How Does Artificial Intelligence Create Value?
A more philosophical view of the technology’s impact
The last decade has been filled with fanfare and buzz — no news there. However, there’s now a distinct shift in various industries (at least from an AI consultant’s perspective) about how the general interest and uptake of new ML technologies and approaches went from esoteric to exotic to fear of obsolescence.
Industries and clients are more ready than ever to partake in prototype and scale-up efforts using relatively new (even unproven) approaches to data manipulation and insights extraction. But as most consultants find out, a simple application of machine learning does not a successful solution make. There needs to be an underlying problem being solved, with some measurable outcomes and returns at least an order of magnitude above the implementation costs.
So how is value usually created, and how does AI contribute towards that value creation? And is there a pattern that can explain why AI is so valuable?
Understanding Value Creation in the Modern Economy
There are many definitions of “value”, but for our purposes, we’ll define it as “the modern economic measure of importance, worth, or usefulness of something” (from Oxford Languages). From this, the idea of value creation in an economic system is directly related to the ability to generate wealth, reduce operating costs, or accelerate and alleviate the work burden of individuals.
(Now, measuring the societal impact of AI is an entirely separate conversation that I won’t even pretend to broach here. Also, I’m fully aware that I’ll be butchering Econ 101 concepts, but I’m generalizing towards the point and main message of this article.)
Beyond the Three-Sector Model of (Old) Industry
A popular model for economic value creation was the Three-Sector Model, where the 20th-century view of economic value was separated into three main areas:
- the primary sector focuses on the extraction and manipulation of resources, such as iron, lumber, and mining;
- the secondary sector focuses on manufacturing, transformation, up to the point of delivering consumer goods;
- the tertiary sector, which is usually described as the services sector, includes all matters of government, consulting, and knowledge workers.
This particular model focused extensively on an extractive basis for value creation, both towards the environment and towards people. There is also an idea of finite resources: there can only be so many skilled workers, or so many trees to cut down. A participant in this economic model was primarily viewed as a laborer: a lumberjack, a smelter, a restaurant waiter. You get paid by 1) showing up to work, and 2) doing the work.
However, this model fails to capture the latent knowledge required to be effective in each and every one of these sectors. Just giving someone a chainsaw near a forest does not mean that I can count on safely and reliably getting a certain amount of lumber every day.
Beyond Three Sectors: The Knowledge Economy
Also known as The Fourth Revolution and The Information Age, the knowledge economy model allows for all matters of information (such as market insights, algorithms, and processes) to be represented as valuable information, and workers now become thinkers.
From Wikipedia:
A knowledge economy features a highly skilled workforce within the microeconomic and macroeconomic environment; institutions and industries create jobs that demand specialized skills in order to meet the global market needs. Knowledge is viewed as an additional input to labour and capital. In principle, one’s primary individual capital is knowledge together with the ability to perform so as to create economic value.
So if knowledge is valuable, isn’t the automated application of that application also valuable? Indeed, that’s where automation as a whole and specifically software come in. I can encode my knowledge in a repeatable series of steps that will yield expected results, even outcomes.
(Fun fact about this encoding process, though: software developers, as a class of people, experience tremendously high rates of burnout. I also often hear my team use language that would make a sailor blush.)
Back to the Future — With ML
Stepping back from these distinct models, we’re looking to answer the high-level question about AI and its general ability to become valuable in most circumstances, given appropriate data.
If we abstract the Three Sector model into activity types, we notice three repeatable aspects of digitally relevant value creation elements. This is a useful inspiration to many chains of value creation in the technology sector. Abstracting these three sectors into general elements of value creation, we get:
- Extraction and collection, which involve the accumulation of a particular unit by separating it, isolating it, or duplicating it from its original environment.
- Transformation, which involves applying one or many changes (such as processing or digestion) to this unit so that it may be more usable in a different state, usually with all other units of its class, and sometimes with multiple classes of these units.
- Information, which is the knowledge or know-how related to the first two categories (such as where to extract and how to transform).
For instance, in the mining industry, knowing where to dig is equally as important as knowing how to operate the machinery that will get those ores out of the ground.
(In fact, a case can be made about how modern consulting is simply capitalizing on information differential — you know how to do something that I don’t, and I need that knowledge, therefore your knowledge is valuable given my present situation.)
Value in the Knowledge Economy: The Data Transformation Spectrum
The way data moves, changes, and informs is the premise behind creating economic value in and beyond the Knowledge Economy. We’re finally seeing the results of the Big Data push of the early 2000s, now that there are ways of making it actionable.
If we take all forms of data (whether they be in their raw, uncollected states, all the way to polished corporate reports) and place them into buckets organized by level of refinement, we can create five broad states: connectivity, visibility, aggregation, insights generation, and decision support.
This Data Transformation spectrum is a simple model that can be used as an inspiration for technological innovation in a highly connected world. It covers many potential aspects of value creation where data acts as a primary resource.
Companies will pay to get something connected, visible, aggregated, comprehensible, and/or automated.
Simply put, value gets created when data moves across the Spectrum.
So Why is AI So Transformative?
Machine learning is the first technological paradigm that can be applied to every aspect of the value creation chain within the Data Transformation Spectrum. Because of its depth and breadth, the combination of data science and machine learning engineering allows for insights generation, acceleration of work, and investigative support for every component along the chain.
The reason AI is so valuable is because it provides elements of valuation creation both within and between every pillar of data transformation.
This is why AI can create so much value: it is applicable in almost any given data-relevant context, provides some sort of value (either through the manipulation of the data or generating results within its context), and allows for more and more automation and insights derived from the data.
Summary
- Value is created through a system that generates money, reduces costs, and alleviates the burden on resources and laborers.
- Historically, value was created by manual labor; now, in a knowledge economy, information and reproducible software meant a lower cost to sharing burden alleviation.
- The foundational elements of value creation through technology are the collection and extraction of resources, the transformation of these resources into new elements, and knowledge on how to perform the first two.
- The ability to extract insights (valuable and actionable information) from data happens throughout a spectrum of transformation.
- AI creates value at every pillar of this spectrum.
Happy Consulting!
-Matt.
If you have additional questions about this article or our AI consulting framework, feel free to reach out by LinkedIn or by email.
