“Artificial Intelligence” (AI) was created as a concept in 1956, about 67 years ago, and since then been a popular wonder driving sci-fi concepts, from the famous HAL to the most recent M3GAN. Of course, in many AI sci-fi flicks, the AI becomes sentient, determines it is smarter than its human creators, and then develops a plan to subjugate the human race. Raised on the sensationalism of this concept, it’s not surprising that there has become an infatuation with AI since November’s ChatGPT research project publicity. From religious leaders to politicians, business to schools, technologists to pundits, everyone is wondering, and often worried, about the future impact of ChatGPT and AI more generally.
I believe that there is a perspective—maybe the most critical one—that is not well documented, and as a result is causing most people to miss the bigger picture of what AI means, both now, and in the future, to our society.
That perspective is simply that AI that is the latest general-purpose technology (GPT), and therefore, it will fundamentally alter how innovation occurs in society, and correspondingly will increase worker productivity. As with all prior GPTs created, this will dramatically change our societies, economies, and government systems. Astute observers are therefore encouraged to understand this in more detail, to be best prepared for what is likely to come.
The concept of “general-purpose technology” is not well known. It was created in 1992 by Bresnahan and Trajtenberg, Stanford, and Tel Aviv University economists, who identified that periods of significantly increased economic growth were correlated with inventions of new technologies that became ubiquitous and supported many different purposes. Those technologies demarcated new eras in innovation, as they became pervasive to be used across a variety of industries and dynamic to support many use cases, which increased work productivity. A classic example is the steam engine, which made everything from transportation to manufacturing more efficient. There are now 24 of these GPTs generally accepted, including the more recent inventions of the computer and the internet.
What Bresnahan and Trajtenberg identified, is that with the adoption of each GPT, the productivity of research and development activities increases, and correspondingly economic productivity increases as the new GPT is adopted broadly.
I believe AI is our 25th GPT, and as a result, should be noted as the demarcation of a new era in innovation. Further, those that harness the power of AI will highly benefit, and those that control AI will be those with the most dramatic economic and political power in the history of the world.
Understanding this requires a short discussion on innovation itself. “Innovation” is a highly misunderstood term, as it is often applied to everything from incremental improvements in everything from toilet paper to vehicles, and disruptive improvements that change markets. It can be used as a noun, verb, or adjective, and as a result, highly confuses people.
I believe the best definition of innovation is this: the process of creating sustainable value from something new. Critically, the result of innovation needs to be net value creation—something more valuable than existed before innovation was conducted.
With this definition, it is clear to see that the purpose of business is in fact innovation – to create sustainable value for its shareholders. It can be argued it is also the purpose of government—to create sustainable value for its citizens. It has even been theorized that innovation is the primary cause of social justice and economic prosperity in the world, and as such, the most critical process one should understand for future success.
Innovation occurs through a process of “work” (often referred to as research and development) where someone uses their ingenuity to do something in a new way that creates future value. That may be fitting two pieces of a widget together in a unique way for the first time, which allows the item to be sold as a new product in the marketplace. It may be identifying new and less costly ways to complete a process, or creating new business strategies that are more profitable or purpose-fit for the economy. In any situation, “work” occurs, which creates innovation.
It's well known that computers can do work that humans have previously done—and that provided our last major technological and innovation era. How AI is different, and how it augments past computer capability, is that AI automates the work of innovation.
In 2018 Cockburn, Henderson, and Stern identified this concept as the IMI, or “invention of a method of innovation”, and when invented would lead to the replacement of traditional research and development in favor of automated methods of innovation and ultimately, automated value creation. As a result, they said: “the potential commercial rewards from mastering this mode of research are likely to usher in a period of racing, driven by powerful incentives for individual companies to acquire and control critical large datasets and application-specific algorithms.” This is why Microsoft just spent $10 Billion to purchase a minority stake of OpenAI, the makers of ChatGPT. They have recognized the race, and are joining themselves to OpenAI as their horse. Similarly, Google has DeepMind, which it recently claimed would release a private beta in 2023. Interestingly, Amazon has not publicly stated a branded general AI, although has significant use of machine learning internally and has a relationship with Stability AI.
In what may or may not be a coincidence, ChatGPT does not stand for chat general-purpose technology, but instead Chat “General Pre-learning Transform”. That describes its technical function: it is a method of general “chatting” (text-based interaction) based upon data it has pre-learned. Although there are many interesting aspects to this, for this article, the most critical is that ChatGPT can work across a huge variety of scenarios, to simplify and accelerate the process of work, inclusive of conducting research, doing analysis, and presenting the results, in a dramatically more efficient way than humans can do likewise. In other words, ChatGPT meets the requirements of the general-purpose technology test.
Now, the question is, is it also an IMI? I think the answer is also similar, yes. Take the simple use case of writing a school essay: ChatGPT can do so in moments (if the subject material is in its dataset). If you were a student, it can save hours of work, creating value. Similarly, for businesses, it can do historical analysis of data sets that would be highly useful for decision-making, identify sales leads, write marketing copy, and more. Again, it will do this dramatically faster than humans, which means that it is creating value for humans through productivity savings.
To see whether I felt ChatGPT could be an IMI, I set about a short exercise to see whether it could help with innovation activities. What I often find in my study and practice of innovation, is that the most common problem people have in discussing it is to have a common definition and framework. To address this, I often use the Doblin 10 Types of Innovation framework to help identify innovation opportunities for my clients, and I elected to do the same with ChatGPT.
As a former Amazon leader, and considering the massive layoffs being conducted recently, I decided to use big tech as an example, to see whether ChatGPT could “out-innovate” our best innovators.
And I think it did—at least with identifying innovation opportunities. Here are some examples prompts, and their results.
Prompt: give 10 real world ideas for how Amazon could apply the Doblin 10 Types of Innovation to be more profitable.
1. Product innovation: Amazon could develop a new line of smart home devices, such as a smart thermostat or security camera, to increase its offerings in the connected home market. 2. Process innovation: Amazon could explore the use of blockchain technology to improve supply chain transparency and traceability, reducing costs and increasing efficiency. 3. Position innovation: Amazon could expand its services to include same-day or even same-hour delivery in urban areas to increase customer satisfaction and attract new customers. 4. Paradigm innovation: Amazon could introduce a subscription-based service for access to exclusive deals and discounts, creating a new revenue stream. 5. Performance innovation: Amazon could invest in developing new technologies, such as artificial intelligence, to improve the accuracy of product recommendations and increase sales. 6. Package innovation: Amazon could offer bundled packages of products and services, such as a Prime membership with access to streaming services and discounts on groceries. 7. People innovation: Amazon could invest in employee training programs to improve the skills and productivity of its workforce. 8. Portfolio innovation: Amazon could expand into new industries, such as renewable energy, to diversify its revenue streams and reduce risk. 9. Place innovation: Amazon could open physical stores in high-traffic locations to increase brand visibility and drive sales. 10. Plan innovation: Amazon could develop new marketing and advertising strategies, such as influencer partnerships or targeted social media campaigns, to reach new customers and increase brand awareness.
If you are familiar with Amazon, this may not be very groundbreaking, as it is already doing many of those things, and that data is likely within ChatGPT’s dataset. But, consider this for a small or medium-sized business. A leader can easily get multiple ideas of how they can increase their innovation, and many of these may be more effective (and less time-consuming) than creating the ideas on their own.
Another interesting test case is to compare approaches to innovation between companies, and begin to ask about more specific scenarios:
Prompt: What are the most likely to be profitable within 10 year innovations that both Amazon Web Services and Microsoft Azure have not yet entered, according to each of the Doblin 10 Types of Innovation Model types
Response: According to the Doblin 10 innovation model, there are ten types of innovation that a company can pursue. Here are some potential innovations that Amazon Web Services (AWS) and Microsoft Azure have not yet entered, that are likely to be profitable within the next 10 years, based on each of the ten types:
1. Disruptive innovation: Both AWS and Azure could enter the field of serverless computing, offering platforms that allow customers to run their applications without the need to provision or manage servers. 2. Incremental innovation: Both AWS and Azure could develop new services that allow customers to more easily manage and optimize the performance of their applications, such as auto-scaling or traffic management. 3. Architectural innovation: Both AWS and Azure could develop new infrastructure or services that allow customers to more easily integrate their applications with other technologies, such as IoT devices or blockchain networks. 4. Radical innovation: Both AWS and Azure could explore the use of new technologies such as quantum computing, or neuromorphic computing which can speed up certain workloads. 5. Modular innovation: Both AWS and Azure could develop a platform that allows customers to easily compose and deploy complex applications using a variety of pre-built, modular components. 6. Component innovation: Both AWS and Azure could develop new technologies or features that can be integrated into their platform to enhance the user experience, such as machine learning or natural language processing. 7. Process innovation: Both AWS and Azure could streamline the process of developing, deploying and managing applications on their platform, making it more efficient and user-friendly. 8. Business model innovation: Both AWS and Azure could develop a new business model for their services, such as a pay-per-use or a subscription-based model. 9. Market niche innovation: Both AWS and Azure could target specific niche markets with their services, such as healthcare or financial services. 10. Network innovation: Both AWS and Azure could leverage their existing network of customers and partners to build new services that are tightly integrated with other products and services.
Now, although this is not completely following Doblin's model, I think this is getting better and I think there are some real ideas here. Keep in mind these are only examples, and each could likely be refined much more (and there are many other parts of the Doblin 10 Types of Innovation model to explore).
Finally, what about one company setting out to disrupt another?
Prompt: what are the most likely to succeed innovations that Amazon could use to disrupt Facebook, according to each of the Doblin 10 Types of Innovation Model types
According to the Doblin 10 Types of Innovation model, the following are some potential areas where Amazon could potentially disrupt Facebook:
1. Progression: Amazon could potentially use its vast data and analytics capabilities to create a more personalized and dynamic social media experience for users. 2. Disruption: Amazon could potentially launch a low-cost, ad-free social media platform that appeals to users who are tired of the current ad-heavy social media landscape. 3. Business model: Amazon could potentially leverage its existing e-commerce capabilities to create a social media platform that seamlessly integrates online shopping and social networking. 4. User experience: Amazon could potentially create a social media platform that utilizes virtual and augmented reality to enhance the user experience. 5. Service: Amazon could potentially create a social media platform that offers a variety of services, such as job search, financial advice, and mental health resources. 6. Platform: Amazon could potentially create a social media platform that allows for easy integration with other services and platforms, making it a hub for all of a user's online activities. 7. Network: Amazon could potentially leverage its existing distribution network to create a social media platform that is easily accessible to users in remote or underserved areas. 8. Ecosystem: Amazon could potentially create a social media platform that is integrated with its existing ecosystem of services, such as Prime, Alexa, and AWS, to create a seamless and cohesive user experience. 9. Channel: Amazon could potentially use its existing reach and influence to create a social media platform that is easily discoverable and accessible to users. 10. Customer engagement: Amazon could potentially create a social media platform that encourages and rewards customer engagement and loyalty.
From a business and innovation perspective, I think it’s clear that ChatGPT and AI can easily support innovative ideation and identification of market opportunities. As has been shown in other research, it can also do everything from writing code to creating marketing copy. In each of these situations, it can perform work much faster than humans, and as shown, can identify areas of innovation. Although it cannot yet create complex innovation, I believe that it will occur soon, and I do consider this to be the first IMI. That’s an astounding feat that needs to be recognized.
So what does this mean for a business person or technologist? It’s simple: a new innovation era, driven by AI, has arrived. Further, we are at a highly unique—astounding even—inflection point, where AI is becoming both a GPT and IMI, which has not occurred together before in the history of the world. Based on this fact, I believe that the pace of innovation is about to dramatically increase—even faster than it has the last several years—and you need to prepare yourself if you want to be successful.
People and companies who succeed in this new innovation era will be those who understand and build practices to harness AI to drive innovation. Next will be those who use AI products to drive efficiency in their work. Finally, will be the laggards, who do not utilize AI nor create an innovation competence, and will be unable to compete with others as the cost of production decreases, and cannot keep up with innovation advances. Do not be in that last group.
There are myriad other aspects to this, from societal to political impacts, where we need regulation and what type, how we separate the owners of the data models from the AI owners, the maturity of AI, etc. I may address those in future articles.
Finally, I will close with a quote and a question for you to ponder. My favorite quote is from Steve Jobs: “Innovation is the only way to win.”
And my question for you, is that if it’s true that AI is going to enable a dramatic increase in the speed of innovation, what are you doing now to create an edge over your competitors, and win?
Author Bio: Trent Gillespie is the Founder and Chief Innovation Officer for Day One Innovation. He is a 20+ year technologist, who spent almost 9 years at Amazon in senior leadership managing global growth initiatives, large innovation and technical portfolios, and defined and provided oversight of Alexa’s Privacy capabilities. Trent is now a full-time innovation speaker, advisor, and practitioner, inspiring innovation improvement in organizations from start-ups to enterprise.