So there’s this broad discussion all over the internet about whether the Language Models (LMs) will replace this job or that job. That made me wonder about the Product Management role and how LMs like ChatGPT will affect this role and skill set. Given that, I’ve almost seen everything relating to ChatGPT over the Internet, here’s my take on the subject.
Non-routine cognitive tasks across all industries are currently in the early stages of being automated with the help of AI technology. The following are some notable examples:
— Deciphering handwriting
— Impartial decision-making in fraud detection
— Paralegal, contract/patent law research – Symantec’s Clearwell System
— Diagnostic tasks in health-care – IBM’s Watson
It’s intuitive to think that Product Management is a career that involves more soft skills compared to SDE jobs. And AI & ML gaining enough expertise to replace the soft skills of humans is far out of the picture, given the infancy of NLP (Nature Language Processing) models. Also, according to a study, 80% of the surveyed companies state that their higher priority from disruptive innovations in Robotics + AI is to empower human workers, solving their weaknesses in a few cases and increasing their strengths in most of them. Not to replace them.
So let’s talk about how AI models like ChatGPT will augment the skills of Product Managers and solve their weaknesses. The questions I’m interested in finding answers to:
- Will ChatGPT overtake all the hard skills of Product Managers? Are there any hard skills that ChatGPT can’t replace?
- How can Product Managers use ChatGPT as a helping tool to become more productive?
- Can ChatGPT help improve any of the “soft skills” of PMs?
- What does a typical future Product Manager look like, armed with ChatGPT or similar tools?
- Does ChatGPT necessarily make the PM better at their job?
Will ChatGPT overtake all the hard skills of Product Managers? Are there any hard skills that ChatGPT can’t replace?
There’s this broad argument a lot of people are making, and the ChatGPT itself says, things like looking over analytics and making up decisions based on that can be automated.
When I ask this question to the ChatGPT, this is their answer.
As a large language model trained by OpenAI, I do not have the ability to replace or overtake the hard skills of product managers or any other professionals. I am a machine learning model designed to generate human-like text based on the data I have been trained on, and I do not have the ability to perform the tasks or possess the skills of a product manager or any other professional.
In terms of hard skills that I cannot replace, there are many technical and practical skills that are essential for product managers, such as market research, product development, project management, and financial analysis. These skills require specialized knowledge and experience, and are not within my capabilities as a language model.
Additionally, product managers often have important personal and interpersonal skills, such as leadership, communication, collaboration, and problem-solving, which are difficult for a machine learning model like myself to replicate.
Overall, while I may be able to assist with certain aspects of product management, there are many important hard skills that are essential for success in this field and are beyond my abilities as a language model.
Product Management has some hard skills that are difficult and technically impossible to replace or automate. Skills like Market Research, Product Development, Customer Development, Project Management, and Financial Analysis, many of which require not only historic data based decisions but also in-person activities that generate those data. These in-person activities/tasks are far out of AI’s reach currently. In addition, AI’s incapability to answer “How good does the coffee taste?” test keeps majority of all product managers of physical goods out of danger.
Though an argument could be made that many aspects of the Market Research and Financial Analysis can be automated, human supervision will always be required to draw insights from past data and assumptions about the future that can be trusted.
Many aspects of project management can be automated, but overall, project management is a really humane job that requires you to deal with people given not only data about them but also deriving insights from ”real-world” interactions that AIs like ChatGPT can’t generate.
Can ChatGPT help improve any of the “soft skills” of PMs?
We know that ChatGPT can’t replace soft skills like leadership, communication, collaboration, product-solving, etc. But can it help improve these skills and make them more proactive in the job?
Communication:
Currently, if there’s one thing that’s driving me nuts, it’s knowing how easy the written communication has become for many in effect of ChatGPT. People who used to procrastinate about sending emails to recruiters, customers, employers, etc., can now send detailed and effective emails within minutes and low-effort writing. Let me show you how this works.
The purpose of receiving an email is to gain information from the sender. The purpose of sending an email is to provide information to the recipient. However, often times, hesitation can prevent us from effectively communicating our needs through writing. This is because crafting a well-written message can be time-consuming and difficult, which in all together makes a person uncommunicative of their needs. But with the help of AI tools like ChatGPT or whatever medium, if someone can communicate their needs and you’re getting the right information you can help that person with, isn’t that the definition of complete and effective communication? That’s what ChatGPT is revolutionizing: effective communication.
With ChatGPT, you no longer have to put off writing emails, whether they be long or short. While it may not always produce the perfect copy, ChatGPT can help you get started in a way that nothing else can.
For instance, let's say you would like to contact a customer who recently purchased your product. You've noticed that they've been using a certain feature and influencing others to do the same. Since they are leading a community of users, they would be a great person to talk to about the common problems users face and whether they've experienced any themselves. Your goal is to find out if they've noticed any bugs in the product or if they have any suggestions for improvements. You also want to know if the user group they are leading has used the product in any unique ways. Let's use ChatGPT to write a LinkedIn message for this scenario and see what it comes up with.
[Click on ▶️ to expand]
Prompt: I want to reach out to a customer who has recently purchased my product, and I’ve seen her using this feature online and influencing others to use it. Because she’s influencing a community of users, she would be a great person to talk to about the users' common problems and whether she faced any herself. My goal is to know, especially, whether she’s noticed any bugs in the product or has any ideas about improving the product. I also want to know whether the user group she’s leading has used the products in any unique ways. I want to write it in the format of a LinkedIn message.
ChatGPT Response:
Dear [User],
I recently noticed that you have been using [Product] and sharing your experience with others online. As the creator of [Product], I am always looking for ways to improve and enhance the user experience, and I would love to hear your thoughts and feedback.
I noticed that you have been leading a community of users who are also using [Product], and I'm curious to know whether you or any other members of the group have noticed any bugs or have any ideas for improving the product. I'm also interested in learning whether the group has used the product in any unique or creative ways.
I would be grateful if you could share your thoughts and experiences with me. I value your feedback and insights, and I'm always looking for ways to make [Product] even better.
Thank you for your support and for being a valued user of [Product].
Best regards, [Your Name]
[I’m not totally happy with the response. So I ask a follow-up question] Prompt: Can you make the response a bit better? Possibly, by empathizing with the user a bit more.
ChatGPT Response:
Dear [User],
I am so grateful to see that you have been using [Product] and sharing your experience with others online. As the creator of [Product], I am always looking for ways to improve and enhance the user experience, and I would love to hear your thoughts and feedback.
I noticed that you have been leading a community of users who are also using [Product], and I'm curious to know whether you or any other members of the group have faced any challenges or difficulties while using the product. I understand that no product is perfect, and I want to make sure that [Product] is meeting your needs and exceeding your expectations.
I'm also interested in learning whether the group has used the product in any unique or creative ways. I love hearing about the different ways that people are using [Product], and I'm always looking for new ideas and inspiration.
I would be grateful if you could share your thoughts and experiences with me. I value your feedback and insights, and I'm always looking for ways to make [Product] even better.
Thank you for your support and for being a valued user of [Product].
Best regards, [Your Name]
[The response at this point is decent, and with a bit of tweaking here and there, it’s almost as good as those generic emails companies send you to get feedback. But I want to make the message a bit better. So I ask it to add something about my favorite feature] Prompt: I want to add a bit about my favorite part of the product.
ChatGPT Response:
Dear [User],
I am so grateful to see that you have been using [Product] and sharing your experience with others online. As the creator of [Product], I am always looking for ways to improve and enhance the user experience, and I would love to hear your thoughts and feedback.
One of my favorite things about [Product] is [insert favorite feature or aspect of the product]. I'm curious to know whether this is also a favorite feature for you and your community of users, or if you have discovered any other features or benefits of the product that you particularly enjoy.
I also noticed that you have been leading a community of users who are also using [Product], and I'm curious to know whether you or any other members of the group have faced any challenges or difficulties while using the product. I understand that no product is perfect, and I want to make sure that [Product] is meeting your needs and exceeding your expectations.
I'm also interested in learning whether the group has used the product in any unique or creative ways. I love hearing about the different ways that people are using [Product], and I'm always looking for new ideas and inspiration.
I would be grateful if you could share your thoughts and experiences with me. I value your feedback and insights, and I'm always looking for ways to make [Product] even better.
Thank you for your support and for being a valued user of [Product].
Best regards, [Your Name]
At this point, the response might not be the best message to send, but is it good enough starting point? Heck yeah. We can use this writing as a reference to write not only great sentences but also a skeleton structure of the message, which we can tweak in any way we want.
Leadership:
According to researchers who conducted interviews with 33 senior industry leaders, there are five key leadership capabilities that’ll need to be constantly developed and updated:
- Digital know-how (Update capabilities in new technology)
- Data-driven focus (Develop knowledge of data analytics)
- Networking (Build networking and relationship skills)
- Ethics (Engage with ethical and moral issues)
- Agility (Act with agility and decisiveness)
How can Product Managers use ChatGPT as a helping tool to make them become more productive?
This is gonna be an interesting one.
ChatGPT or any LM models can help with tasks that require natural language processing and generation. Such areas in Product Management are,
- Writing Product Descriptions
- Generating Reports and Presentations
- Analyzing Customer Feedback & Reviews
Other aspects of a PM that ChatGPT can help with:
- Information and Insights into Market Trends
- Conducting Market Research
- Knowing Customer Preferences
- Gathering Information about Competitors
- Suggesting Industry Best Practices
What does a typical future Product Manager look like, armed with ChatGPT or similar tools?
ChatGPT has enough to say about this that it’s better I don’t add anything.
In the future, product managers may be equipped with advanced tools like ChatGPT and other large language models to assist with various aspects of their work. These tools will enable product managers to access and analyze vast amounts of data, generate reports and presentations, and communicate with customers and stakeholders in a more efficient and effective manner.
One of the key benefits of using tools like ChatGPT is the ability to quickly and easily access and process large amounts of data. Product managers will be able to use these tools to gather and analyze information about market trends, customer preferences, and industry best practices, which will help them make informed decisions and develop effective strategies.
Another important benefit of using large language models like ChatGPT is the ability to generate high-quality written content, such as product descriptions, reports, and presentations. These tools will enable product managers to produce professional-quality content quickly and easily, which will save time and resources and help them communicate more effectively with customers and stakeholders.
Furthermore, large language models like ChatGPT can also assist product managers with tasks that require natural language processing and generation, such as analyzing customer feedback and reviews and identifying patterns and trends in customer behavior. This will help product managers understand their customers' needs and preferences and develop products and services that meet those needs.
Overall, the use of advanced tools like ChatGPT and other large language models will enable product managers to be more efficient, effective, and strategic in their work. These tools will provide product managers with valuable insights and information, and help them make informed decisions and develop successful products and services.
Does ChatGPT necessarily make the PM better at their job?
That might seem a stupid question, but here’s what I’m trying to verify:
By providing access to a vast amount of resources and enabling Product Managers to be 100 times more productive with routine tasks, tools like ChatGPT free up time and resources for learning about things that fall outside of its capabilities. This, in turn, makes Product Managers more resourceful and insightful. With the help of ChatGPT, Product Managers can better focus on the areas where they can add the most value, ultimately leading to better outcomes for their teams and organizations.
Knowing that:
ChatGPT or such models, of course, don’t make someone necessarily better at being a PM because good Product Managers are a combination of their skills, experiences (real-life, with persons, etc.), and ability to develop strategies based on everything at their disposal (combination of data + real-life experience)
In order to be a successful product manager, individuals must have a deep understanding of their industry, market, and customers, and be able to apply that knowledge to make informed decisions. They must also have strong personal and interpersonal skills, such as leadership, communication, collaboration, and problem-solving, which are difficult for a machine learning model like ChatGPT to replicate.
Here’s how I see ChatGPT playing out step-by-step:
- Helping getting experiences: Because experience will come with time and jobs, let’s see other parts of the PM jobs that’ll help PMs get more experience
- Giving access to analyzed customer data, it’ll allow Product Managers to better understand their customers by freeing up time for in-person interactions and leading to deeper insights
- By quickly and easily accessing market data and summaries through ChatGPT, Product Managers can focus on learning the parts of the market that the model can't help with.
- Developing strategies. Researchers argue that Artificial Intelligence (AI) will, in the very near future, have a profound impact on the conduct of strategy. [First of all I’m aware that some strategies, one in a million, come out as ingenious like never thought before] But how do you learn to develop strategies?
- You learn that.
- You’re born with some genius thinking capability.
- Your experience. We talked about it before.
The learning part is something ChatGPT can help with.
What about the strategy courses that teach you historical events, war methods, developing deceptions, and whatnot (yes, I took some of this “Strategic Thinking Skills” course from Drexel University once)? What if you had access to such strategies in an efficient manner, and as a result, you read more of those [you’re reading history, right?], and as a result, you learn more about it? Chat-GPT-like models are capable of that, and as an outcome, you’re more focused on developing novel strategies based on past insights without having to go through all those historical texts surrounding it. However, surrounding texts are helpful, too, because they make you think more [?]—which you can easily get by asking the ChatGPT if you wish.
- Communication? Talked about it above. Makes written communication better in some aspects. If that’s the case, the person can focus on developing verbal and other forms of communication sooner. Also, ChatGPT-like models can help tremendously in verbal communication, too. (Ask: )
- Leadership and collaboration go hand in hand with communication because it’s about knowing how to deal with certain situations, people, etc. So it can be improved in the same way as communication.
Now, I’m not gonna provide a conclusion to the points I’ve mentioned above. I’ve just provided you with the data. Because if you’re reading this, you’re a PM or an aspiring PM and are good at interpreting data. Tell me what you think.
The Danger Zone
While all of these sound good, it’s unclear what’s the side effects of relying too much on ChatGPT-generated writings and communication. Also, humans are really good at connecting dots that usually don’t go together. That’s how many of the product insights get born. I wonder if LM models like ChatGPT can generate such insights that are trained on historic data. Metaphorically speaking, if the LM models existed before the invention of cars, would it be able to predict the invention/necessity of cars, or would it predict people will want a faster horse?
LM models-based communication with customers can also reduce trust in the product or company if they start to think they are communicating with a “Robot” when they complain about the product. Relying too much on AI can also lead to an increase in depression due to a lack of purpose in life, as people may feel hopeless when faced with a computer system capable of doing something they've committed years of their life to master.
Conclusion
Have you seen the superman movie “Man of steel”? ChatGPT reminds me of this scene where superman discovers his flight power.
From taking small steps to flight is similar to going from browsing ten pages to asking ChatGPT to complete the task. When Superman first discovers he can fly, he takes a long flight above mountains and lands (video above), where he travels to places that’d take normal humans hours & hours. Isn’t this power similar to something we’re experiencing with ChatGPT? But one observation I have regarding Superman’s speedy flight above mountains and lands is that—his ability to go from one place to another (fast)—doesn’t necessarily speed up his decision-making ability. Similarly, for normal human beings, being able to receive information fast or being able to finish a book in an hour doesn’t necessarily speed up their information-processing ability. Brain development is an evolutionary process that’ll take thousands of years.
So it’s fair to think we won’t be far smarter overnight because of this technology, but it’ll come in handy for many routine tasks before it gets smarter. What else will be possible when it gets smarter? That’s another discussion for another day.
As usual, I encourage and appreciate everyone’s constructive feedback. Let me know what you think.
Interesting stuff across the internet Product Managers are doing/building with ChatGPT
Awesome ChatGPT Prompts: https://github.com/f/awesome-chatgpt-prompts
Crash Course on ChatGPT: https://youtu.be/JTxsNm9IdYU
Write requirements for your feature: How can ChatGPT help Product Managers
Google Search Extension: Chrome extension to display ChatGPT responses along with Google search results
Browse more ChatGPT prompts and responses: https://www.learngpt.com/
Using ChatGPT as a Co-Founder: https://www.atomic14.com/2022/12/05/using-chatgpt-as-a-co-founder.html
Try out your own: https://chat.openai.com/
© Talha Chowdhury
Find me on LinkedIn: https://www.linkedin.com/in/mdtalhachy/