Quick note, software update 19 is still undergoing final testing, as a coterie of bugs surfaced from our recent testing. Fixing wouldn’t take too long, we’ll be done with a few more days.
Software update 20 in iOS, in addition to mirroring biofeedback training and other upgrades from update 19, would consist of a major upgrade to machine learning. Machine learning would now have the capacity to learn from what you teach her, in addition to from your physiology pattern.
Taking away the jargon - what this means is that, the machine learning would be able to understand you to a greater depth than previously possible. As a result she would give you better predictions. Predictions would become more stable and robust, because, previously, machine learning would have to make best guesses without any form of feedback from you, but now, it could integrate a new source of data - from you. This would allow her to cut a huge amount of degrees of freedom, to make much more robust predictions.
To translate to a relatable experience, when you train your FOCI’s machine learning after the automatic Machine Learning is completed, emotion inference would become more robust, less prone to noise, more accurate (hopefully, because it could be fun to deliberately confuse FOCI’s machine learning, as some of my colleagues are bent on confusing my FOCI. Thus it could technically be less accurate if you decided to play and confuse the Machine Learning). Luckily, we also designed in a reset button, to tell her that what was taught to her was false, and wipe that off the equation :)
Stay tuned, we hope to release this coming software update 20 mid-August.
Mick & The FOCI Team
FOCI’s Focus Training Manual is constantly evolving with FOCI, from many users’ feedback. We have the newest updated version here.