The Double-Edged Sword: Accelerating Product Development with AI Wisely

In Product Development, artificial intelligence tools like ChatGPT are becoming increasingly prominent. There's a buzz around how these AI tools can expedite processes such as crafting user stories or even creating project plans. Maybe we could set it up in such a way that it can look at our past data and automatically assign story points for stories?? What if we can automatically add the story to the Jira board and assign that to a sprint/a resource1? It sounds like we can achieve more than twice the work with half the effort!?Wait a minute. What is ChatGPT writing? User Stories? Plans??

Hold That Thought: Is More Always Better?

Writing user stories at lightning speed - what could possibly go wrong? In my consulting and training engagements, I often come across scenarios where a business analyst drafts a story and tosses it over the digital fence to the developers. Or even worse, a ScrumMaster creates stories and assigns them to developers. Now, with AI, we’re looking at potentially doing this faster than 100 words per minute. ?

Speaking of speeding things up, take a look at this Logitech mouse2 designed for AI-enhanced productivity. Does anyone find it useful?? Just because you can do things quicker doesn't mean you're doing them better. As it seems, even your mouse can get ahead of itself!?

Going back to Story Writing3.? The essence of user stories is not about writing down requirements; it's about conversation. It’s about creating a common understanding of the user needs for everyone. Yes, then we document what we communicated too!?

And what about project plans? Planing is something that we collaboratively do, and plans are something that we create from that activity. When we respond to changes, we change our plans. ?

Smart Use of AI: Enhancing, Not Replacing

While Large Language Models (LLM) like ChatGPT are adept at generating content that sounds accurate and authoritative, there's a hidden risk. These models can sometimes produce outputs that are plausible, yet factually incorrect or contextually inappropriate. In other words, the output can be very convincing but wrong. That is a dangerous combination (more on that in a later article). In Product development settings, this could mean generating user stories that seem perfect on the surface but miss crucial nuances of the user's actual needs or the project's specific requirements. After more than a year of interacting with ChatGPT and other LLMs, we now know that having a human in the loop is important. ?

If we shouldn't let AI replace the human touch in story crafting, what’s its role? Let's explore a couple of scenarios (below) where AI doesn’t just spit out user stories like a factory line but supports a more thoughtful Agile process. See examples of prompts that can help at the bottom of this article.?

Engage and Reflect

The allure of AI tools like ChatGPT in Product Development is undeniable, promising to expedite tasks and streamline processes. However, the rush towards automation risks further eroding the foundational Agile principles of collaboration and human engagement. As a community, we have been ignoring these principles even before the advent of Language Models in the last year.? As we harness these powerful technologies, let us not forget the core value: collaboration. This mindful approach will enable us to utilize AI not just for speed and efficiency, but help us in creating valuable products that customers want.

How are you using AI to enhance your product development? What successes and challenges have you encountered? Please share your experiences and insights, as we all navigate this evolving landscape together.

———

Scenario 1: AI as a Reviewer and Formatter

In this role, AI does not start from scratch but enhances what the human team has created. After a brainstorming session where the Agile team drafts a preliminary user story, AI can step in to refine this draft.

Here’s a simple prompt that you can start from for ChatGPT:

'We've drafted the following user story: "As a user, I want to filter search results by date so that I can find the most relevant articles quickly." Can you review this for compliance with the INVEST principles? Also, please suggest any improvements to make this story more clear and actionable.'

This prompt directs the AI to ensure the story is Independent, Negotiable, Valuable, Estimable, Small, and Testable, and it invites suggestions for enhancement.

Scenario 2: AI as an Analytical Assistant

Imagine an AI agent that is not only familiar with the basic principles of user stories but also integrated with the specific context of your project—understanding user needs, product goals, and company vision. Such an agent could analyze discussions with customers or lists of defects and suggest potential user stories on the context. How about an agent that can listen in on a conversation with a set of stakeholders and come up with a list of possibilities/ideas from that conversation? These examples don't replace human judgment; rather, it enhance the brainstorming process, presenting ideas for the product owner and team to consider and refine.?

Creating an agent is beyond the scope of this article (more on that in a future article). But here's an example of how you might engage an AI tool to extract actionable items from a meeting summary:

'Here is a summary of our latest discussion with the stakeholder regarding the project's next phase: "The stakeholder emphasized the need for enhanced security features and a more intuitive user interface." Based on this, can you suggest potential user stories that align with these needs?'

This prompt helps the AI understand the context and apply its knowledge of user story creation to generate ideas that the team can then evaluate and refine.

Reference:?

1. That was a satirical reference. Please do not use the term resources when you mean people - as Paul Boos would say “If you mean people, say people (not resources)”
2. Tech brands are forcing AI into your gadgets—whether you asked for it or not
3. How should user stories be written? by Ron Jeffries?

More Actions to learn

Join a community to explore these ideas more

Agile Denver Meetup Group’s? SIG for AI

The AI Product Development Network - a Meetup group

Attend a Workshop

Product Development with AI

Note

This article was authored by Manoj Vadakkan , with assistance from ChatGPT4. All original ideas and the initial draft were created by Manoj. ChatGPT4 was employed to enhance paragraph organization and correct grammatical inaccuracies. Additionally, the concept and creation of the accompanying image were developed by ChatGPT4.