I would like an algorithm or service that would suggest arguments, opinions, and points of view from smart people trusted within their communities but with whom I am likely to disagree or whose communities I am underexposed to. I do not think I am alone in this desire.
I attempt to get some of this out of who I follow on Twitter (and it was a great use for Google Reader -- may it rest in peace), but that is a pretty imperfect system. I also routinely ask others to suggest sources I might like to fulfill these needs, but I have found that many struggle to make good suggestions.
One of the tricky things about this algorithm or service is that it would need to distinguish between those arguments and communities that I care about, those that I do not, and those I am repulsed by. For example, I am probably underexposed to cricket enthusiasts but I don’t care much about cricket anymore and don’t want more information. Another example is that I have not read anything about the Parkland victims being actors conspiracy theories but I would be actively repulsed if a service suggested that I should read about it.
My suspicion is that one of the reasons services serve up filter-bubble content based on the engagement metrics of friend groups and similar users is because it is much easier than finding good, challenging material to suggest to users. That said, I wonder if the later might be more fulfilling to the user over the long term and result in a stickier service if it could be achieved.
Do you know of a service doing a good job of this? Do you have ideas for users or publications that would fit this bill for me? If so, please send them my way at @amac.
I went to the last Internet & Jurisdiction gathering in Paris. I can’t make it to the one that starts today in Ottawa, but I would have come if I could. I’ve been thinking about the last one all year because it was full of good, smart people trying to make progress on coherent and practical Internet jurisdiction. What I also loved about it was that I came away strongly disagreeing with the direction they were going. More on that below, but first some background.
Dwarf Galaxy Caught Ramming Into a Large Spiral Galaxy
(NASA, Chandra, 08/14/13) from NASA Marshall Space Flight Center.
This and the other space images accompanying this blog post appear
to be in the public domain, in spite of NASA's weird licensey language
to the contrary.
Background: Internet Jurisdiction
Jurisdiction is one of the oldest and thorniest questions for Internet policy: “Which government(s) get to regulate what and who, where?”
As John Perry Barlow put it in his 1996 manifesto declaring the Internet’s independence from government regulation, “[Cyberspace] is a world that is both everywhere and nowhere, but it is not where bodies live.” In that piece, he argued that regulation of the Internet by governments was both unwise and impractical. Others saw it differently. As Tim Wu wrote in 1997, “it is possible to regulate the Internet, and ... countries, corporations, organizations, and private individuals are already doing so.” The first important legal cases involving the extent of government jurisdiction over the Internet were decided shortly thereafter.
As the Internet has grown, become more mainstream, and increased in importance, particularly with respect to real world consequences that governments have historically regulated, questions of which governments get to regulate who and what online have become increasingly frequent. These questions get “answered” in courts, as governments make laws, and by corporations and individuals as the architecture, norms, markets, and regulation of the Internet develop.
There has been no straight line of consensus “progress” from one point of view to another. Even now, there are big questions that are being actively fought, including the United States Supreme Court considering Microsoft’s challenge to request by the United States for user data stored in Ireland, and the Supreme Court of Canada asserting the ability to order content removed globally only to have a U.S. District Court disagree.
The technological landscape has also changed dramatically. Over the last twenty years, as billions of people started using the Internet it has morphed from an incredibly decentralized landscape of personal websites hosted from tiny service providers, often at the very edge of the network, to a more centralized set of cloud-storage service providers serving a large percentage of the population. If the FBI wanted to find out whether I had sent an email to a particular person in 1996, they would have had to come to my house to get my computer and take a look at my locally stored email, if I hadn’t already deleted it. Today, all my email is on Google’s servers, just like that of more than a billion other people from all over the world. The public content I created used to be housed on a server in my closet. If someone thought I was saying something illegal, they would have likewise most likely have had to come to me in order to get it removed from the Internet. While it is true that, in some situations, some other avenues existed to get my information or remove my content, they were not very broadly available or used. By contrast, now, most of my online content is served from large U.S. corporations, like Google and Github. If they decide my content shouldn’t be online, they can remove it and force me to go look for another publisher. In some cases the online service providers are so important that no suitable replacement would exist.
Towards a coherent, if abhorrent, Internet jurisdiction policy
The Internet & Jurisdiction Conference (I&J for short) focuses on three broad tracks: data, requests for private user data; content, requests to render content inaccessible; and the internet domain name system. I’m most interested in the first two and these comments are mostly meant for them.
I&J had a wide variety of participants and many more government and law enforcement types than I generally find at the Internet policy conferences. The conversation was therefore more oriented towards those stakeholders than at some other conferences, and it was quite similar in tone to the types of conversations happening in governments and courts all over the world right now.
In both the data request and content removal areas, these conversations are moving towards a coherent, if abhorrent, policy of allowing governments almost everywhere to get data about any internet user or remove any content without needing to engage the users themselves or the court systems of their jurisdictions. Most discussions exclude certain governments from the club that should have this type of power, but the idea that data should be able to be given over and content should be able to be suppressed through interactions between governments and repeat-player intermediaries was so ingrained in many of the discussions as to be an assumption. Convenience and speed are touted as principal advantages.
For example, a Facebook user in Mexico should have their data given to authorities in the England on a request to Facebook. A Canadian Microsoft user should have their post suppressed, at least in Thailand if not all over the world, via a request to Microsoft. Even if the user is known to the complainant, no direct approach to them is contemplated. At some companies under some circumstances the user might get a notice, but that is left to the companies and to the circumstance. This is not just “All your base are belong to us” but “All your base aren’t ever belonged to you in the first place.”
A challenge: Center Internet policy on users citizens
If I were able to come again this year, I would. Indeed, my favorite conferences are those at which a majority of the attendees are smart, passionate advocates with direct experience with the subject matter and with whom I disagree (see e.g. the excellent Fordham Intellectual Property Institute). If I was there, I would challenge the attendees to propose a way forward that centers on the user rather than removing them from the equation. It is more convenient to just go to the big corporate repeat players. They are well known to the governments and can be counted on to pick up the phone. However, an Internet jurisdiction policy that regularly circumvents the user will encourage countermeasures to return power to the user -- the emerging prevalence of end-to-end encrypted services is one good example of this trend. More importantly, those users are our countries’ citizens, they deserve our respect and, at least, to be able to face their accusers and challenge the accusations. There may be cases where expediency trumps all, but this is the tiny minority of cases, not the norm on which policy should be based.
The YouTube clip above is of me trying to make a similar point at the end of last year's I&J.
PS If you want to read more about the current intermediary liability battles, please follow Daphne Keller and Eric Goldman and take a look at their excellent sets of resources on the topic at Stanford and on Eric’s blog. Graham Smith also wrote a couple of posts in the run-up to this year’s I&J.
Many of us have more than a few screens. Mostly they lay idle. A tablet for plane trips, an old phone, a Chromecast connected LCD TV.
I also have a bunch of images I want to see more of. More than 100,000 personal digital photos. And, lotsmoreincredibleimagesout in the world. Millions.
One of the things I loved about the Obama White House was that Pete Souza and his team's wonderful photos were everywhere and changed relatively frequently. Walking to a meeting you'd see a picture of a co-worker and her kids hugging the President. In a meeting room there would be a beautiful photo of Half Dome in Yosemite and Marine One. It was great.
President Barack Obama visits with Natalie Quillian,
Advisor to the Chief of Staff, and family in the Oval Office,
Aug. 27, 2014. (Official White House Photo by Pete Souza).
Public Domain.
I'm working on using idle screens to bring more of that into my home with my images and those from incredible online collections. The idea would be that the screens could show a playlist (automatically generated or hand-curated) of images from a wide variety of sources and on a wide variety of devices / screens. Incidentally, I haven't been able to find the equivalent of .m3u for images. I don't think it exists. It would be wonderful if people could trade image playlists.
Automatically generated playlists could show images relevant to the day of the year, or types of images, or ones that will look good on that particular screen, or ones I might like based on what I've liked before. An earlier version of this was surprisingly good if it just showed images from Christmas, New Year's and Thanksgiving.
Solar Flare, August 31, 2012, Nasa Goddard Space Flight
Center. Hosted by Wikipedia. Public Domain.
I don't know of a thing out there that does this across image collections. Do you?
I mostly program to learn things. This project will help me really learn the techniques I've been studying through the Andrew Ng's Coursera on Machine Learning. It will also brush up some of my full-stack web development skills as the database will be Mysql, the backend will be Python 3 and Django 2, the frontend will be HTML and CSS with a lot of Javascript (including ajax, which I haven't used too much). And, I'll get to know both JQuery and Cycle 2 well. It has been fun so far and I will share updates as I go, and, eventually, the code.
I'm not doing this as an "entrepreneur." The idea isn't to make a bunch of money or get a million users. Mostly I'm doing this for myself and to learn. I have found that deeply understanding technology is really helpful to my legal and policy work (not to mention my life). I also really enjoy coding for fun. If you haven't tried that, you should!
I'm writing about it here to further commit to finishing it and so that others can share their good ideas.
153rd New York Infantry, ca. 1861, from Metropolitan Museum
of Art. Hosted by the Internet Archive. Public Domain.
If you are interested in using or working on this, please let me know. I'd be interested in understanding other use cases. I would also love pointers to great image repositories (preferably public domain). And eventually, I'd like people to help me rate the pics from some of the public repositories. My email is "lawyer" @ the popular email service run by Google. Or you can @reply me on twitter @amac.