When it comes to releasing new features and products, many companies today practice “release and forget.” This practice is about to change in a most radical way because, for AI-driven products, GA is only the first step in the lifecycle of a learning product.

The Learning Product

UX leaders have always advocated that releasing a product or a feature is just a starting point in development – that’s where the real work starts. However, most companies have failed to implement this practice. This is all about to change with the release of more AI-driven products. In the past, we have covered the new RITE methodology with the AI-Data component here: https://www.uxforai.com/p/the-new-ai-inclusive-ux-process It bears repeating: certain aspects of the new AI capabilities make it imperative to take AI learning strategies and resulting product evolution into account and make it the central premise of user-centered product development:

As depicted in this diagram, the AI-Data component of the development is just as crucial after the release as it is during the RITE phase of product development. The system's usage by customers generates more data, necessitating AI re-training, which leads to the emergence of new inputs and user behaviors in the system, thereby generating more new data. This AI-Data learning component is a significant factor in the product lifecycle, presenting a real and tangible change. It's not a consultant’s cliche like 'thinking out of the box.' In essence, it's a new opportunity for your company to succeed in the AI-driven normal, one that you can actively manage and nurture.  

To use the DnD LitRPG analogy, going GA is like playing Baldur’s Gate and just getting out of Candlekeep. That means you have just unlocked your power and experience progression. GA is not the end of the lifecycle; it’s literally the beginning of the real game when the training wheels come off.

The Learning Enterprise

In the olden days of yore, when mobile tech was just coming into its own, many enterprise companies ignored the changing product landscape. Companies like SAP, Oracle, BMC Software, IBM, and many others have initially ignored the new “mobile movement.” Many product leaders at these companies were known for saying that “mobile is fine for ‘consumer’ tasks like email, shopping, and social media, but ‘real’ enterprise work was going to be done on their platforms for many more years without any significant changes.”  

How wrong they were. 

Mobile has very quickly changed the standard of usability, flow, ease of use, and even patterns of use of all software everywhere. Early adopters of “mobile first” thinking were also huge beneficiaries. Companies like Amazon and Salesforce have quickly recognized both the opportunity and the inevitability of this systemic product strategy change and have quickly adapted to create modular software designs and app markets that allow customers to purchase only what they wanted, configure new modules with a few clicks without writing any code, and pay only for what they used. It gave rise to cloud computing, multi-platform inter-operability, and vastly improved usability and a self-learning model for enterprise software consumers. This was a revolution that rocked the industry to its core. In order to survive, most enterprise companies have been forced to adapt to the same paradigm quite quickly, often in a matter of just a few years. 

Now, we are at the cusp of another such systemic change. The AI-driven enterprise is here, and the change is happening even faster than last time: the transformation that took years last time is now measured in months. Software that is evolving and learning, not just allowing a few admins to change some configuration, is the new norm. Companies that do not embrace this new “Learning Enterprise” model are going to be left behind. 

To continue with another DnD LitRPG analogy, the software is no longer a plumb, square house where admins are adding rooms and air-conditioning units. Software is not a thing built of dead wood and concrete, measured, processed, ossified. Modern AI-driven software is now more akin to a treehouse of the Elves: a living, growing, evergreen structure made of trees, always growing in harmony with people, and business challenges, not a dead thing made of lifeless wood and stone. (NOTE: while we are talking about AI living in harmony with humans as part of the AI-driven enterprise, many folks are worried about AI’s impact on our actual, you know, environment. You can read about it here: https://www.scientificamerican.com/article/ais-climate-impact-goes-beyond-its-emissions/)

This means that we can no longer count on the walls around us to be the same plumb and square walls (that may be painted a different color this year or reskinned with whatever skeuomorphism or material design de jour happens to be.) We need to start treating software as a living, evolving entity that learns, changes, and adapts over time. Learning and self-healing systems, made with half cunning exquisite elvish craft and half with magic – learning systems exist in harmony and for the support of the people who inhabit them. 

The change is profound and significant. To truly succeed in this new normal, it will take some new skills and deep organizational changes. It is no longer enough to just “be OK” with living in a living treehouse. We must embrace, enjoy, take pride in, and lead this organic evolution as a matter of survival. Marcus Kirsch, the author of The Wicked Company (https://www.thewickedcompany.com/), writes eloquently about the new companies having to “surf the wave of change… or be swept under” – I agree. Which is why we are going to continue discussing and delivering thought-provoking content and practical advice for the fast-evolving UXers who are leading this change. 

Because UX is the surfboard your company needs to stay on top of this wave.

Surf’s up!

Greg Nudelman (with Daria Kempka, Contributing Editor)

P.S. Our design problems are getting super interesting. For instance, when we moved from print to the web, it was a new paradigm of designing for an interface that could change how the interface would display: the size, the colors would be different on different screens, etc. Things kept evolving from there to the point that we had responsive design and dynamic content that would change based on what you’d looked at before or on some other info that was collected about you. Now we’re designing something that can design itself. Like…whoa. It’s hard to even fathom. This is why we appreciate that you chose to subscribe to our newsletter and receive actionable ideas in your inbox. We hope that if you enjoyed the article, you consider sharing it with a friend or collegue using the link below. When you do, we’ll be happy to send you a free book (retails for $20). Just click the link below. Thanks.