The Future of Maintenance – Part 10 – ChatGPT Vs Search Engines

To create a better maintained world as we believe at SimpleWays, Maintenance teams need awareness of the technology trends. One of the most talked about technology trends in the last few months was ChatGPT. So, maintenance teams need to be crystal clear about what can ChatGPT do and what it can’t do for them. Only then, they can decide to grasp this as an opportunity or escape it as a distraction.

Diverse capabilities of different team members enrich the possibilities of the team as a whole. Even more, diverse opinions among team members enrich the workplace in a more profound way. That’s because it let them accept new technologies differently. These differences filter the proposed add-ons to the team tools to reach what is best for the whole..

This intro was important because ChatGPT created a disruptive technology wave that affects everyone’s life. So even if you didn’t explore it till now, you at least heard about it. Moreover, you might have created an opinion about it even before trying it. This opinion reflects your general reaction to the rapidly digitalized world. This doesn’t mean that ChatGPT or similar apps are bullet proof or they are the magical solution to every problem. But it really provides a good starting point if you feel lost and don’t know where to start from as far as I explored it.

Also, it is not a bubble that appeared and will disappear with time. At least its robust startup and progress within few months doesn’t imply this. It created a strong start like Uber and Airbnb, or maybe a stronger start. Those disrupting startups still face resistance in some parts of the world. In some places laws were put in place to restrict or even ban their existence.

So, we need to demystify what is ChatGPT? How it works? What is a search engine? How it works?

What is ChatGPT? Why it is different?

I liked the answer of Wikipedia; ChatGPT  is an artificial intelligence chatbot. You can refresh your mind about what is a chatbot and its various types in this chat. The Future of Maintenance – Part 9 – Does Maintenance need a chatbot?. By the way one of the sources for ChatGPT self learning was Wikipedia.

In short a chatbot is a software program that gives answers to its users through some sort of conversation. When we say it is an AI chatbot, then it has some sort of intelligence. That makes it comparable to human intelligence. So, it can understand the question which we call in software world “The Query”. It senses the environment or the context through which this query came. Then, it come up with several possible answers and judge by itself what is the best answer. This judgement is based on the definition of success predefined in its software algorithm. Unlike search engines which provide you with lots of possible responses for your query. Those answers are ranked based on the search engine ranking model i.e. definition of relevancy or correctness or success. Then the user selects what best suits him.

Is providing a single answer good or bad?

So an AI agent returns a single answer to the query very much as most of the humans. When you continue your chat with another human being you can reach to a little or no variation of his original opinion. That’s typically what happens with ChatGPT. If you click on regenerate response a couple of times what will happen? You will see that the new response emphasizes more on one of the points of the previous answer rather than the other ones. Then, after a couple of times it return back to the original answer. Another common point with human responses come in Wikipedia definition for ChatGPT says “ChatGPT has displayed a tendency to confidently provide inaccurate information“.

So what makes AI services feels human is providing the user with a single answer. Sometimes this counts against the freedom of choice. So when an AI chatbot as ChatGPT returns to me a single answer then it can be seen as limiting my alternatives or available options. However, as usual, there is an opposite side of the scale. That’s the side of the “Choice paradox”. Barry Schwartz in his book The Paradox of Choice says:

Learning to choose is hard. Learning to choose well is harder. And learning to choose well in a world of unlimited possibilities is harder still, perhaps too hard.

Barry Schwartz, The Paradox of Choice

Facing too many choices may seem to elevate the freedom of choice. However, psychological studies says that when the number of choices increase, selecting only one of them raises the doubts that this might not be the best choice. This happens even if the choice satisfies the requester needs. Moreover satisfaction decrease specially for those who targets the best choice not only the one that causes personal satisfaction. So, the ChatGPT saves many users from the doubts of the choice paradox even though the answer is not fully correct.

What makes ChatGPT different? And, how does ChatGPT define itself?

What made ChatGPT different that it gained 100 Million users in a couple of months and raised its developing company OpenAI worth to 29 Billion USD was simply its good training!! What we mean by that? let’s carry on.

If you ask ChatGPT about itself or simply: “Who are you?”. It answers:

“I am ChatGPT, an AI language model developed by OpenAI. I’m designed to assist and provide information on a wide range of topics. Think of me as a conversational agent that can engage in text-based discussions and help answer your questions or provide explanations. I’m based on the GPT-3.5 architecture and have been trained on a diverse range of internet text to provide informative and helpful responses. How can I assist you today?“

ChatGPT answer to “Who are you?”

You can receive more in depth technical answer if you ask the ChatGPT about the ChatGPT

ChatGPT is a large language model developed by OpenAI. It is based on the GPT (Generative Pre-trained Transformer) architecture and is specifically designed for generating conversational responses in natural language. ChatGPT is part of the broader GPT family of models, which have made significant advancements in natural language processing and generation.

Using deep learning techniques, ChatGPT has been trained on vast amounts of text data from the internet, allowing it to learn patterns, context, and generate coherent and contextually relevant responses. It leverages the transformer architecture, which enables it to capture long-range dependencies in the input text, making it effective at understanding and generating human-like responses.

ChatGPT can engage in text-based conversations with users, answering questions, providing explanations, giving suggestions, and engaging in interactive dialogue. It can be used for a wide range of applications, such as chatbots, virtual assistants, customer support systems, language translation, content generation, and more.

ChatGPT response for “What is ChatGPT”

ChatGPT Disclaimer

ChatGPT includes the below statement in different variation at the end of its answer to all technical questions that I had tested. It is some sort of removing the legal liability of incorrect answers or incorrectly used answers that draws negative consequences to whomever applied this answer. Or it simply asks you to cross check the answer or consult an expert.

It’s important to note that ChatGPT is an AI model, and while it can generate impressive responses, it has limitations. It may occasionally produce inaccurate or nonsensical information, and it should not be solely relied upon for critical or sensitive tasks. Additionally, OpenAI encourages users to provide feedback on problematic model outputs to help improve the system and address any biases or issues that may arise during interactions with ChatGPT.

OR: If you encounter any difficulties or are unsure about the process, it is recommended to consult the user manual, contact Ingersoll-Rand customer support, or seek assistance from a qualified technician familiar with the equipment.

Or: It’s important to consult the specific user manual, documentation, and software provided by Siemens for the CP 1430 TF module, as the exact steps and procedures may vary based on your application and the software version you are using. Following the manufacturer’s guidelines and recommendations is crucial to ensure a successful and reliable operation of the CP 1430 TF module.

ChatGPT disclaimers

Technologies used in ChatGPT vs those used in Search Engines

As you can see from the replies of the ChatGPT above, ChatGPT uses Large Language Model, deep learning techniques and Generated pre-Trained Transformer. Those technologies are used for developing the AI associated with the chatbot. On the other hand search engines uses crawling, indexing and ranking. But do search engines use AI techniques and ChatGPT uses search engine techniques? And what do each one of them mean in simple language? That’s what we are going to chat about next.

In conclusion,

The increased awareness of the maintenance team leads to a better maintenance process. They improve their performance continuously by picking up what suits them from the new emerging technologies and techniques. On the other hand if they reached a conclusion that this new technique is not fitting their context, they make a wise informed decision about this. This can only happen if we present them with a simple straight forward explanation of those emerging techniques and how they can help maintenance performance improvement.

IIn this chat we came closer to understanding ChatGPT as one of the disrupting technologies that may change how we work and study. We explained what is ChatGPT and how it looks when compared to search engines. Also, we used the choice paradox to explain some psychological backgrounds around how it reached 100 million users in 2 month which never happened before for any app.