I don't know why, but it took me until about halfway through the article to realize it was micro-interactions not micro-transactions. My brain just automatically read it as the latter.
I was picturing this dystopian scenario where clothes would be paid for through microtransactions using some kind of network aware fibers that would automatically charge your bank account per use or something.
This is interesting! Seems we can make everything interactive these days with a simple mix of electronics and ML. Discoverability might be an issue though, I already have trouble remembering all the gesture combinations on the Macbook trackpad.
It’s also not clear if the ML-based classification scheme can handle the vast individual differences among the general public, in real-life scenarios. (Wasn't this why Motion Sense on Pixel 4 wasn’t as precise as initially expected?) Performance may take a further hit with wear and tear, if these are embedded in clothing like hoodie drawstrings.
We're witnessing the birth of a new category of cybernetic languages (smart gestures? tech-tactile? techtile?). Glad that Google is doing their HCI research to figure out how the tech is bounded by the human body. I imagine that at some point, you could add vibrations as another information channel to supplement sight and hearing.
How long before we see competing gesture languages? How long before there's an industry standard?
You are not alone. I first read it as "E-Mail microtransactions", then "E-Mail Microinteractions". Just now I realize it's not about E-Mails. Not sure If I should be happy or sad now.
I think this is a fascinating question and I don't honestly have an answer to it. There is a lot of interest in the broader HCI community around wearable tech, embodied tech, and ubiquitous computing. Naturally Google - a major employer of HCI researchers - has a lot of people who are into this sort of thing. But that just pushes the answer back to why is the HCI community so into this? My best guess is that the HCI research does not really have a central canon or paradigm (a lot of people come from other departments (CS, Psychology, ...)). Consequently the community is very open minded about what constitutes valid research and creative expression and potential valued more than practicality.
I've spent some time in the HCI community, even going to a SIGCHI once, and my take is that input methods are a very wide open field. So it's easy to get work here.
Some of it can be downright sexy, and giving good demo gets you more prestige and money. I'm sure GATAP is no different.
What puzzles me is that we're certainly headed down new roads that other groups have paved already, like two-way audio interfaces and AR/VR. Tactile is great, but also presents a huge gap in affordances. Any interface you need to train for is a failed one. Anyone with an AirPod and Siri has moved a generation past a clickable headphone button. So why keep pursuing a one-dimensional interface method like a hoodie string?
I get that there's an application here in decoding waveforms using "AI" to translate signals to gestures. And maybe that's the point.
11 comments
[ 6.6 ms ] story [ 33.4 ms ] threadhttps://m.youtube.com/watch?feature=youtu.be&v=pbAvY6bwZD8
I was picturing this dystopian scenario where clothes would be paid for through microtransactions using some kind of network aware fibers that would automatically charge your bank account per use or something.
It’s also not clear if the ML-based classification scheme can handle the vast individual differences among the general public, in real-life scenarios. (Wasn't this why Motion Sense on Pixel 4 wasn’t as precise as initially expected?) Performance may take a further hit with wear and tear, if these are embedded in clothing like hoodie drawstrings.
How long before we see competing gesture languages? How long before there's an industry standard?
https://www.theverge.com/2017/10/6/16428338/google-atap-levi...
Some of it can be downright sexy, and giving good demo gets you more prestige and money. I'm sure GATAP is no different.
What puzzles me is that we're certainly headed down new roads that other groups have paved already, like two-way audio interfaces and AR/VR. Tactile is great, but also presents a huge gap in affordances. Any interface you need to train for is a failed one. Anyone with an AirPod and Siri has moved a generation past a clickable headphone button. So why keep pursuing a one-dimensional interface method like a hoodie string?
I get that there's an application here in decoding waveforms using "AI" to translate signals to gestures. And maybe that's the point.