Evolution of business models over the past 5,000 years
How business models mirror the evolution of civilization
A business model is a framework of how a company creates & delivers value for its customers and generates revenue. But look deeper, and you’ll find it mirrors our civilizational stack—how we think, organize, and scale human effort.
Let’s walk through the epochs.
1. Agrarian Age (Pre-3,000 BC): Labor and Land
Flow: Create → Deliver → Sell
Since the invention of agriculture, the challenge wasn’t market demand—it was survival. You had to till the land, grow the crops, and get them to market. Demand for food was always there, so the value chain bottleneck was making food.
Total value was determined by labor and land available to making and deliverying crops, within the constraint of the weather.
2. Industrial Revolution (1760s-): Machines and Persuasion
Flow: Create → Deliver → Sell
Machines redefined the creation and delivery of goods. We mass-produced physical goods (like textiles) in factories, and transported these goods by cars at a scale and speed far beyond what humans alone could have done. The value chain bottleneck shifted to sales, leading to the rise of marketing, branding, and psychology in commerce.
Total value was determined by total ability to persuade consumers to buy goods, within the physical constraints of factories and vehicles.
3. E-commerce Era (Early 1990s): Codes and Discovery
Flow: Create → eDistribute → eSell → Deliver
The internet unbundled geography and broke up traditional function of delivery into (i) distribution (making products accessible) and (ii) delivery (making products in the hands of consumers). We distributed and sold products online, to then, post-purchase, delivered the goods. The value chain bottleneck shifted to distribution, or having the products discovered, leading to digital marketing, SEOs and clickbait.
Total value was determined by total ability to distribute products online, within the physical constraints of factories and vehicles.
4. SaaS Era (2000s-): Codes and Engineers
Flow: eCreate → eDistribute → eSell → eDeliver
The digitalization of products eliminated the physical constraints of goods. We created software to be distributed, sold, and delivered online, all on code. The value chain bottleneck shifted to building (the initial creation of and subsequent iteration of codes), where engineering replaced factories as the source of scale.
Total value was determined by total ability to distribute and sell digital products within the constraints of coding.
5. Gen AI Era (2020s-): Codes to code codes
Flow: AICreate → eDistribute → eSell → eDeliver
Generative AI has now eliminated the labor constraints of coding. AI models can now build, distribute, and generate content for the product (not in one go yet, but sooner or later). The new value chain bottleneck is back to persuasion, where getting consumers to decide to buy is ultimately the only constraint.
Total value is total ability to persuade consumer to buy AI products within the constraints of total amount of attention by consumers.
What is the next evolutionary path?
What comes next isn’t just automation. It’s autonomy.
As apparent throughout the evolution of business models, value is generated at the ultimate bottleneck of value chains, which attracts capital and talent to solve for it.
We are now heading toward an era where:
AICreate: not just code, but physical products through generative design + manufacturing. We will use AI to manufacture machines to then manufacture goods. This will eliminate the labor constraints of physical goods .
AIDistribute: via AI agents that act on your behalf, surfacing products contextually. We will have LEO (LLM Engine Optimizations) or AEO (AI-agent Engine Optimizations) to have our products discovered and babyfed (placed right in front of customers to just say “yes”). This will eliminate the human-targeted distribution needs.
AISell: via AI agents/ AI-assisted human sales that optimize offers for intent and behavior dynamically, 24/7. We will have AI to make us want to buy things (but think it was by our own will). This will eliminate the traditional sales logic.
AIDelivers: via self-moving vehicles, leveraging real-time intelligent logistics or even predictive needs. We will have embedded systems to deliver physical goods ASAB (as soon as bought). This will eliminate the physical constraint of warehousing and delivery logistics.
Such applications of AI are quite obvious and straightforward— it is only a question of when not if. And we will see a great amount of talent and capital pouring into these (for founders interested in the next trillion-dollar companies- these are it).
But this doesn't fundamentally change the flow of business models we’ve seen to date. Is there anything that will?
The fundamentally new business model
All the above have been predicated on the assumption that we humans are making the decision to buy products and services. But what if we don’t have to? What if AI can make the decision for us?
We are entering the Decision AI Era, where AI will buy products and services for us on our behalf. Initially, a few immaterial, eventually semi-full decisions will be made by AI to optimize for happiness, wealth, status, attention, you name it. Once we start to delegate purchase decisions to AI to optimize for a certain goal or function, the entire value chain will be flipped.
New Flow: AISell → AICreate & AIDeliver
Total value will be total amount of purchasing decisions we’ll let AI make.
In the Decision AI era, products won’t win because they’re better. They’ll be made as what AI chooses for us.
The logic of business is evolving—from serving human wants to anticipating them, then eventually overwriting them.
And so the question then isn't how much AI can do. It's how much we’re willing to delegate before we forget what it means to choose.
AI will handle the “how.”
But only we can choose the “why.”
The future will need more sovereignty.
To decide or not to decide, that is the question.
That’s it for this week. If this resonated, share it with those interested in the next generation of business models. And if you haven’t yet, subscribe to keep exploring deep thoughts.
With Love,
Koshu
Love this framework for understanding the evolution of business models. As AI starts making decisions for us, will we still value products for their inherent qualities, or will our tastes be shaped by algorithms? It’s a thought-provoking shift in how we define value.