Skip to main content

Rewriting product and engineering roles in the age of AI

The future of software development

As the evolution of AI tools and use cases continues to accelerate, we’re seeing a fundamental shift in how work is done across the software development life cycle. How can tech industry leaders prepare for this AI-driven future of work?

Writing user stories. Troubleshooting code. Translating outdated software. Converting wireframes. All these tasks are central to the day-to-day lives of product managers, software engineers, and others involved in the software development life cycle (SDLC)—and are increasingly being augmented and accelerated by artificial intelligence tools.

Generative artificial intelligence (GenAI) has the potential to disrupt workflows across the enterprise, but use cases at each stage of the software development life cycle are largely more advanced and have achieved greater adoption. A recent GitHub survey of 2,000 globally distributed developers found that more than 97% of respondents had used AI coding tools at work, and 88% in the US indicated at least some company support for AI use. 

As the evolution of AI tools and use cases continues to accelerate, a fundamental transformation of the work done by product managers, software engineers, and others involved in adjacent workflows is underway.

How AI is already transforming workflows

As AI-enabled solutions enter the market, the greatest value has been realized when GenAI is embedded throughout the entire SDLC, rather than focusing solely on coding. The following effects on roles across the SDLC have emerged: 

Planning for the future

Given this degree of disruption and model evolution just a few years into GenAI adoption, one thing is for certain: Nothing is for certain. It is therefore essential for tech leaders to engage in scenario planning to seek answers to fundamental questions about their product and software teams’ potential size, shape, workflows, and ways of working. Key uncertainties to consider in planning include:

As of the writing of this article, Generative AI model development and advancement of agentic capabilities are evolving at a dizzying pace. The degree to which this is sustainable is perhaps the most significant uncertainty; AI companies are optimistic, but there are several hurdles that may hinder development. 

It remains to be seen to what degree the GenAI ecosystem (models and applications) for product engineering will be fragmented or integrated, closed or open source, etc. Speed of adoption across the entire development life cycle will likely depend on how easy it is to leverage tools within existing technology ecosystems and how fast it is to build necessary integrations.

The regulatory environment regarding technology, and especially AI, continues to come into focus but is far from finalized. While these decisions remain in flux or at odds, certain use cases are likely to be avoided, especially by enterprises (for example, those relying on personally identifiable information). Where companies choose to draw the line in the grey area may have as much to do with risk tolerance, culture, and competition as it does regulatory pressure.

The need for AI chips, energy, and other natural resources, as well as training data, is likely to continue to grow as model development and adoption grows. These resources are constrained and may cause a slowdown in development.

With great power, as they say, comes great responsibility. Leaders in the tech industry are likely to be the first to face widespread task augmentation and automation from AI, and they should take seriously the responsibility to prioritize the workforce experience. 

The promise of Generative AI in the software development life cycle is significant and, in many cases, already proven. Leaders should begin taking action as soon as possible to prepare employees to take advantage of this tremendous opportunity, and to design an intentional future that can benefit all. 

Did you find this useful?

Thanks for your feedback