OpenAI — Eating It’s Customers for Breakfast, Lunch and Dinner
The Smiling Assassin, and Innovate, Leverage, Commoditise (ILC).
Introduction
TLDR
Through the lens of the Innovate-Leverage-Commoditise (ILC) model, we explore how OpenAI’s strategy might unfold. OpenAI could potentially utilise its platform as an innovation incubator, learning from the endeavors of its users.
Subsequently, it could leverage these user-driven innovations, integrating them into the platform as new or enhanced services. This ILC-based approach could significantly shape the platform’s evolution and its impact on users and competitors alike.
Understanding The ILC Model
This will require some knowledge of Wardley Maps. You can find more on Wardley Mapping here.
The Innovate-Leverage-Commoditise (ILC) model serves as a powerful strategic tool for platform owners, allowing them to fuel growth and gain a competitive edge in their ecosystem. OpenAI, with its unique position as a platform provider, appears particularly poised to harness this strategy. This approach would enable OpenAI to co-opt user innovations, convert them into valuable offerings, and subsequently, establish market norms.
“Innovate: Platform users and developers create new features, services, or use-cases, adding a layer of innovation to the platform’s existing offerings.
Leverage: The platform owner observes, learns from, and profits from these user-driven innovations, amplifying the successful ones for broader adoption.
Commoditise: Subsequently, these innovations become standardized and integrated into the platform’s core offerings, reinforcing the platform’s indispensability while simultaneously reducing reliance on individual innovators.
The idea is to take an existing, well-defined product and turn it into an industrialised utility with an easy-to-use API. This way, other companies can innovate by building on top of your utility and reduce their cost of failure while increasing their agility. These companies building on top of your utility are your “outside” pioneers or what are commonly called an “ecosystem.”
The more companies you have building on top of your utility (i.e., the larger your ecosystem), the wider the scope of new innovations, and the more things your “outside” pioneers will be building.
Your “outside” ecosystem is, in fact, your future sensing engine. By monitoring metadata such as the consumption of your utility services, you can determine what is becoming successful. It’s important to note that you don’t need to examine the data of those “outside” companies, but only the metadata. Hence, you can balance security concerns with future sensing.
You should use this metadata to identify new patterns that are suitable for provision as industrialised components. Once you’ve identified a future pattern, you should industrialise it to a discrete component service provided as a utility and exposed through an API. You’re now providing multiple components in an ever-growing platform of component services for others to build upon. You then repeat this virtuous circle.
Companies in any space that you’ve just industrialised might complain — “they’ve eaten our business model” — so you’ll have to carefully balance acquisition with implementation.
On the upside, the more component services you provide in your platform, the more attractive it becomes to others. You’ll need to manage this ecosystem as a gardener, encouraging new crops (“outside companies”) to grow and being careful not to harvest too much. This creates an ever-expanding platform.
The larger the ecosystem that is built, the more powerful the benefits become. There is a network effect here.
The ILC model has some subtle beauty. If they consider the ecosystem as the companies building on top of the discrete component services, then the larger the ecosystem, the greater the economies of scale in our underlying components. This creates more metadata to identify future patterns, which in turn leads to more innovative components built on top, and a wider future environment that they can scan.
The ecosystem can also be a fertile hunting ground to recruit
As a result, they become highly efficient as they industrialise components to commodity forms with economies of scale, and highly customer-focused by leveraging metadata to find patterns that others want.
All these desirable qualities will increase with the size of the ecosystem, as long as they mine the metadata and act as an effective gardener.
Accelerators, Decelerators, and Constraints
Within the interplay of the Innovate-Leverage-Commoditise cycle, there exist factors which can either accelerate or decelerate the process for OpenAI. For instance, user adaptability and market demand could function as accelerators, while regulatory constraints or technical debt might decelerate the cycle. Understanding these dynamics can help us predict how OpenAI may navigate its strategic path within the ILC framework.
Open source efforts through collaboration with others have produced stunning technology that out surpassed proprietary efforts in many fields. In many cases, open source technology has become the de facto standard and even the commodity in a field. Open source seems to accelerate competition for whatever activity it was applied to.
There are however counter forces that exist such as fear, uncertainty and doubt. This was often applied by vendors to open source projects to dissuade others by reinforcing any inertia they had to change. Open source projects were invariably accused of being not secure, open to hackers, of dubious pedigree and of being a risk.
However intellectual property is used for ring-fencing to prevent a competitor developing a product.
The evolution of a component can be constrained by a component it depended upon such as knowledge.
Certain actions would accelerate competition and drive a component towards a commodity whilst others could be used to slow its evolution.
Point 1: An open approach, whether it’s open source or open data, can accelerate the evolution of a component.
Point 2: Fear, uncertainty, and doubt can slow down the evolution of a component when crossing an inertia barrier, and patents can be used to protect a technology.
Point 3: Constraints in underlying components can affect the evolution of a component. For example, converting compute to a utility would cause a rapid increase in demand due to new uncharted components built upon it or the long tail of unmet business needs. However, this requires building data centers, which is not as rapid as providing virtual machine
OpenAI — The Play
OpenAI’s potential adoption of the ILC strategy could reshape the AI landscape. By nurturing a broad ecology of innovation on its platform, OpenAI can gain insights from a plethora of user-driven advancements.
This perpetual loop of learning and commoditising can allow OpenAI to anticipate market trends and introduce services that meet evolving demands, thereby reinforcing its market position while simultaneously shaping the dynamics of the AI ecosystem.
In the Wardley Map above about Prompt Engineering shows the various components required, the value chain and dependencies, along with their current state of evolution from genesis, through to commodity.
More on Prompt Engineering over on the blog.
What do we know?
What can we learn from the Wardley Map.
It’s key that prompt templates, agents and a vector database is required to support any form of chat history. To store your own data a vector database is required, along with converting your data into embeddings.
We’ve seen the fear, uncertainty and doubt play from OpenAIs competitors.
We’ve seen OpenAI use Open Source to accelerate their components within the platform.
Embeddings are used to enhance LLMs with private data. Embeddings are numbers, numbers can’t be copyrighted.
What can we (OpenAI) learn?
- OpenAI is building a very large ecosystem based on its platform.
- It has a significant capability to learn from data outside and within it’s ecosystem.
- It has a significant capability to learn from metadata within it’s platform.
Predictions
If OpenAI fully embraces the ILC model, we could envision several scenarios:
- Enhanced user engagement as OpenAI becomes a hotbed for innovation;
- Potential market disruption as user-driven advancements are commoditised swiftly;
- Heightened competitiveness among AI platforms striving to outperform each other in driving and leveraging user innovation.
Recommendations
- Double check the privacy settings. These will change and mature over time.
- Make sure you understand the Terms of Service and Privacy Policy. More here
- OpenAI may eat your Prompts, the best ones will be included under the hood of ChatGPT, to support all the other xxxGPT services that will be released.
- OpenAI may eat your business model
- Most importantly, OpenAI may eat your data. Not really, only the embeddings, it’s not really your data, it’s just a bunch of numbers …….. or is it?
How to protect your business model and data?
Check the following settings within your GPT.
Want to Learn More?
You can read the very large book on Wardley Mapping or you can try using AI to ask questions about the book, this is a small AI app I’ve built, using OpenAI. You can ask any question.
Sources:
- Prompt Engineering Wardley Map, https://onlinewardleymaps.com/#clone:kOzAzisCJpejteQDUB
- Wardley Mapping, https://medium.com/wardleymaps
- Chat with the Wardley Mapping Book
- Learn Wardley Mapping using an AI Assistant
