Replied to Issue [#327]: Happy New Year! by Doug Belshaw (Thought Shrapnel)

I had a bit of an epiphany when I realised that one of the main uses of AI is, or will be, voice assistants. In practice, that means a lot less time spent looking at displays and a lot more time interacting with devices using natural language. To do that, voice assistant need to know the context in which you operate, so they need to have data on you.

I'm not delighted to be handing over so much data to Google, but given their GDPR-compliant controls, I'm willing to give it a try. The Lenovo device is in our kitchen and has replaced our DAB radio, and the Onkyo speaker is in our bedroom. Both of them have hardware switches which mute the microphone when they're not being used.

I really am not sold on all this move to smart devices Doug. My wife recently purchased an iWatch and has taken to messaging directly from it. I now need to check if she is talking to her phone, watch or me. I have also noticed this on public transport. I have two particular reservations:

  1. What if everyone was talking at once? What would that look and sound like?
  2. What about the conversations that may not be appropriate for speaking out loud in public or in private.

I respect there are some who see such constructive uses as a God send (read Richard Wells reflection), however this depends upon an appropriate space.

My other question is uses beyond the novel. Yeah I can ask Google a question or play a track from The National, but what else? I am really interested in what particular workflows you develop in conjunction with your smart things.

NOTE: I have written this response in the open web and respect your desire to restrict such conversations to paying subscribers, which I am not one, sorry.

Liked The Next Data Mine Is Your Bedroom by Sidney Fussell (The Atlantic)
Just this month, the insurance company United Healthcare began partnering with employers to offer free Apple Watches to those who hit certain fitness goals. Insurers might also offer benefits to residents whose homes prove their fitness or brand loyalty—and punish those who don’t. Health insurers could use data from the kitchen as a proxy for eating habits, and adjust their rates accordingly. Landlords could use occupancy sensors to see who comes and goes, or watch for photo evidence of pets. Life-insurance companies could penalize smokers caught on camera. Online and in person, consumers are often asked to weigh privacy against convenience and personalization: A kickback on utilities or insurance payments may thumb the scales in Google’s favor.
Liked Facebook - Trust us! by Daniel GoldsmithDaniel Goldsmith (View from Ascraeus)
Facebook - sure, we may have sold your most intimate data to the Russkies, installed a cryptofascist in the whitehouse, engendered genocide in Myanmar and the slaughter of hundreds of innocent people across the developing world, and (just this last week) got caught leaking user data of at least 50,000,000 people, but you should totally allow our always-on microphone and camera into your home! Trust us!
Liked Article 13 makes it official. It's time to embrace decentralization by Ben WerdmüllerBen Werdmüller (Ben Werdmüller)

Although it uses incredibly imprecise language, it can be reasonablly inferred that the directive targets large service providers like Google and Facebook. It doesn't target small communities or people who are independently hosting their content.

...

All of which means that peer-to-peer decentralized social networks are exempt, if you're hosting your profile yourself. Nobody on the indie web is going to need to implement upload filters. Similarly, nobody on the federated social web, or using decentralized apps, will either. In these architectures, there are no service providers that store or provide access to large amounts of work. It's in the ether, being hosted from individual servers, which could sit in datacenters or could sit in your living room.

Replied to Maths eats robots for breakfast - Issue 83 - Dialogic Learning Weekly  (Dialogic Learning)
Most of my week has been spent thinking about, advising on and reviewing future school designs. I have noticed the rising influence of the interior design of workplace on the aesthetic of secondary and senior learning spaces. It reminds me of this article outlining how WeWork (a co-working business) designs spaces using rich datasets and machine learning. I wonder if future schools will have responsive learning spaces based on similar sets of data about usage and pedagogy? It is not such a big leap, my home thermostat continually learns the patterns of how we heat the house and creates a schedule for us. Imagine a campus that can respond in a similar way to the patterns it predicts from how we use it.
Another great newsletter Tom.

I remember Ross Halliday focusing on what might be deemed as ‘IoT for education’ at GTASyd. It is an interesting space. I can see the potential for it in education, but at what cost? For what impact? Here I am reminded of Marshall McLuhan’s tetrid:

  • What does the medium enhance?
  • What does the medium make obsolete?
  • What does the medium retrieve that had been obsolesced earlier?
  • What does the medium reverse or flip into when pushed to extremes?

I recently finished reading Ben Williamson’s book on Big Data in Education. Although not solely on this topic, definitely relates and worth reading.

Counter-surveillance

Jim Groom reflects on the challenges of data surveillance for open education. The solution that he, and the team that he was collaborating with, came up with was that we need a form of counter-surveillance to take power and ownership back.

The only way to challenge surveillance is through counter-surveillance Source

It is interesting to juxtapose this with a comment that Mark Burden recently made that it is the Internet of Data Collection Instruments.

In terms of the device collectors, in some ways they are delighted about this passivity because it reveals behaviours that we wouldn’t necessarily reveal if we knew data about us was being recorded. So in that sense when you think about what is now called the internet of things, the very label ‘the internet of things’ is a misleading label, in fact it’s a label that I think should be put in a wastepaper basket. What we are really talking about is the internet of data collection instruments. And these instruments rely on our passive behaviours in order to collect the data from the environment and about us in relation to what we do in those environments. And what we are now starting to see is that the smart home, or what is becoming increasingly the smart home, is being packed with these devices.Source