This is an exceptionally effective demonstration of some of the things websites can do to draw conclusions about you. How Normal Am I? is an art project/tech demo website by Tijmen Schep that I found on Metafilter. It says it's using face-recognition to gauge how "normal" you are but as the Metafilter post says, "At least that is the link-baity hook. What it is really doing is giving a short but effective demonstration of what is possible in the field of recognizing and categorizing people." It's definitely worth trying out, but be warned: as the first comment notes, "this turned the camera on on my laptop, then never turned it off."
This isn't a bad article, though of course an ethical approach is going to take much more than can be outlined in a simple article. Reid Blackman argues against three common approaches: the academic approach, which isn't business-focused; the on-the-ground approach, because practitioners lack ethics expertise; and high-level principles, because they are too vague. Instead, he outlines "building a customized, operationalized, scalable, and sustainable data and AI ethics program." The problem with this approach - and yes, there is one - is that you could follow all the steps outlined in the article and come up with a risk-management framework that has nothing really to do with ethics. That might in fact be what businesses want to do - but then they shouldn't call it 'ethics'.
This post is an introduction to a short report (6 page PDF) on a company called Avature, which provides recruiting and onboarding technology for large companies. It reads a lot loke marketing content, so take it with a grain of salt. It's worth noting here because of the role recruitment tools play in the online learning ecosystem - after all, for many people, the output of online learning is the input for recruitment (put that way it sounds really crass, but that's the way to think of it). And the caution here is to avoid designing for specific recritment objectives - becuse the goal of recruitment technology is flexibility. This is going to be especially the case for competency requirements and job qualifications. "As a company starts to tweak, adapt, and expand its recruiting practices, there are always new workflows, business rules, and reports to add. Recruiting is not a transactional process – it’s more like a continuous flow – so the platform has to be adaptable and easy to change."
A common problem with privatized public services is the need for stringent oversight. Private companies are dedicated to making a profit, which means that the public interest might take second place. We see that problem evident in the world of non-degree certifications. "There’s really no centralized resource for finding out about them," says Martin Kurzweil, "and a lot of providers don’t participate. It’s difficult to get quality information, so people find information directly from providers, creating a situation that leaves learners vulnerable to fraud and abuse." We might perhaps see some short-term solutions, such as the Credential Engine Registry, but this just pushes the problem back a level. Over time, what we need to anticipate and plan for is the erosion of the value of certifications generally (with perhaps the exception of major brands like MIT and Google).
Like Doug Belshaw, I take a really hard line on notifications - I accept them for text messages and Strava, but that's it. Why? Well as Doug Belshaw says, our "‘social norms’ have been shot to pieces by companies employing machine learning to growth hack the human brain." In other words, he says, "they’ve optimised engagement with their platforms to such an extent that it bypasses human rationality." Don't let them into your head! Some things you can do: "switch to the mobile web version of social networks rather than having a native app installed." And when "prompted to allow notifications do not allow them." Also, I would add, in Android go to 'Settings > Apps & notifications', where you can turn them all off in one place. On an iPhone it's harder, but you can still turn off notifications.
This newsletter is sent only at the request of subscribers. If you would like to unsubscribe, Click here.
Know a friend who might enjoy this newsletter? Feel free to forward OLDaily to your colleagues. If you received this issue from a friend and would like a free subscription of your own, you can join our mailing list. Click here to subscribe.
Copyright 2020 Stephen Downes Contact: firstname.lastname@example.orgThis work is licensed under a Creative Commons License.