How to navigate the human-machine partnership in testing

As the march of the machine continues in the form of analytics and machine learning, we're witnessing what seems like endless encroachment on jobs in industry after industry. And the pace is quickening.

The test engineering community has already been through its first phase of automation, in the form of scripted testing, and now continuous testing. With advancements in analytics and machine learning in testing, jobs once thought immune to automation are now being replaced by machines that can perform Cognitive-based tasks. Those jobs are gradually being transformed as organizations increasingly delegate sets of tasks to automation. Meanwhile, the test-engineering world is in the early stages of using smart assistants in its testing processes.

Going forward, test engineers will remain in demand. But tomorrow’s most valuable engineers will not be geniuses in cubicles; they’ll be those who can build relationships, brainstorm, collaborate—and lead. They will form a human-machine partnership.

 

 

World Quality Report 2018-19: The State of QA and Testing

Humans aren't going anywhere

That's what Geoff Colvin says in his book, Humans Are Underrated: There will always be a role for humans, even as machine learning marches forward. The global cultural norm is to demand human accountability in key roles such as generals, judges, doctors—and test engineers— even if their decisions are derived with the help of machines.

The author makes two additional points. First, people keep changing their understanding of what the problem is and what their goals are. Do changing requirements come to mind? These are issues that people must work out for themselves. Requirements evolve, and testers work together with developers and product owners to establish expected behaviors and define acceptable quality standards.

Second, people want to work with other people in solving problems, tell them stories, hear stories from them, and collaborate to create new ideas. Test engineers deal with these situations every day. Developing and refining these skills is a natural byproduct of your job as a tester.

People manage relationships; machines don't

Peter Drucker coined the term knowledge worker in the 1950s, but today the most valuable people are relationship workers. Towers Watson and Oxford Economics asked employers what skills they’ll need most in the next five to ten years. The top results were these skills of social interaction:

  • Relationship building
  • Teaming
  • Co-creativity
  • Brainstorming
  • Cultural sensitivity

In her keynote presentation at STARCanada last year, agile consultant Janet Gregory talked about this when addressing the skills and attributes testers need to develop to be successful. The informal results that she gathered from her work with practitioners seem to confirm the results of the Towers Watson study.

Skills are the demonstrated ability to do something well, she said, something to be learned and practiced in order to achieve expertise. Attributes, on the other hand, are characteristics you ascribe to someone. They reflect things that they do or how they want to be seen.

The top nine skills, as ranked by Gregory's students, include communications, asking questions, giving feedback, and problem solving. All of those have more do with human skills than technical prowess.

Collaborate to succeed

Getting ready for a partnership with the machines is important, but it also requires us to become better humans. The good news for test engineers is that software development and testing has evolved to be more collaborative than ever. Agile, extreme programming, and paired programming are all about developing software socially.

We need to resist distractions. Just as the social dynamics of software testing is skyrocketing, technology in the form of our smartphones and social networking, is detracting us from our opportunities to continually improve our social skills.  Simon Sinek has a great video interview millennials in the workplace that highlights just how detrimental the cell phone has become to social interactions.

Machine learning, in the form of cognitive quality assurance and cognitive automation, is coming to software testing, so you should make every effort to improve what you can do on your side of the human-machine partnership.

Identify tasks that require quiet time to accomplish, and do those remotely. But for anything collaborative, do it face to face whenever possible. Even if you work remotely, try to get face time with your co-workers. Finally, continue learning the latest test practices and technologies, as well as how best to leverage machine learning.

In sum, if you want to succeed as a test engineer in the era of machine learning, learn how to build relationships, brainstorm, collaborate, and lead.

Want to know more? During my STAREast 2018 Orlando conference session, "What’s Our Job When the Machines Do Testing?" I'll offer more tips on how to work effectively with machine-learning tools. And be sure to use the promo code SECM for $200 off your registration.

Topics: Dev & Test