Agile AI: Will artificial intelligence end agile as we know it?

It’s becoming clearer by the day that AI will profoundly impact all industries and professions in the coming years. This realization evokes a range of emotions. Some find it alarming, questioning whether their jobs will become obsolete. Others enthusiastically embrace it, hoping to reduce their workload. 

While the reviews of GitHub Copilot imply that AI coding assistants are not yet ready to transition from “Copilot” to “Pilot,” there are other aspects of developers’ work that they may gladly welcome AI to take over, such as agile project management.

This begs the question: How might artificial intelligence revolutionize agile practices, and will it fundamentally change or even end agile practices? In this article, I will discuss how agile AI will shape and possibly even end agile as we know it. 

Putting an end to agile-related overhead

When it comes to agile and any form of project management, a considerable amount of overhead and facilitation is expected. From crafting epics and user stories to organizing Agile events, it is understandable why dedicated Scrum Masters and Agile coaches have become commonplace. But what might this look like if we introduced AI?

No more standups or status updates

One area where AI excels most is in summarizing and drawing conclusions from existing information–you’ve probably already seen chatbots like ChatGPT flexing their muscles in this regard. Because of this, it’s only natural that one of the first agile events we may expect AI to tackle is daily standups.

The goal of standups is simple–provide a status update and a rundown of current blockers. In the future, this will likely look like an AI-generated summary of where everyone’s currently at, with blocker notifications sent directly to teammates who can unblock the task. 

To take things a step further, these AI-generated project summaries likely will go beyond standups, evolving into daily to-do lists that can keep devs focused on their priorities. 

Story point estimation will become redundant

While status updates seem like an obvious first step for AI, story point estimation is a more brow-raising prospect. 

With the right amount of data (i.e., A historical sprint record and well-documented user stories), AI could generate some seriously accurate story point suggestions that developers simply will have to approve or reject. The rub? AI will likely need help explaining why a coding problem is complex. This will place a greater emphasis on the “discussion” aspect of this agile practice rather than the estimation itself. This is why I believe estimation itself will likely become redundant in the age of AI.

Backlog refinement and sprint planning will no longer be a “meeting”

Prioritization is king in Agile, and sprint planning and backlog grooming are how it’s done. While AI requires context for this, with enough historical sprint data and goal clarity, it’s likely AI will be managing the sorting, planning, and prioritization of our backlogs and sprints. 

But does this mean an end to these 2 agile events? In the short term, it will likely mean reduced time spent in these meetings. However, as AI gets more accurate, it will likely require only one person to review and approve the AI-suggested sprint plans and backlogs, eliminating the events entirely. Pretty neat, right?

Manual categorization of work will be a thing of the past 

We’ve talked a lot about how important it is for AI to have data. And that data needs to be accurately categorized to be useful for AI (and humans!). Luckily, AI is already improving the accuracy and ease of categorizing work. 

AI categorization suggestions like Zenhub labels analyze your Issue data to present accurate recommendations for labeling your issues. This is huge for ensuring that AI has proper context to enable future AI suggestions (like those discussed above), not to mention making it easier for humans to filter data in the meantime. 

In the future, AI will likely take over other aspects of Issue metadata (e.g., tagging assignees), generating Issue descriptions, and linking Issues to Epics, Projects, Sprints, or PRs. These increasing levels of automation would likely reduce or entirely put an end to manual work associated with agile. 

Will agile change forever? 

Will AI put an end to agile as we know it? Yes. As technologies shift, ways of working change, and in ten years, it goes without saying that agile won’t look the way it does today. 

Despite this, at the end of the day, agile will always involve humans working together and having conversations with each other – not just bots. With less time spent on status updates, organizing projects and agile events, and estimating stories, we’re optimistic that AI will give teams more time to collaborate, explore new ideas, be creative, and solve the most pressing problems. 

If you’re interested in the future of Agile AI and how we’re building these experiences into Zenhub, join our AI waitlist to get the latest updates on new AI releases.