The STARWEST conference is the leading conference for testing innovation, bringing together the brightest minds in quality and testing. Not surprisingly, one of the main issues that was top of mind this year was AI and its impact on development and testing.
If you did not get a chance to attend (or even if you did), here are four key insights we observed throughout the event. They are based on the event sessions and the hundreds of conversations we held with attendees and thought leaders in the industry.
1. AI Adoption Will Increase in the SDLC
AI is being widely-adopted throughout the SDLC, from coding through testing to deployment and release. In the upcoming year, approximately half of engineers will be using AI for a wide range of needs, and percentages will significantly increase in the upcoming years. In addition, in the next year AI’s abilities to augment, improve and transform engineering steps will increase as well, further driving adoption.
AI will simplify processes like interpreting code and debugging, speed up phases like test generation, improve capabilities like security code analysis, and support overarching needs like creating monitoring plans. Overall, AI will become a huge productivity driver that supports developers across multiple tasks, significantly accelerating time-to-market.
Back to top2. Testing as a Major Focus for AI Integration
Throughout the SDLC, AI is primed to provide special value for testers. AI can help with automated script generation, test planning, requirements analysis, analyzing results, maintaining automated test scripts, and many more tasks.
Testers are poised to benefit from AI’s ability to speed up the tasks they wish to avoid. This will free them for more creative work, as well as for directing and guiding AI and creating and overseeing testing strategies.
Back to top3. AI Is the Next Productivity Frontier for Developers
What are the numbers behind the productivity promises? According to McKinsey, they are unprecedented. With AI, code generation can take 35-45% less time and code refactoring 20-30% less.
AI is found to be mostly helpful for repetitive work, for jumpstarting tasks, or for refining existing code. However, it is not as valuable when it comes to complex tasks. While it can provide support, it is no replacement for the human brain. This is food for thought for testers, since it is also where testers can shine when using AI for testing.
Back to topIn the realm of application testing, integrating AI is essential for teams to stay efficient and ahead of the curve. Discover how Perforce’s AI-powered capabilities drive testing efficiency, mitigate risks, and accelerate your application’s time to market.
4. Testing the Testers: The Importance of Testing AI
AI comes with risks and issues, which need to be addressed when incorporating AI in software and processes. Risks include bias in results, non-compliance with guidelines, and safety risks. Such faulty AI might lead to operational setbacks, increasing costs, financial and reputational damage, and losing a competitive edge.
Rigorous testing of AI systems ensures accuracy and transparency, which build user and customer trust. These are foundational for widespread adoption of AI. Testing of AI includes:
- Unit testing of the AI systems
- Performance testing AI model efficiency, speed, and scalability
- Bias and fairness testing of outputs
- Explainability testing of the decision-making process
- Testing to ensure no private data is exposed to public AI systems
When choosing a testing tool vendor and when using AI in your testing process, ensure it is verified and transparent AI.
Back to topBottom Line
We believe that in the upcoming 1-2 years, AI-driven testing will become prevalent. Companies and testers that wish to maintain a competitive advantage can:
- Consume resources about the latest advancements AI in testing.
- Understand how AI will transform test automation and analysis.
- Develop a plan for incorporating AI and AI-driven tools in their testing strategy.
- Implement guardrails to reduce AI testing risks.
For those interested in learning more about how they can positively impact their testing through the use of AI, testers can request a demo of BlazeMeter's Test Data Pro.