10 minutes to read - Apr 18, 2023

Is ChatGPT a Game Changer?

VISIT
Is ChatGPT a Game Changer?
ChatGPT is not simply retrieving information but is generating information, based on rules and relationships.
Table of Contents
1Is ChatGPT a Game Changer?
Impressive but not always accurate
What you put in matters
Assumptions that go too far and inconsistent behavior
The future of ChatGPT and knowledge management

Digital assistants are taking a larger role in digital transformations. They can improve customer service, providing more convenient and efficient ways for customers to interact with the organization. They can also free up human customer service agents by providing quick and accurate responses to customer inquiries and automating routine tasks, which reduces call center volume. They are available 24/7 and can personalize recommendations and content by taking into consideration role, preferences, interests and behaviors. All of these contribute to improved productivity and efficiency. Right now, bots are only valuable in very narrow use cases and are unable to handle complex tasks. However, the field is rapidly changing and advances in algorithms are having a very significant impact.

For example, there has been a great deal of hype and attention around ChatGPT, the chatbot based on the GPT3 language model from OpenAI. GPT stands for Generative Pretrained Transformer and is an AI algorithm that can create new content based on ingesting large amounts of text and data and deriving language rules and relationships. The generation of text in response to an input can be incredibly nuanced and creative, giving the impression that one is conversing with a human.  

Unlike a search engine, it is not simply retrieving information but is generating information, based on rules and relationships derived through enormous amounts of data that the algorithm has processed.

But wait, let me have ChatGPT explain:

Seth: What is Generative AI?

ChatGPT: Generative AI refers to a type of machine learning model that is capable of generating new, original content based on a given set of input data. This can be in the form of text, images, or other types of data. Generative AI models are trained on a large dataset of examples, and they use this training data to learn the patterns and relationships that are characteristic of the data. They can then use this learned knowledge to generate new content that is similar to the training data, but which is not simply a copy of the training data.

It is pretty amazing.

Impressive but not always accurate

However, it frequently provides incorrect responses that are plausible. It also sometimes cannot find information that is easily retrievable through a web search and may generate different answers to the exact same question. For a creative task such as copywriting that is a good thing. However, in an enterprise context, consistency in retrieving information is critical. The point is that as useful as it may be, generative AI will not be the complete answer to powering enterprise digital assistants and customer service bots.

To illustrate the inconsistency and understand how the system uses context, I asked ChatGPT about myself. (Who doesn’t google oneself from time to time?) 

Seth: Who is Seth Earley?

ChatGPT: I'm sorry, but I don't have any information about a person named Seth Earley. Can you provide any additional context or details about the person you are looking for?

And my company...

Seth: Who is Earley Information Science?

ChatGPT: I'm sorry, but I do not have information on a company or individual named "Earley Information Science." Could you provide more context or clarify your question?

How disappointing! I was crushed!

What you put in matters

However, when I provided information about the company and asked for an elevator pitch, something interesting happened. The system first returned a good pitch:

Seth: Earley Information Science is a firm that specializes in product data, content operations, knowledge management and customer data. The web site is www.earley.com. Can you write an elevator pitch for the company’s services?

ChatGPT: Earley Information Science is a leading provider of product data, content operations, knowledge management, and customer data services. With a team of experienced professionals and a comprehensive suite of solutions, Earley helps organizations optimize their product and customer experiences by improving the quality, consistency, and accessibility of their data. Whether you need to streamline your product content creation process, enhance the accuracy of your customer data, or build a scalable knowledge management system, Earley has the expertise and tools to help you achieve your goals. With a focus on innovation and collaboration, Earley is committed to delivering value and driving business success for its clients.

Great response. Thanks, ChatGPT – I might use that one.

When I then asked who Earley Information Science was, the response included my role in the company. (Which I had not provided.)

Seth: Who is Earley Information Science?

ChatGPT: Earley Information Science is a consulting firm that provides information management and technology services to organizations. The company was founded by Seth Earley, who is a thought leader and expert in the fields of information management, knowledge management, and enterprise search. Earley Information Science helps organizations to optimize their information management systems, improve knowledge sharing and collaboration, and increase productivity through the use of advanced technologies and best practices. The company’s services include strategy development, system design and implementation, and training and support.

The system appeared to have retained context from the prior query, but it produced much more detail than what I had given it.

When I repeated the query about my name, this time the system appears to have used information in prior chats to inform its response.

Seth: Who is Seth Earley?

ChatGPT: Seth Earley is the founder and CEO of Earley Information Science, a company that specializes in… etc.

Interesting indeed.

Assumptions that go too far and inconsistent behavior

In a new chat session, asking the same question produced another version that had significant inaccuracies. (Bold added for emphasis.)

Seth: Who is Seth Earley?

ChatGPT: Seth Earley is a recognized expert in the field of information management and knowledge management. He is the founder and CEO of Earley Information Science, a conslting firm that specializes in helping organizations create, manage, and leverage their knowledge and information assets.

Earley is a well-known speaker and author in the field of knowledge management, and has written several books on the subject, including "The New Knowledge Management: Complexity, Learning, and Sustainable Innovation" and "The Knowledge Management Toolkit: Practical Techniques for Building a Knowledge Management System." He is also a frequent contributor to industry publications and has been quoted in a variety of media outlets on topics related to knowledge management and information management.

You went a little too far there, ChatGPT. I contributed to "The Knowledge Management Toolkit," but I did not write it. And I had nothing to do with the other book mentioned, “The New Knowledge Management."

During this session, I did not provide any information about me or my firm. This time the system was able to recognize me without more context.

But, not so fast. In another chat session, the same query once again stumped the system.

Logging out and back in produced this result:

Seth: Who is Seth Earley?

ChatGPT: I'm sorry, but I don't have information on a person by that name. It's possible that you have the wrong name, or that the person you are asking about is not a well-known figure. Without more context, it is difficult for me to provide any additional information.

But now asking about the company produced this response:

Seth: Who is Earley Information Science?

ChatGPT: Earley Information Science is a company that provides consulting and technology services in the field of information management, etc. … but then it added: It is not clear from the information available online who the founder or owners of Earley Information Science are.

I had not asked about the founder, but OK. But then when I asked about myself again, this time it knew I was the founder.

Seth: Who is Seth Earley?

ChatGPT: Based on my research, Seth Earley is the founder and CEO of Earley Information Science, a company that provides consulting and technology services in the field of information management etc.….

…He has written several books on these subjects, including "The ABCs of KM" and "The Digital Workplace."…

These details are not true. I did not write "The ABCs of KM" and "The Digital Workplace." I have written one book (The AI Powered Enterprise) and coauthored another (Practical Knowledge Management) but thanks for the undeserved kudos, ChatGPT.

When I logged out and back in the behavior changed again.

Seth: Who is Earley Information Science?

ChatGPT: I'm sorry, but I am not able to find any information about a company or person named "Earley Information Science." …. Etc.

This was the same response that I had gotten in the prior session before I had provided some company information. When I asked about Seth Earley again, the system recognized me and provided a bio. A further inquiry about Earley Information Science produced a different description and a statement that it was not clear if the company was still in operation (we are).

One last experiment produced the oddest set of incorrect details, listing books I had not written and a teaching position I do not hold.

The future of ChatGPT and knowledge management

Some people believe that the power of ChatGPT is going to eliminate the need for knowledge management approaches and principles. When I put that question to the system, it responded with a logical, meaningful answer:

ChatGPT: Knowledge management, knowledge architecture, and knowledge engineering play important roles in the design of chatbots.

Knowledge management involves the processes and technologies used to capture, store, and share knowledge within an organization. In the context of chatbot design, knowledge management can help to ensure that the chatbot has access to a wide range of relevant information and can provide accurate and useful responses to user queries.

Knowledge architecture refers to the structure and organization of knowledge within an organization. In the context of chatbot design, knowledge architecture can help to ensure that the chatbot's knowledge base is well organized and easy to navigate, which can improve the user experience and the effectiveness of the chatbot.

Knowledge engineering involves the process of designing and building knowledge-based systems, such as chatbots. It involves identifying the knowledge and expertise needed for the chatbot to function effectively and designing the chatbot's knowledge base and information architecture to support these needs.

Overall, knowledge management, knowledge architecture, and knowledge engineering are important considerations in the design of chatbots because they help to ensure that the chatbot has access to the knowledge and expertise it needs to provide accurate and useful responses to user queries.

To summarize, sometimes the answers from ChatGPT were insightful and articulate, while other times they were inconsistent or outright wrong. Next time, I will delve into more detail abot the value of generalized language models, whether ChatGPT can replace knowledge management, and how ChatGPT relates to knowledge management tools such as ontologies and knowledge graphs.

Article source
Author: Seth Earley , Earley Information Science
Seth Earley is an expert with 20+ years experience in knowledge strategy, information architecture, search-based applications and information findability solutions.
loading...