The ability of ChatGPT to integrate knowledge from various and diverse knowledge bases and generate human-like language allows innovation teams to increase the quantity, quality, and diversity of the opportunities they identify. This post provides insights and examples on how to integrate this groundbreaking tool into your opportunity discovery process.

Blurring the line between internal and external knowledge

Trained on a large set of data in an endless range of topics, ChatGPT opens up the opportunity for innovation teams to integrate knowledge that might lay outside their area of expertise, thus enabling them to explore larger problem and solutions spaces. The model’s ability of understanding, generating, and adapting language makes it even more powerful and easy to apply.

Thus, ChatGPT can be a great addition to your toolkit as you apply Worksheet 1 of the Market Opportunity Navigator for generating your Market Opportunity Set. It can provide a broad view on where your innovation can be applied, who might need it, and what is the value that different market segments can gain from it.

Applications for technologies

To understand the value of this model in the identification process, we first asked:

How can ChatGPT support the identification of applications for technologies?

The reply we received highlights the key aspects of this search process: understanding the technology’s core capabilities, learning from existing applications, and fostering discussions between experts from different fields. Here is the model’s reply (bolding done manually):

There are several ways that ChatGPT could potentially support the identification of applications for technologies:

1. ChatGPT could be trained on data about different technologies and their potential applications in order to identify relevant applications based on user-provided descriptions of the technologies.
2. ChatGPT could be used to generate ideas for potential applications of a technology by asking questions about the technology and its capabilities, and using its language generation capabilities to suggest possible uses.
3. ChatGPT could be used to help research and gather information about existing applications of a technology, by asking relevant questions and providing summaries of the information it finds.
4. ChatGPT could be used to facilitate discussions between experts in different fields about the potential applications of a technology, by asking questions and providing relevant information to facilitate the exchange of ideas.

Overall, ChatGPT’s ability to understand and generate human-like language could make it a useful tool for supporting the identification of applications for technologies.

Example: Applications and customers for ChatGPT

To demonstrate this ability, we asked the model to identify applications stemming from its own innovative technology and list the customers who will find these valuable. We asked:

Where can ChatGPT be applied? What type of customers might need it?

The reply provides some initial ideas for applications and users (bolding done manually):

GPT-3 is a language generation model developed by OpenAI that is designed to generate human-like text. It could potentially be applied in a wide range of contexts where generating natural language text or content would be useful, such as:

1. Chatbots and customer service: GPT-3 could be used to power chatbots that provide customer service or assist with tasks such as booking appointments or answering FAQs.
2. Content creation: GPT-3 could be used to generate articles, social media posts, or other written content in a variety of styles and formats.
3. Translation: GPT-3 could be used to translate text from one language to another, potentially improving the accuracy and fluency of machine translation.
4. Data annotation: GPT-3 could be used to generate descriptions or labels for data sets, making it easier for people or algorithms to understand and analyze the data.

Customers who might need GPT-3 could include businesses that want to automate customer service or content creation tasks, organizations that need to translate large volumes of text, or researchers who are working with data sets and need to label or annotate the data.

You can dig deeper to understand how your invention can address the needs of a specific customer segment. To demonstrate this, we chose to focus on a group of users that were mentioned above (academic researchers) to learn more about the jobs that ChatGPT can do for them. We asked:

How can ChatGPT bring value to academic researchers?

The answer offers additional use cases which were not mentioned in the first prompt. Also pay attention to the last sentence – indicating that we need to verify and double-check the model’s suggestions. Here is the reply (bolding done manually):

There are several ways in which chatbot GPT-3 could potentially bring value to academic researchers:

1. Data annotation: GPT-3 could be used to generate descriptions or labels for data sets, making it easier for researchers to understand and analyze the data.
2. Text generation: GPT-3 could be used to generate large amounts of synthetic text data for use in natural language processing research.
3. Text translation: GPT-3 could be used to translate research papers or other texts from one language to another, potentially making it easier for researchers in different parts of the world to collaborate and share ideas.
4. Text summarization: GPT-3 could be used to generate summaries of research papers or other texts, making it easier for researchers to quickly scan and understand large amounts of information.

It’s important to note that GPT-3 is a language model and does not have access to real-world information, so its suggestions should be taken with a grain of salt and may require further research or verification.

Using the power of ChatGPT for the Flyability case

We often use the Flyabilty case to show the value of the wide-lens perspective provided by the Market Opportunity Navigator. This company developed an innovative drone that can fly to inaccessible places due to its unique protective cage, and applied Where to Play to figure out where to start building their business (you can read more about this case here).

To demonstrate how ChatGPT can help innovators in the identifications of market opportunities, we asked the following question:

We developed a new drone with a camera and a protective cage around it. It can fly or roll in complex and confined spaces or next to humans. Who might need this drone?

If you teach or train others how to apply the Market Opportunity Navigator, show them the reply below (or run this exercise live with them). In a split of a second, we were introduced to seven different target markets for the company. Of course, some of the suggestions are not completely accurate (and there are others which were not mentioned), but it’s definitely a great springboard for the discovery process.  Here is the reply (bolding done manually):

There are potentially many customers who might need a drone with the capabilities you described. Here are a few examples:

1. Industrial and commercial inspections: Your drone could be used by companies to inspect hard-to-reach or hazardous areas, such as the inside of a pipe or a building under construction.
2. Search and rescue: Your drone could be used by rescue teams to search for and locate people in difficult or confined spaces, such as collapsed buildings or dense forests.
3. Law enforcement: Your drone could be used by police or other law enforcement agencies to search for suspects or gather intelligence in difficult or confined spaces, such as urban environments or buildings.
4. Military: Your drone could be used by the military for a variety of purposes, such as intelligence gathering, surveillance, or search and rescue operations in complex or confined spaces.
5. Agriculture: Your drone could be used by farmers or other agricultural workers to inspect crops or livestock in confined or hard-to-reach areas.
6. Film and television production: Your drone could be used by film and television production companies to capture footage in confined or hard-to-reach areas, such as inside a building or in a dense forest.
7. Real estate: Your drone could be used by real estate companies to capture footage of properties or land for sale, including hard-to-reach or confined areas.

These are just a few examples, and there may be many other potential customers for your drone depending on its specific features and capabilities.

Concluding thought: The concept of ‘Hybrid Intelligence’

Chat GPT allows innovation teams to quickly and easily tap into a larger pool of knowledge than was previously possible. However, transformer-based language models are not without limitations: they are not completely accurate and the data presented might simply be wrong. At the moment, ChatGPT provides great value in tasks that require a more general perspective (such as applications for technologies), but should be used cautiously in tasks that require accurate data such as the evaluation of market opportunities (e.g., market size or competitive landscape).

These limitations highlight that such technology should not be trusted and used blindly. Indeed, a recently published paper in the Journal of Product Innovation Management proposes that AI models will best support innovation practices in the form of hybrid intelligence, i.e. the combination of human and artificial intelligence: “By providing the language model with their own knowledge and ideas through natural language, innovation teams can integrate the AI into their existing processes, as if it would be a new colleague”.

Integrating this remarkable tool can thus greatly improve the productivity of innovation practices in general, and the identification of opportunities in particular.

Learn more about it in our upcoming certificate program.