As I work on more and more Artificial Intelligence and Machine Learning engagements, I’m seeing more similarities to UX engagements than differences. Compare and contrast:
User Experience design is an iterative process. UX practitioners learn, define a candidate experience, test said experience with users, then tweak based on user feedback. You tweak, you test, you implement – and in an optimal application of this process, you start the whole thing over again so you can tune the experience after it is rolled out to the end user.
In an AI/ML project, you get relevant data, you analyze and “finesse” the data to align to a smart model that is “trained” to analyze it, and you then validate the model through testing. Once you implement, you constantly tune the model to become smarter and better at supporting a smart output of the model.
In both situations a key aspect is “Learn, Implement, and refine.” The only real difference is in the inputs and validation criteria – in one case the inputs are qualitative (user testing, feedback and general sentiment/response) and the other is quantitative (successful data analysis, predictions, and results).
The parallels are there, but more than that, I also see an opportunity. Too much of the UX design process is based on “soft” numbers – user feedback, standard usability scores, customer sentiment. And Machine Leaning is about the “hard” numbers – data points that can be processed in the billions, but without the humanistic touch that is seen in user centered design.
I can see a hybrid process, one that leverages the best of both, to bring forth technology solutions that aligns both with data and sentiment. I am trying to pursue this on projects that I am brought into a,d I hope you read this and start thinking about how you can bring a similar “all-up” perspective in any projects you work with.
Let’s leverage the best of everything to make the lives of users better.