Useful beats truthful: Survival and repetition of ideas in the information ecosystem

Viewpoint

Utility trumps truth in the world of information: useful information survives and gets repeated, even if it's not accurate. Truthful information goes extinct if it's not useful.

Background

About 20 years ago, I had the opportunity to participate in a corporate training program with a presentation from our company's chief scientist. One of the few things that I still remember from that training program was his assertion that within 10 or 15 years, "Everything that can be digitized will be digitized." I suppose he reached this conclusion because the rapidly falling cost of digital storage seemed to suggest that storage would eventually be nearly free and inexhaustible.

Here we are twenty years later, and a lot of information has been digitized, but there is also a lot that has not yet been saved. It's conceivable that much of the information that exists at a given moment may never be saved, let alone repeated. The question that occurs to me is, "why does some information survive and reproduce while other information evaporates?"

Introduction

We would like to think that truthful information is more likely to survive and false information is more likely to go extinct - that human knowledge is getting perpetually closer to the target of "truth". But, I'm starting to question this.

A little under a year ago, I posted about Donald Hoffman's book, The Case Against Reality: Why Evolution Hid the Truth from Our Eyes. Shortly after that, I read The Selfish Gene by Richard Dawkins. The Selfish Gene, of course, is a famous book because it introduced the idea of the meme as an evolutionary information concept that is analogous to the biological gene.

If we set these two readings against today's information landscape, maybe it provides some insight into the pervasive reach of "fake news", "misinformation", "disinformation", "malinformation", "Community Notes", and "fact checkers" that seem to dominate much of what we see, hear, and read from modern media sources.

Let's recall some relevant ideas from the two books and see how they apply to the survival and reproduction of information.

Hoffman: The Case Against Reality

Donald Hoffman is a computer scientist at UC Irvine. In "The Case Against Reality", he describes research that he has conducted with his students and other colleagues investigating the nature of perception and reality. He has investigated the topic from the perspective of theoretical physics and from the perspective of evolution/game theory.

As I recall, the book said that he and his collaborators have run numerous simulations of perceptive agents in virtual worlds, where agents that perceived the truth competed against other perceptive agents that perceived something that interpreted their virtual world surroundings in different ways.

What the simulations revealed was that perceiving "the whole truth" is computationally expensive. Agents that used this strategy inevitably went extinct. The agents that dominated the simulations were the ones that "rephrased" reality in a way that somehow conveyed actionable results in an efficient way. In the domain of perception, Hoffman termed this, "Fitness beats truth".

Hoffman's conclusion is that the reality that we perceive doesn't exist as we perceive it. In fact, it's not even an approximation of reality. Instead, it's an abstraction that only emerges as a result of our observation. Beneath the world of perception, he argues that a deeper layer of consciousness exists, and our perceptions give us no insights into the actual nature of that deeper reality.

He compares it to the GUI that we use to access our computers. We slide the mouse icon over a folder, click on it to see the files inside, and choose the one we want to open. None of those constructs actually exist inside the computer, though. Not the icon, not the folder, not the file, and not the words on the page. If we had to navigate the bits that were stored in the computer without benefit of the "Desktop" abstraction, it would be impossibly time consuming.

You can hear Hoffman describe it in his own words, here:

Dawkins: The Selfish Gene

I read "The Selfish Gene" because I wanted to learn about Dawkins' ideas on memes. It turns out, however, that the meme chapter was almost an afterthought in the book. The book dedicated many more pages arguing that the gene is the fundamental unit of evolution - not the individual animal, the family, or the group. In Dawkins' description, the gene is a replicator that is perpetually engaged in an effort to have itself repeated. On the other hand, the plants and animals that hold the genes are best thought of as "survival machines". Over the course of millennia, the genes have constructed more and more elaborate survival machines to aid in their efforts to be reproduced everywhere.

After building that case extensively, Dawkins added a chapter at the end where he argued that the meme is a new and more efficient form of replicator. In this view, a meme is a fundamental cultural unit that might be something like an idea, a belief or a behavior, and we can imagine that the related "survival machines" are human brains.

Combining the ideas

The things that these two books have in common is that they're both looking at things from the perspective of evolutionary theory. Hoffman's looking at metaphysical reality, and Dawkins is looking at the spread of information. So, what happens if we transport Hoffman's conclusion into the domain of information?

I have to say that I'm not totally persuaded by Hoffman's ideas about reality and perception. I guess my main concern is that I'm not sure how accurate a simulation can be when it is attempting to argue about the fundamental and unseen nature of reality. I don't doubt the results of his simulations, but I question whether they can be usefully applied to the real world.

However, I like the framework that he set up for comparing "fitness" and "truth", and I think the same framework might be relevant for the evolution of information. So, I have spent some time thinking about the nature of truth and usefulness in the modern information landscape. Basically, what we have is something like a 4-quadrant table:

Useful vs. Truthful	Truthful	Not truthful
Useful	Q1	Q2
Not Useful	Q3	Q4

It shouldn't be hard to agree on the first and fourth quadrants. If information is truthful and useful, we would expect it to be preserved and repeated. If information is neither truthful nor useful, we would expect it to be forgotten. But what about quadrants 2 and 3?

I would argue that information in quadrant 2, useful but not truthful, is more likely to be preserved and repeated than the truthful but not useful information in quadrant 3. Let's look at these quadrants more carefully.

Quadrant 3: Truthful, not useful

A true answer exists (or existed) to the question, "What did I eat for lunch on January 5, 2024?" However, I am confident that no one will ever know the answer to it. That information has been forgotten because it was of no known use to anyone. I suspect that we could come up with an infinite variety of questions like this. The world is littered with things that are true and almost immediately forgotten.

For survival and replication, then, it seems that truthfulness is not a sufficient condition. To decide if truthfulness is a necessary condition, we need to back-up to quadrant 2.

Quadrant 2: Useful, not truthful

We have to think about two different classes in this quadrant. Things that are approximations, and things that are just plain wrong. Let's start with the approximations.

As part of my day job, I am frequently on the phone with people from America and others from Europe. Often, when we are waiting for calls to start, the topic turns to the weather. The challenging thing about these conversations is that one side of the Atlantic talks in degrees Fahrenheit, and the other side talks in degrees Celsius. Of course, we all know the formulae to do the conversions:

• F = ⁹/₅ C + 32; and
• C = ⁵/₉ (F - 32)

But, just try doing those conversions in your head at the speed of conversation. So, I came up with a little mnemonic device for myself - DASH30 - Double and Add, Subtract and Halve.

• Double the C temperature and add 30 to get Fahrenheit; or
• Subtract 30 from the F temperature and halve it to get Celsius

Clearly wrong, but it turns out to be a close enough approximation for conversational purposes, and if I can remember DASH30, I can do it quickly in my head. False. But useful.

Now, what about things that are useful but just plain wrong?

One of the things we need to note is that the question of "useful" needs an object - "useful to whom?" This is the thing with lies. They are useful to the liar, if only in the short term. So the question of whether an untrue idea will survive and repeat depends on the number and the power of the people who benefit from it. I'm reminded of the old phrase, "A lie can travel half way around the world while the truth is putting on its shoes."

For examples of a false idea that survives and replicates, we can consider any number of urban legends that have been repeated for years, decades, maybe even centuries. For example, the myth that lemmings commit mass suicide was popularized by a Disney documentary in the 1950s, and it survives today. I have to assume that these ideas survive because they're useful to someone - perhaps for their entertainment or illustrative value.

Discussion

So there you have my reasoning. I conclude that truthfulness is neither necessary nor sufficient for an idea's survival and repetition. In contrast, I suspect that some degree of perceived usefulness is both necessary and sufficient. Let me restate my thesis:

Utility trumps truth in the world of information: useful information survives and gets repeated, even if it's not accurate. Truthful information goes extinct if it's not useful.

Based on this thesis, we can reason about our own relationships with information. If this is correct, then proving an idea false is not enough to stop it from spreading and repeating. In order to stop an idea from replicating, the utility of the idea must be eliminated, or the idea must be replaced by a different idea that provides better utility.

Another implication, if this opinion is correct, is that if we want our true ideas to spread, then it's not enough to show that they are true. We must also show the other meme survival machines how the ideas are useful.

Finally, this would mean that when we receive information, we always need to be aware of the incentives that are connected to it. We should consider not only whether the idea is true, but also who benefits from it, and how?

Here is a closing discussion question, courtesy of ChatGPT:

How can we ensure we're getting accurate information in a world where usefulness often outweighs truth? What steps can individuals take to verify the information they encounter and avoid spreading misinformation?

---

Update: It's surprising that the Steem blockchain is seemingly the first place that the phrase, "useful beats truthful", has been used on the Internet.

---

Thank you for your time and attention.

As a general rule, I up-vote comments that demonstrate "proof of reading".

---

Steve Palmer is an IT professional with three decades of professional experience in data communications and information systems. He holds a bachelor's degree in mathematics, a master's degree in computer science, and a master's degree in information systems and technology management. He has been awarded 3 US patents.

---

^{Pixabay license, source}

Reminder

Visit the /promoted page and #burnsteem25 to support the inflation-fighters who are helping to enable decentralized regulation of Steem token supply growth.