thus his tweet claiming creativity was universal — assuming here I understood him correctly) and then when I called him out on it, didn’t realize that he had done this and started using “creativity” to mean that which evolution does — which is not equivalent to a universal explainer. If my conjecture is correct, that was why he thought I had misrepresented his views. In reality, he had simply unconsciously moved from one legitimate definition of “creative” to another.

Another possibility here is that I had in fact misunderstood him. I would like to see Brett clarify his first Tweet and resolve what his original intended meaning was. But it looks like I am not disagreeing with Brett in any capacity. (He hasn’t weighed in on whether or not he believes dogs are automatons or if they are creative but not universal explainers.)

--

--

Bruce Nielson

Bruce is a Master's student specializing in Machine Learning and Artificial Intelligence at the Georgia Institute of Technology.