4 ms·
> Machine learning is used to help the iPad's software distinguish between a user accidentally pressing their palm against the screen while drawing with the App
by bambax 6y ago
> Machine learning is used to help the iPad's software distinguish between a user accidentally pressing their palm against the screen while drawing with the Apple Pencil, and an intentional press meant to provide an input. It's used to monitor users' usage habits to optimize device battery life and charging, both to improve the time users can spend between charges and to protect the battery's longterm viability. It's used to make app recommendations.
The problem with machine learning altering the behavior of a device is it shortcuts human learning. The human brain is very good at learning deep insights about things and its environment and alters its behavior accordingly.
If things change while we're learning about them, it confuses and upsets us. A dumb machine is much easier to use than a "smart" one.
- kumarvvr 6y agoThis might be true for mostly user interface stuff. But for things that are happening under the hood, a human mind will get accustomed to a gradually improving user experience. Apples mantra seems to be to make everyone use the device as per their whims and fancy, and let the device figure out how to deal with it. Ultimately what this leads to is a user experience lock in, where other devices that dont adapt feel clumsy or stupid.