Ted Newcomb (tcn) Tue 25 Sep 12 10:53
David, this is sort of theoretical, but I'd be interested in your response. It seems to me that at the moment we have this huge amount of data being sifted both by humans and machines. I read on the Net today that 90% of the data on the Web has been created in the last two years. Yikes! Before we even get to the point of understanding and knowledge and wisdom we have two very different processes going on in the data sift. Humans and machines approach and process the bits differently, and maybe that's a good thing -- different perspectives, different assessments. I'm not thinking either we or the machines are arriving at any singularity any time soon, if ever. All that being said, what sorts of different knowledge networks might evolve, and how might they best interface? How do we take these differences into account? What's the near and far future for all of this?
David Weinberger (dweinberger) Tue 25 Sep 12 19:04
Ted, re: 76, you ask "What's the near and far future for all of this?" I have decisive and clear answers to both parts of this question: The near future will probably be pretty much what we currently have, and I have no idea what the far future will be. The easy part is why I have no idea. It's because the future exists to make fools of us all. Right now, it looks like we're going to become more and more dependent on machines to sift through the data for us, and to notice connections that we can't see. Machines are likely to scale at this faster than we can either as individuals (for sure!) or in various social formations. Of course, you never know what will be invented, but that's how it seems now. So, I assume that computers will come across more and more -- not fewer and fewer -- weird correlations, like the fact that if you're obese, your friends are more likely to become obese. And while a site like Galaxy Zoo may enable humans to classify millions and millions of galaxies as spiral or ovoid, I expect computers are going to be far far better at noticing farflung anomalies in the data coming in from radio telescopes and the like. As for the best interface among the knowledge systems, we're seeing strides in standardizing before the fact and in building systems that can make sense of huge, messy, ad hoc piles of data. In between these two poles, there's the Linked Data format that makes it much easier for computers to make connections -- crosswalks are the new metadata -- without having to get everyone to agree to standard vocabularies, etc. ahead of time.
Jon Lebkowsky (jonl) Thu 27 Sep 12 05:15
In these kinds of systems leveraging big data and big computing power, how can we best leverage the role of the human and the role of the computer? It seems to me that sense-making requires both, and I find myself wondering what systems and disciplines might evolve that we couldn't have imagined before we had all this computing power, and a network to connect all people and computers to each other. What new roles are emerging that weren't there before?
Ted Newcomb (tcn) Thu 27 Sep 12 05:21
Piggy backing on Jon's question, can you expand on the knowledge economy and the changes taking place within human culture? The old days were driven by individuals plying knowledge. Now it's all about collaboration with other humans as well as machines. What kinds of changes is that bringing about?
David Weinberger (dweinberger) Thu 27 Sep 12 07:12
Jon (re: #78), we've already seen new competencies, and new value arising from old competencies. We all know people who are unbelievably good at pointing us to links. (Sometimes this is a brain function, as with Peter Kaminski, to name one friend, and sometimes it's a brain + computer function, as with Jerry Michalski (to name another friend) who supplements his own massive brain with TheBrain.com to record his life notes. For another example: designing ontologies is a new skill, although not without precedents. Or, to take another random example, someone like Dan Brickley has an astounding ability to design systems that navigate gigantic clouds of linked data. Someone like Esther Dyson combines technical, business, and cultural knowledge with a social network that encompasses most of the tech world. You can ask questions of Peter, Jerry, Dan and Esther that you couldn't reasonably have asked of anyone before. But, without taking a single thing away from these awesomely brilliant people, in each of these cases, the role is enabled by the Net. We can ask your question, Jon, about roles by flipping it and asking what new services the Net is providing by enabling these roles. So, to get to your question: "[W]hat systems and disciplines might evolve that we couldn't have imagined...? What new roles...?" My answer is: I don't know. I really don't know what's going to happen. No one does. "Yeah yeah," one might want to reply. "But your just being coy. Or maybe a coward." But I'm not being coy. The best I can do is point to some traditional areas that are not currently well-supported. Would that help? - Taxonomies, nomenclatures, classification. Having common ways to refer to things is really helpful. We can make up for them to at least some degree by cross-walking and mapping. It's always going to be messy. The rise of unique IDs and namespaces is helping a great deal. - Filters. We used to not worry about filters because all we could get was the filtered product. Now we have to worry about them all the time. But we also now _filter forward_ rather than filter out: When the site TheBrowser.com puts together a front page with 10 items on it from around the Web, all the other items that didn't make it onto the front page are still fully available; TheBrowser.com has merely shortened the number of links it takes to get to its ten. - Consensus. We used to think that we "all" agreed on some things. We had authorities we "all" trusted. Now we have communities of belief. Links and conversation can help us get past the fragmentation that makes us stupid, but not past all fragmentation. But we should keep in mind that we've lost these old formations to a large degree because they don't scale, and because they presented themselves to us under false pretenses: they were never as baked into the world as they seemed. It's _our_ knowledge now.
Jon Lebkowsky (jonl) Thu 27 Sep 12 21:35
David, we want to thank you for joining us here at Inkwell.vue. This has been an excellent conversation (archived for sharing at http://bit.ly/2big2know). The conversation officially ends today; we're hoping you'll join us again! Thanks also to everyone who contributed. And we'll keep reading your posts at http://www.hyperorg.com/blogger/!
David Weinberger (dweinberger) Fri 28 Sep 12 05:17
Thank you! You Wellians ask wicked good questions! Thanks for the insights, the challenges, and the fun.
Ted Newcomb (tcn) Fri 28 Sep 12 09:53
Thank you! This was great and your book and work are valuable resources.
Craig Maudlin (clm) Fri 28 Sep 12 11:02
(fom) Wed 3 Oct 12 22:03
I hope you'll stay around to answer a few more questions. I'm still trying to think of a good one. Love the book -- and the other night when I had friends over, they pounced on it: "What is this about?" "This looks good!" etc etc, so I had to promise to lend it. Can you say more about that taxonomies category?
Members: Enter the conference to participate