Innovation!

2025-06-25

Innovation!

It always has a positive ring to it. Sometimes, it’s not just a positive ring, but spit-dribbling religious awe, stary eyed distant look, as the rapture of a paradise procurred by innovation invest the corporeal form of the orator.

So what’s innovation?

It’s doing the same thing differently.

Well, maybe not. Because, Doing the same thing differently is, as a sentence, a paradox. The way of doing thing is part of the thing itself. You know how free-range-hen eggs cost more than other types of eggs? Well, that’s because it’s not just the end product that matters, but also the way it was made. Otherwise, the two types of eggs would be fungible and have the same price.

So let’s try to go with a less dead-end definition:

Innovation is changing a process, so that certain characteristics of that process (generally the output) is the same, or lightly changed, while other aspects are radically transformed. But most importantly, that change must be good.

It’s not as short and snappy as the first definition. But let’s break it down:

As a software developer, I’m deeply aware that change is not a panacea.

To jump outside of the world of pure abstract ideas, and into the domain of programming for a second, here is an often overlooked aspect of programming:

When programming, you can implement the same thing in many different ways. For example, the same program may be written in Common Lisp or Python. That’s not a random example, reddit was originally written in Lisp, and later switched to Python. Twitter was first written in Ruby, then Scala, then Java. And that’s just about programming languages! It’s almost irrelevant to the end-result, other choices matter much more.

The truth is, there is no best choice, as there is no best programming languages1. The choice depends on the context. A choice in a specific context may be good, not because it’s just good, but because the bad aspects of it are less relevant in the given context, and the good aspects strengthened. And of course, with time, the context might change. The choice may even influence the context, etc.

1

Except Rust, of course.

This is a big mess. And we are talking about programming, you know, pure logical processes and formal methods. When it comes to our earthly planet, and lowly human interactions, context is both much more important and its impact much more fuzzy.

In fact, a program, as a process of production, has many dimensions: How much memory does it use? Is it parallelizable? How many instructions does it use? What is the memory-locality of the algorithm? How good is the result? How much knowledge of the system do I need to change elements of it? Knowledge of the system in its particularity, or systems of similar kind? How many bugs has it? Is the system vulnerable to hacks? How much working people does it need for its everyday operation?

When a process can be judged on all those orthogonal2 questions, a value judgement has exaclty one dimension: “Is process A better than process B?” This is what is called projection in math, meaning: taking a set of values, and applying on it a function to get a single value back3.

2

orthogonal in the sense that answering one question tells nothing about the answer of another one.

3

Or a smaller set of values, for all the Mr. Pedentics reading.

So, innovation depends on this function that takes a complex thing and labels it as good or bad. Of course, this function depends entirely on what you think is good or bad. You may ignore some aspect, and place much more moral value on another aspect.

In fact, innovation is nothing else than using a different set of tradeoffs. Just like in programming.

Innovation comes in three forms:

  1. The context has changed, while the process has not. The context here is material ressources, or scientific knowledge. While the process was good in the past, it isn’t in the new context. Changing the process to make it better suited to the current context is good, and therefore is an “innovation”.
  2. Social values have changed, while the process has not. Now, innovation is about changing the system of humans relations to comply with the new social values. In the past 40 years, this means giving less to the ones who have little (as we think them as responsible of their own misery) while giving more to the ones who already have a lot. Rail privatization in the UK and France is a good example. It was justified by “innovation” but in this context, “innovation” means little else than busting up rail unions. The savings made by making rail workers’s life a misery didn’t help reduce the rail fare (the opposite is true!) but fell into the pockets of a few shareholders.
  3. Finally, a particularly nasty kind of innovation, is one that is fully conscious of the ambiguity of its moral nature. This kind of innovation is not changing systems in order to apply to a new social or material context, but rather the inverse: changing the social or material context to justify a specific way of doing things, that will later be labelled as innovation. Example: Uber was breaking the law. Uber justified itself by accusing the law of being immoral. The reasoning is: Uber is innovative, Uber is breaking the law, everything that is innovative is morally superior, therefore the law is morally inferior.

In fact (3) should be linked with (2). This is a strategic use of the word Innovation based on an equivocation.

On one side, we have the dictionary definition of Innovation as positive change of process; On the other side we have the economic definition of Innovation, a similar definition, but with a very narrow understanding of positive as incures less cost.

The common sense understanding is that reducing cost at the expense of everything else is bad, the exact opposite of what Innovation means in the economic sense. Yet, this definition of innovation was used to justify the major social changes that led to our society.

In fact, there was an inversion of the source of morality: in the recent past, a process was awarded the title of innovative whenever it was a good change in society4. In the 1980, this got reversed: (some) economists redefined innovative as reduces costs, then gave this title to a set of policies, and as a result, those policies were considered as good changes in society.

4

Up to the mid XXth century, innovation as in trying to change something that already works had a pejorative meaning.

The sleight of hand has gone unnoticied for decades. But now, the material conscequence of this change is hitting people in the face.

The previous definition of innovation as a moral good based on social changes and material improvements to human life led to a technological boom and prosperity. What “moral good” and “material improvement” to pursue might have been flawed, or biased by personal interest, or differed from people to people. But the little that overlapped produced prosperity, and that, despite personal interest, not thank to it. The current narrow definition led to financialization, innequality, and exploitation of the planet and the people, as using profit optimization as sole goal cannot encode the rich and complex needs of a society.

We need to remember that the market is not an end in itself, but a tool. We should judge policies not by how much they favor one specific tool, but rather how they affect society at large, and if that’s the society we want to live in.

Innovation is a rebalancing of tradeoffs, looking exclusively at the market impact and ignoring all other aspects will invariably lead to innovations that work at the expanse of everything outside of the market.