You've probably been having conversations lately about whether to use AI for testing.
I've even gotten comments from some of the testers in my videos proclaiming that ALL AI is snake oil.
In fact, you're likely getting tired of all the talk about AI in testing by now.
Listen—I get it, but I believe testers must be familiar with this topic and must squash some of the misinformation around it.
One overhyped area is Generative AI—but is that all there is? And how does it impact the role of a QE or Testers?
Some folks have been asking me, “Do we still need human testers when AI can create and test software?”
I'll try to address these questions and concerns in this post.
AI Expert Guest Mark Creamer
I recently had the opportunity to discuss this subject with Mark Creamer from ConformIQ, a company that is leading the way in AI-driven test design automation technology solutions.
Mark brings over four decades of industry expertise to the table and shares perspectives on how AI is revolutionizing testing practices and influencing the trajectory of quality assurance for the future.
But first how do I address the snake oil comment?
Is AI in Automation Testing Snake Oil
It's definitely frustrating to see so many companies making big promises without delivering real value.
But I use AI all the time and have spoken with many engineers who have the opposite experience of you.
That is why I always recommend each tester do a automation testing tool POC for themselves to see if it works for their env/user case. If it does great if not move on.
But as you will see Mark Creamer is no snake oil salesmen – he really know his stuff no B.S!
Now on with the post :)
AI Testing Unveiled
When discussing AI in testing, it can be tempting to focus on the excitement surrounding AI tools such as Chat GPT; yet, according to Mark Creamer's insights, AI was being integrated into testing practices long before most people were aware of it. In fact, he worked on an AI project during his MBA studies in the early 1980s.
So, what has changed?
AI's presence in testing isn't the concern; it's more about how visible and easily available it has become to everyone, thanks to Gen AI leading the charge in bringing AI into the limelight as a seemingly fresh advancement when, in fact, different AI forms have been silently assisting testing tools for quite some time.
Symbolic AI: The Unsung Hero of Testing
One such form of AI that's been instrumental in testing is Symbolic AI.
Unlike the flashier Gen AI, Symbolic AI operates more like an embedded technology, working quietly in the background to optimize test case generation and execution.
Mark emphasizes the effectiveness of Symbolic AI in generating expected test scenarios:
“With those criteria in mind, it will ensure coverage of the requirements for testing in a deterministic manner.”
In high-stakes industries such as finance and healthcare, where dependable and consistent testing is essential, Symbolic AI's ability to anticipate outcomes and maintain consistency proves beneficial.
Gen AI: The New Kid on the Block
Symbolic AI has been making advancements in testing for quite some time without much attention.
In contrast, Gen AI has recently made a grand entrance with its remarkable capability to produce text and code that closely resembles human work.
This development has captivated the interest of many professionals in the field.
Mark advises against viewing Gen AI as a substitute for existing AI technologies in testing scenarios by emphasizing that its strength lies in enhancing the capabilities of individuals, as opposed to transforming beginners into experts due to reported instances of Gen AI producing results.
Mark proposes that Gen AI should not be considered a substitute for testers or other AI technologies, but rather viewed as a supportive tool. Gen AI is particularly effective at tasks such as aiding in model creation, generating test concepts, or assisting in the development of test scripts.
However, when it comes to developing consistent test suites, Symbolic AI still maintains the upper hand.
The Ultimate Combination: Merging AI Innovations
Mark argues that the real power comes from combining different AI technologies to create a more comprehensive testing approach.
For instance, Gen AI is handy for producing system-level models based on user anecdotes or behavior-driven development (BDD) scenarios.
These models can then be inputted into Symbolic AI systems to produce test cases that encompass the system rather than just specific parts.
This method enables teams to use the advantages of both AI technologies.
Gen AI can comprehend and produce text and code that resonates with human language patterns.
Symbolic AI's ability to create test cases in an optimized manner is noteworthy.
The outcome is a testing procedure that's more effective and comprehensive than what individuals or a solitary AI technology could attain.
AI: Enhancing, Not Replacing, Human Testers
One key lesson we learned from our chat with Mark is that AI isn't meant to replace testers but rather to empower them and enable them to concentrate on valuable duties.
Mark emphasized that he believes General AI is highly effective in enhancing individuals' abilities and increasing their productivity levels. This viewpoint extends to the utilization of AI in the field of testing.
By automating tasks and producing test scenarios, as well as aiding in the development of comprehensive system-level frameworks, AI enables human testers to concentrate on intricate and subtle facets of quality assurance.
In addition, Mark mentioned that using AI-created models can significantly improve teamwork within a group.
“The visual representation in the model provides benefits for grasping system interactions and devising testing plans,” he explained.
The Collaboration of Human and Artificial Intelligence in Testing Evolution
When envisioning the future of software testing for us, it's evident that AI will have significance; nevertheless, this doesn't signal the exclusion of human intervention in the testing process altogether.
Instead, what lies ahead is a scenario where intellect and artificial intelligence collaborate, each enhancing the strengths of the other.
Human testers contribute creativity and intuition, along with the capacity to grasp business scenarios—abilities that AI has not yet mastered fully.
AI adds speed and consistency to the mix, along with the capability to handle volumes of data.
When combined effectively, they create a synergy that elevates software quality to new levels.
See AI in Action Webinar
You have gotten this far, so I assume you believe in the value of AI's potential in testing.
Do you want to learn more about leveraging Symbolic AI and Gen AI to enhance your testing processes?
We're excited to announce an upcoming webinar in which Mark Creamer will discuss these topics and answer your questions live.
In this Webinar, you'll learn:
- How different types of Artificial Intelligence are used for software testing
- Correlation between automation and AI
- Practical AI use cases in software testing
- Tips on how to gauge your readiness
- ConformIQ's take on Requirements to Automation
Don't miss this opportunity to gain valuable insight from one of the industry's leading experts on AI in testing.
Register now for our webinar “To AI or Not to AI in Testing: Navigating the Future of Quality Assurance” by clicking the link below:
Software testing is evolving rapidly, and AI is at the forefront.
By understanding and embracing these new technologies, we can create more efficient, effective testing processes that produce higher-quality software.
Join us for this Webinar and take the first step toward the future of testing!
