By Dr. Martin Block, Professor Emeritus, Northwestern University, Retail Analytics Council, Ronald Lunde, former senior executive in the grocery and advertising industries, Gary Drenik and Phil Rist, Prosper Insights and Analytics
The digitally connected global consumer challenges leader and management teams to focus on nonlinear interactions, to see their future market and non-market models as complex adaptive systems, not as historic equilibrium systems. Digitally enabled consumer strategies need a constant juggling of multiple theories, evolutionary technologies, strategies, and models. Result? The winners take most market outcomes.
The next wave of extraordinarily impactful and beneficial technological change is about to be delivered…a future enabled by ‘4 Big Thing’ concepts — Artificial Intelligence (AI); Artificial General Intelligence (AGI); General Adversarial Networks (GANs) and … Chat BOTs (ChatGPT).
You may think that all this is science fiction today—but tomorrow?
Four Big Things
- AI excels at completing specific tasks or types of problems.
- AGI has higher capabilities than AI, theoretically approaching the ability to perform any cognitive task that a human can. AGI is, however, still a developing technology.
- GANs consist of two ‘opposing’ neural networks. One is a generator that generates new data instances, while the other, a discriminator that evaluates the data for authenticity and accepts or rejects the input.
- ChatGPT interacts with humans in a Natural Language (NLP) conversational way. ChatGPT can admit its mistakes, challenge incorrect premises, reject inappropriate requests, and answer follow-up questions.
Recognize that the above ‘Four Big Things’ are inter-related and inter-dependent. They do, therefore, require a clear use strategy; infrastructure investment; rigorous employee training and the right data.
One More Big Thing
What’s next after what’s next? Evolving Neuromorphic Computing: Scientists have tried to duplicate human reasoning and perception (AGI) for decades. The breakthrough might well be Neuromorphic Processors which achieve roughly 10 trillion operations per second. The high-performance architecture and functionality used in neuromorphic computers is significantly different from traditional Von Neuman architecture. AI, machine learning, neural network architecture used in consumer and enterprise technology will be theoretically and exponentially changed. For example, Neuromorphic Processors might diagnose diseases at first stages by analyzing the contents of exhaled breath.
Chat BOTs
Our friends and colleagues at Prosper Technology (Gary Drenik and Phil Rist) gave OpenAI’s ChatGPT product a test drive. ChatGPT demonstrates AI, and a smattering of GANs and AGI. We think you will find their results on the following pages prescient, interesting and informative.
Last Big Thing
Greylock partner and Silicon Valley legend Reid Hoffman recently said, “Let’s have people realize that everything that you do as a profession is going to have an [AI] copilot.” “…whether or not you’re a journalist, whether you’re an investor, an academic, a lawyer, a doctor, there’s going to be a copilot within maximum five years, maybe two.” Six Questions for a Chat Bot:
QUESTION:
1. Please Tell Me What A Chat Bot Is.
ANSWER: A chatbot is a computer program designed to simulate conversation with human users, especially over the Internet. Chatbots are often used in customer service and online shopping to provide quick and automated assistance to customers. They can be programmed to respond to specific commands or to use artificial intelligence to understand and respond to more complex inquiries. Some chatbots are designed to be used on messaging platforms, while others are integrated into websites or mobile apps. Chatbots can be useful for businesses because they can handle a large volume of interactions with customers, freeing up time for human customer service agents to handle more complex or high-priority tasks.
QUESTION:
2. What Types Of Inquires Can Chat Bots Respond?
ANSWER: Chatbots can be programmed to handle a wide variety of inquiries, depending on their design and the specific business or service they are intended to support. Some common types of inquiries that chatbots can handle include:
- Basic information: Chatbots can be programmed to provide basic information about a business or service, such as store hours, location, and contact information.
- Customer service: Chatbots can be used to handle common customer service inquiries, such as answering questions about products or services, handling returns and exchanges, and providing assistance with account issues.
- Order processing: Chatbots can be used to process orders and handle payment transactions.
- Appointments and reservations: Chatbots can be used to schedule appointments or make reservations for services or events.
- Troubleshooting: Chatbots can be programmed to troubleshoot and diagnose problems with products or services.
- Personalized recommendations: Chatbots can use artificial intelligence to provide personalized recommendations to customers based on their previous purchases or interactions with the business.
Overall, chatbots can be programmed to handle a wide range of inquiries and interactions with customers, depending on the specific needs of the business or service.
QUESTION
3. How Do Chat Bots Differ From Search Engines? Please Answer In Two Paragraphs.
ANSWER: There are a few key differences between chatbots and search engines:
- Purpose: Chatbots are designed to simulate conversation and provide assistance to users, while search engines are designed to help users find information on the Internet. Chatbots are typically used to provide assistance within a specific context or domain, such as a business or service, while search engines are designed to provide information on a wide range of topics.
- Input and output: Chatbots typically accept input in the form of natural language text or voice commands, and provide output in the form of text or voice responses. Search engines, on the other hand, accept input in the form of keywords or phrases and provide output in the form of a list of relevant web pages or documents.
- Context and personalization: Chatbots can use artificial intelligence to understand the context and intent of user inquiries and provide personalized responses, while search engines typically provide the same results for a given query regardless of the user or context.
Overall, chatbots and search engines are different tools that serve different purposes. Chatbots are designed to provide assistance and support within a specific context, while search engines are designed to help users find information on a wide range of topics.
QUESTION
4. How Important Is It To Input Accurate And Representative Data For A Chatbot To Analyze?
ANSWER: It is very important to input accurate and representative data for a chatbot to analyze, as this can significantly impact the performance and effectiveness of the chatbot. Chatbots rely on data to learn and understand the context and intent of user inquiries, and to generate appropriate responses. If the data used to train a chatbot is inaccurate or unrepresentative, the chatbot may have difficulty understanding user inquiries or providing appropriate responses. This can lead to poor user experiences and may even cause users to lose confidence in the chatbot.
To ensure that a chatbot is able to accurately analyze and respond to user inquiries, it is important to input a diverse and representative set of data that accurately reflects the types of inquiries and interactions that the chatbot will encounter in practice. This may include training data that covers a wide range of topics and languages, as well as data that reflects different user demographics and contexts. By inputting accurate and representative data, businesses can help ensure that their chatbots are able to provide high-quality assistance and support to users.
QUESTION
5. Does GIGO Impact The Analysis Of A Chat Bot?
ANSWER: Yes, the principle of “garbage in, garbage out” (GIGO) can definitely impact the analysis of a chatbot. GIGO refers to the idea that if inaccurate or poor-quality data is input into a system, the output of that system will also be of poor quality. This principle applies to chatbots as well as other types of computer systems.
If a chatbot is trained on inaccurate or unrepresentative data, it may have difficulty understanding user inquiries or providing appropriate responses. This can lead to poor user experiences and may even cause users to lose confidence in the chatbot. To ensure that a chatbot is able to accurately analyze and respond to user inquiries, it is important to input a diverse and representative set of data that accurately reflects the types of inquiries and interactions that the chatbot will encounter in practice. By following the principle of GIGO, businesses can help ensure that their chatbots are able to provide high-quality assistance and support to users.
QUESTION
6. Can Chatbots Analyze Data And Make Recommendations?
ANSWER: Yes, chatbots can be programmed to analyze data and make recommendations. Chatbots can use artificial intelligence (AI) techniques, such as machine learning, to analyze data and make predictions or recommendations based on that data.
For example, a chatbot designed for online shopping could analyze a customer’s past purchases and browse history to recommend products that the customer might be interested in. A chatbot designed for customer service could analyze customer inquiries and interactions to identify patterns or trends, and make recommendations for how to improve the customer experience.
Overall, chatbots can be very useful for analyzing data and making recommendations, as they can handle a large volume of interactions with users and quickly identify patterns or trends that may not be immediately apparent to human analysts. By leveraging the data analysis capabilities of chatbots, businesses can gain insights into customer behavior and preferences, and use those insights to improve their products, services, and overall customer experience.
Continue reading.