That’s the biggest worry, certainly, but I think it’s very possible to get this right. We do not yet know how this feels, but one positive indication that they

That’s the biggest worry, certainly, but I think it’s very possible to get this right. We do not yet know how this feels, but one positive indication that they payed lots of attention, to exactly this, one of the two central riddles to solve to get this tech right (the other being the places in the UI you use it and to what end), is that implementation details in the UI also telegraph how they want to communicate this feature. So, first of all, there are three levels and the first level is a peek, nothing serious and totally non-committal. Frequently accidentally engaging this would probably still be super annoying, but if it happens once in a while and is also super responsive then that’s not a big deal. Making this first step non-committal is probably a big part of making this all work. Second, the peek seems to show off some direct input based on force (with the preview dynamically resizing based on it), so that you are not blindly guessing the pressure you have to apply. This direct feedback is probably also a big part of the puzzle. If they get that right and given their implementation (distinct, few and pretty clear and consistent places to force touch, at least within their applications, that all do predictable and understandable things) I think this is a winner and will make using everything much more of a joy to use. It could also suck. We don’t know yet. I’m just really happy someone is trying stuff like that. The vocabulary on touch devices we have is nice, but it is still sparse. Maybe this extension will work out, maybe it won’t. Either way, I’m excited.