I really love going on conferences. So often they are too commercial, you don’t always get what you expect even though you read the abstract a few times and carefully curate your program. Conferences are like multidimensional room where you learn, connect, take yourself out of your bubble and see the world from different angle. The question is whether you manage to see it through different lenses and diversify your experience or you remain focused on one specific thing.
I am going to WebSummit in Lisbon, Portugal regularly. It’s not strictly web nor tech conference as the name might suggest. It’s like a wide ocean where you can find anything – how technology is evolving and impacting every industry, how we as society adapting or maybe struggling with it, how current mainstream of technology creates next opportunity.
I work in software quality for many years. I love it, so I stay in the discipline for ages. I am also fascinated that the quality as subject is still a huge challenge that mainly discussed in testing community. What was interesting to see this time on Developer Summit stage that the key themes of the talks were shared between AI and testing. Yes, testing. I finally thought we were being heard. At that moment, I felt like I was on the other side of the room, the side where someone was pitching testing to me. I felt that the door of echo chambers, where we testing professionals stayed for a while, started opening. I couldn’t be more happy to have more people pitching for testing.
With absolute certainty I can say that last 5 years been hectic. Go to market approach was and still is primarily focused on speed. Deliver something fast to user’s hand and reiterate. Do we really have time to reiterate and improve? We often tested in production while sharing a playground with real users. We traded speed for quality for very long time and we tried to make it work.
At DeveloperSummit stage I came across not only one, but a few testing talks. That was unusual for a conference at such scale. That deserves our attention.
Antithesis is building a platform that solely focused on testing, helps teams find bugs in complex system exercising non conventional testing like property-based testing, fuzzing. Earlier that week I read a blogpost from iOS developer sharing his experience on parametrised testing in Swift and bringing examples of property-based testing. Honestly saying the majority of unit tests I have been through were example-based ones. You assert against the examples you created yourself defined in pre-computed part of the test. It was refreshing to learn something new, put it into my ever expanding learning list and try out at when I have some time.
Another talk that touched on testing in AI era is Zero-UI engineering. Essentially you build AI agent into your workflow that has an access to your API, event streams and non UI part of application letting them examine failure, suggest the fix and still keep human in the loop. You can try and play with the approach completing this challenge the speaker shared on the talk.
AI mass adoption is on the rise now. We passed the magic phase where we thought AI was something unreal. Soon we will step into utility phase if some of us is not there already. At some point it might bring sameness.
Quality is a mindset and strategy at the same time to stand out of crowded market of services
Nevertheless, quality will not magically follow if we aren’t designing for it, if we aren’t thinking out of the box how to failure proof it. Quality won’t magically follow if we deliberately don’t test our systems, if we don’t build systems that test our systems. Quality won’t magically follow if we don’t engineer system that are resilient, reliable and can recover in seconds when unexpected moment happens.
Nowadays we observe the speed is getting 10x faster. Quality is getting even more fragile. On the other hand, there is a glimpse that cuts through dense world of modern technology that quality is a differentiator. At least I see some signs of it.
The Roundabout Cafe 
Leave a comment