episode 1 - This is the series of post we are creating on usability, applicability and technical aspects by using ChatGPT, Prompting and Prompt Engineering.
Before, I move further and express my thoughts on ChatGPT, let me go back to around 10 years back when we used to have NLP (natural language processing) and ML (machine learning) based implementations.
Nowadays, we often hear about the Generative AI, Prompt Engineering etc. If I can recall previous days of working, we also used to have Prompting based development and implementation in the software product development.
What is Prompting and Prompt Engineering?
Prompting and Prompt Engineering are related concepts in the field of NLP - natural language processing and specifically in the context of the language models like GPT.
What is Prompting - it refers to the usability as proving a specific instruction / query to a language model to guide and receive response that involves direct query.
For example - Prompting is commonly used in the chatbot applications where it refers to the specific instruction on questions to guide the conversation. A Product management people can use prompting at various stages of a conversation which is happening between a user and bot. I am sharing example of a bot from Tanishq. You might have experienced bots and various process enabled optimisations in your enterprise as well as day-to-days product usage.
What is Prompt Engineering - It involves the systematic design and refinement of prompts to achieve desired behavior from the language model. Prompt engineering may involve various applicability and techniques, such as word modification, constraints definition, adding constraints which are used to shape the output.
What is ChatGPT?
ChatGPT is a AI language model developed by OpenAI. It is based on the GPT [Generative Pre-Trained Transformer] architecture. specifically GPT 3.5. It is designed to understand and generate human-like text responses based on the input it receives. ChatGPT is trained on a vast data of text from the internet, that assist it to generate relevant responses to a wide range of prompts and queries.
In a simple word, ChatGPT is somewhat similar to a virtual friend who get engaged with you on your query or prompt with large amount of internet data which is summarised to a possible relevant response - Anubhav
Disclosure - to make it more relevant in the context of ChatGPT usability interaction, few or more sections, data / screen shots are taken from the ChatGPT.
Let's take a look into the prompt query done with - Question - How can I use ChatGPT in product management and we can see response received:
Eventually, when asked, what more we can do using ChatGPT in the product space and herein, we can see the response:
Gradually, after reading these information above, I went further to create more search towards competitive analysis - How can I use ChatGPT in competitive analysis and as below I found:
I would like NOT to miss the call-out by ChatGPT itself -
"Remember that while ChatGPT can provide valuable insights, it should be used in conjunction with human analysis and validation. It's important to critically evaluate the generated response, cross-reference information with reliable sources, and consider additional factors such as market dynamics and customer feedback for a comprehensive competitive analysis."
Also, I tried to get more information about the purchasing and buying tend for the BSE that will give me 5x returns and herein, I got the response from ChatGPT:
Question asked:
- Which stock should I Purchase in India BSE that will give 5x return
- Which stock can give me better returns in BSE
Conclusion
I am sharing these information and inputs based on my experience:
Do not be so much involved and dependent only on the ChatGPT, and/or similar forms of Generative AI.
Consider these spaces as a area to gather more information in quick one click, moreover, do not miss to bring your own understanding or consensus Data-Driven / Evidence based Decision Making, consider this as tool only that can be used by product people as a tool that product managers can utilized
Revalidate your data as human-based Cognitive evaluation approach
Try not to disclose your IP [intellectual properties] data
But there are many other ways ChatGPT can assist PM’s, when used correctly.
Ps: I am working on the series of the ChatGPT and will be sharing various scenarios where we can talk about usability, applicability and implementation of Generative AI.
About the Author
Anubhav Sinha is a co-founder as well as the course developer of the Product Capability Uplift. In this role, Anubhav leads the development of the PMA as well as works as the product thinker of the Product Capability Uplift PMA.
Anubhav Sinha is a product coach, a product management practitioner and technology product geek with around one and half decade of the product management and development experience that ranges widely in the B2B and B2IB product space. He is known for contributing and creating products majorly in the start-up space, helping start-ups in their early stages and contributing industry product organisations as user-experience flow optimiser. He had served industry as Principal Product Owner [co-founder], Product and Design Thinking Coach, Product Owner and Transformation Coach.
Anubhav holds a Post-Graduation in Marketing - IB and Bachelor of Engineering in Electrical and Electronics.
コメント