As we try to stay ahead of the curve in software development, the trend in the tech sector is all about AI. I recently had an interesting conversation with a colleague while we eased down the escalator after SmartBear’s successful GTMKO (Go-to-Market Kick Off) in Boston.
The conversation started around how we would incorporate AI into some of our software development tool sets. As the escalator took us to our final destination, we discussed the complexity of doing this as well as using AI to generate data from the application that could be exported and used in other systems. This, you say, is not really AI, and that’s correct.
We were talking about the alternative approach to helping those on an AI journey, to get to a stage when they in fact want to implement AI, they would have the tools to get started. Whether we like it or not, 83% of companies using AI strategy see it as top priority.
Many on an AI journey – like with other journeys such as Behavior-driven development (BDD) – are investigating it and investing but not executing, leaving a gap and executives scratching their heads saying, “What just happened?” Rushing into AI is becoming quite common, and this is setting them up for failure. So, back to our escalator journey and how my colleague and I went a full 360 degrees, as we entered the final steps of our journey, we came to three conclusions:
- The AI journey is more of a discovery than an implementation.
- Have the right structure, and more importantly, the right data available to start.
- Be patient.
Many articles are written on strategy and how you implement AI; one such example is an interview with Andrew Ng and Beena Ammanath from Deloitte. They had some very interesting insights especially looking from a top-down approach and getting C-level to do an introductory course on AI.
With any journey comes roadblocks and complexities such as:
- Data Quality: Data regulations can have an impact on what data is available so the full story may not be available.
- Skills Gap: It’s a relatively new space so skills are in demand.
- Integration with Existing Systems: For legacy systems, integrating AI into existing infrastructure can be complex, especially when dealing with legacy systems that may not be designed to support modern AI applications.
- Interoperability: Ensuring seamless communication and integration between different AI components and existing technologies are top challenges.
- Ethical and Regulatory Compliance: It’s important to address biases and fairness in AI algorithms to avoid discrimination and ensure fair outcomes. With regulatory compliance, navigating complex regulatory landscapes, including data protection and ethical standards, can be a significant roadblock.
Some companies are going headfirst into their AI journey rather than taking a step back first to see where the benefit might be or even how you can set things up like what we talked about above in relation to data quality.
Saying you are “doing AI” is not a case of getting the latest AI tool and plugging it into your system without knowing what the expected output should be. It’s about embracing and slowly introducing the concept such as using ChatGPT to create a template your organization might use or using GitHub Copilot to write a piece of code or a unit test you can run in your application across the software development lifecycle (SDLC).
With the advancements in AI in a software testing environment, for example, we are seeing focus on:
- Test case creation
- Test data creation
- Analysis of executed test
- Detecting defects and root causes
- CI/CD integration
- Monitoring of real time systems
What about the benefits of AI?
AI is more efficient, and with more analytics and quick responses to changes, costs can be reduced and predicting downtimes in critical environments can also be improved as well. AI also mobilizes your workforce, empowering them to control how they work and giving them time back to prioritize other more important work. According to Forbes, 64% of businesses expect AI to increase productivity.
And, AI can come with more accuracy when it comes to important decision making, having lots of information at hand to quickly make decisions more accurately and more impactfully. Last year, Harvard Business Review saw poor decision-making costing big companies on average 3% of profits.
These benefits will continue to be more of a focus for companies wishing to improve quality of applications, processes, and procedures. This is key to the forward thinking of companies that wish to look toward AI as a natural progression in an ever-evolving world of software quality, processes, and procedures.
Some of us will embrace AI and how we can get to a more productive and efficient way of software quality, and some of us may park it for now, but as AI develops and evolves, most of us and our companies will eventually do a full 360, coming back to easier workflows with AI as part of that strategy.
The future is now with AI; let’s embrace it.
EuroSTAR Huddle shares articles from our community. Check out our library of online talks from test experts and come together with the community in-person at the annual EuroSTAR Software Testing Conference. The EuroSTAR Conference has been running since 1993 and is the largest testing event in Europe, welcoming 1000+ software testers and QA professionals every year.
