Unlock true productivity gain in AI with quality
A Japanese proverb says: "If it looks like miso and smells like miso, then it is not the best miso." As for any proverb, there can be multiple explanations about what it 'truly means'. One could say that it talks about how true quality is found beyond appearances and you shouldn't concentrate on stamping it on things at a later stage. It could also be interpreted like: Art that stands the test of time is often concealed under apparent crude simplicity.
When we find something exquisite, we might want to explore how it's built and try to do it ourselves. Following the technical steps, however accurately, for creating your favorite dish can often lead to disappointment. Being able to consistently deliver great products that hit the right user experience at the right time is nearly impossible. When we come up with something great, we must be able to ensure its quality in different circumstances, as well as over time. The tools and processes we use for creation must support getting quickly rid of things that don’t work and preserving those that do. They form a delicate constellation that must be cared for.
Our digital world is enormously complex. Companies use hundreds of applications to support their business endeavors. The cloud native computing foundation (CNCF) landscape consists of nearly 200 projects right now, many of which are relevant to a professional in IT. The cognitive overload we experience as creators and users of digital services is running out of control. We have looked to all kinds of app wizards, search engines, and most recently generative AI tools to simplify and speed up our daily tasks in this ever-changing world.
Right now, the software industry is on the brink of oversaturation. Companies drive for ultra-productivity with all the new GenAI tools at their disposal. We must remember that when activities scale up, writing new feature code is only a fraction of the value stream in the digital product lifecycle. Most of the time is spent in taking care of that delicate constellation we use for creation and daily operations. This includes dealing with infrastructure, supporting the operative functions of the company and making sure that our users and creators stay happy as everything around us keeps changing.
Pushing for faster code creation and value delivery without a clear strategy for quality improvement will ultimately lead to new problems. Our work at Hidden Trail is to uncover how GenAI tools boost productivity in the long term. We look for concrete ways to improve the quality of the product, the process and especially the experience of users and developers. There are ways to cross-check and develop business requirements, R&D specifications, test strategies and insights from operations to achieve holistic views into what we are (and should be) doing. Balancing the right metrics of productivity, quality, agility and predictability help us steer business transformation initiatives with facts.
Technical capabilities are evolving at speed right now and we are all responsible for trying them out, providing feedback and helping them improve. We can't rely on verifying quality against what we thought good looked like before, or is generally considered as the norm. We must demand better than average, which is by design what generative AI output provides us. Quality provides us a competitive advantage instead of being just a cost item, as long as we are ready to give it the attention it deserves.