On Fri, Oct 20, 2017 at 5:56 PM, Andreas Kolbe jayen466@gmail.com wrote:
I think it would be good to do some legal work to gain that clarity. The Amazon Echo issue, with the Echo potentially using millions of words from Wikipedia without any kind of attribution and indication of provenance at all, was raised on this list in July for example.
There is some basic attribution in the Alexa app (which keeps a log of all transactions). As I said, I don't see a reason not to include basic attribution in the voice response as well, but it still seems worth pointing out. Here's what it looks like in the app (yup, it really does say "Image: Wikipedia", which is all too typical):
I'm all in favor of a legal opinion on bulk use of introductory snippets from Wikimedia articles without attribution/license statement. While I'm obviously not a lawyer, I do, however, sincerely doubt that it would give you the clarity you seek, given the extremely unusual nature of authorship of Wikipedia, and the unusual nature of the re-use. I suspect that such clarity would result only from legal action, which I would consider to be extremely ill-advised, and which WMF almost certainly lacks standing to pursue on its own.
If CC-BY-SA is not enforced, Wikipedia will stealthily shift to CC-0 in practice. I don't think that's desirable.
Regardless of the legal issue, I agree that nudging re-users to attribute content is useful to reinforce the concept that such attribution goes with re-use. Even with CC-0, showing providence/citations is a good idea.
An interesting question to me is whether, with the explosion of information available, people will spend so much time with transactional queries across a large number of diverse topics that there is little time left for immersive, in-depth learning of any one of them, and how that might gradually change the type of knowledge people possess (information overload).
It's a fair question; the Internet has certainly pushed our ability to externalize knowledge into overdrive. Perhaps we've already passed the point where this is a difference in kind, rather than a difference in degree, compared with how we've shared knowledge in the past; if [[Neuralink]] doesn't turn out to be vaporware, it may push us over that edge. :P
That said, people have to acquire specialized domain knowledge to make a living, and the explosive growth of many immersive learning platforms (course platforms like edX, Coursera, Udacity; language learning tools like Duolingo; the vast educational YouTube community, etc.) suggests that there is a very large demand. While I share some of your concerns about the role of for-profit gatekeepers to knowledge, I am not genuinely worried that the availability of transactional "instant answers" will quench our innate thirst for knowledge or our need to develop new skills.
I'm most concerned about information systems that deliver highly effective emotional "hits" and are therefore more habit-forming and appealing than Wikipedia, Google, or a good book. The negative effect of high early childhood TV use on attention is well-documented, and excessive use of social media (which are continuously optimized to be habit-forming) may have similar effects. Alarmist "Facebook is more addictive than crack" headlines aside, the reality is that social media are great delivery vehicles for the kinds of little rewards that keep you coming back.
In this competition for attention, Wikipedia articles, especially in STEM topics, have a well-deserved reputation of often being nearly impenetrable for people not already familiar with a given domain. While we will never be able to reach everyone, we should be able to reach people who _want_ to learn but have a hard time staying focused enough to do so, due to a very low frustration tolerance.
I think one way to bottom line any Wikimedia strategy is to ask whether it results in people getting better learning experiences, through WMF's sites or through affiliates and partners. Personally, I think the long term focus on "knowledge as a service" and "knowledge equity" is right on target, but it's also useful to explicitly think about good old Wikipedia and how it might benefit directly. Here are some things that I think might help develop better learning experiences on Wikipedia:
- a next generation templating system optimized for data exploration, timelines, etc., with greater separation of design, code, data and text - better support for writing/finding articles that target different audiences (beginners/experts) - tech standards and requirements for embedding rich, interactive "explorable explanations" beyond what any template system can do - commissioned illustrations or animations for highly complex topics (possibly organized through another nonprofit) - assessment partnerships with external groups to verify that learners get what they need from a given resource
In practice, this could translate to:
- beautiful animations illustrating concepts like the immune system, the Big Bang, or the inner workings of different engine types - custom interactive explanations for concepts in statistics or mathematics, such as the ones in http://students.brown.edu/seeing-theory/ - code that you can interact with in articles _about_ code like https://en.wikipedia.org/wiki/Quicksort - highly visual explorables for topics that benefit from it -- Thedore Grey's award-winning "Elements" app is a nice example: https://www.youtube.com/watch?v=7FesjAdIWBk - better ways to go from one article to the next: data visualizations, topic maps, dynamic lists, etc.
The reason I think this matches well with what's stated in the strategy is that it's clear that Wikimedia cannot do it alone. Many interactive applications will require the kind of open data platform that Wikidata will hopefully become. Revision metadata APIs (with some form of write access) may make it easier for folks to help with the assessment of content quality.
The international education space (schools, colleges, unis) may often seem intractable and difficult to navigate. But from what I can tell, there's been a slow and steady shift away from crappy Flash/Java applets to more reusable HTML5 components and open repositories. The value of open licensing has become increasingly apparent to countless public institutions.
By sharpening their own role in these networks, WMF and other movement organizations may be able to positively influence decisions on questions like licensing, internationalization, and technology choice.
Since we last discussed this, I've come across a great research paper on Meta, "Considering 2030: Future technology trends that will impact the Wikimedia movement", prepared for WMF by independent consultants Dot Connector Studio (Philadelphia) and Lutman & Associates (St Paul):
https://meta.wikimedia.org/wiki/Strategy/Wikimedia_movement/2017/Sources/Con...
The sections "Things to keep in mind" and "Questions for the Wikimedia movement to consider" most closely reflect my own concerns.
I agree with the authors of this paper that WMF should carefully position itself between early adopter and "laggard" when it comes to new tech. Finding ways how tech can aid learning/collaboration, and become part of the commons, turns WMF into a leader from the perspective of many other organizations that are concerned with delivering knowledge and learning, and a follower from the perspective of tech companies. It's a special place to be. :)
Erik