Re: Vote for my Semantic Web presentation at SXSW

On 2011-08-16, Juan Sequeda wrote:

> In the past two years, I've tried to get people together to submit 
> panels and presentations about Semantic Web to SXSW. Unfortunately, it 
> has barely been successful. [...]

I think it's not successful because the Semantic Web itself is not 
successful. It still lacks a killer app, and the integration, and 
especially the visual candy that rules over everything else, 
adoption-wise. It's still a solution in search of a real problem.

One of the surest signs to me is that pretty much every SemWeb 
presentation I've seen a) starts with the same, already-tired, academic 
litany of theoretical promises, a layer cake or whathaveyou, and b) is 
presented by somebody supported by a grant/tenure/government 
salary/whatever. I see absolutely *no* stuff from private, venture 
funded entrepreneurs which tell me they successfully solved a pressing, 
real life problem using SemWeb technology, and because of the tech, more 
rapidly retired with a hefty trust fund.

Because that, honest to God, is the only criterion of a Solution. It's 
the criterion *even* if the technology was primarily poised to solve a 
problem of a public goods nature where you have to go through the nasty 
gymnastics of convincing a government to make its data open, and linked. 
That's just not going to happen unless the private sector is already 
thriving around your data model, vision, solution, usability and 
consumer candy-appeal. What instead happens is that you flat out lose to 
PDF (textual description of your data), and in particular to Facebook 
(dynamic, social description, again over unstructured text).

Now, I'm not saying SemWeb is dead in the water. Far from it: I'm a big 
believer in the basic principles of it. But as of now, the focus remains 
totally wrong. First, FOAF has lingered on as a potential killer app for 
a while, and stagnated. Second, I'm seeing no Android/iOS/HTML5 apps 
which make serious use of the semantic web, *while substantially and 
measurably benefiting from it*. Third, that's prolly because the 
plumbing isn't there or is too heavy to be deployed incrementally and/or 
cheaply. Fourth, heavy duty data really doesn't sit too well with the 
basic encodings like RDF/XML; or would you happily run your production 
database over it/them? Fifth, where's the truly transparent and 
user-satisfying integration with established media? Et cetera, ad 
infinitum.

The semantic web holds great promise, but it always has been and sadly 
seems to remain more of an academic exercise than something truly 
practicable and profitable. More a tentative solution to a hypothetical 
problem, than a real solution to a pressing need. Then, it stagnates for 
lack of profitable investment, as it has for its entire duration. Like 
some relic, preserved by W3C's saving graces or reverence to TimBL The 
Great Weaver.

I think instead we should have a fast and dirty triple serving protocol, 
or perhaps even a protocol which breaks with the triple model as such 
for efficiency. Then a flashy app for distributed social networking, 
based on some revived derivative of FOAF, on *all* of the app stores 
around. Embeddable and integrable. That'd already go *miles* towards 
adoption. Then do the same for the rest of the Linked Data.
-- 
Sampo Syreeni, aka decoy - decoy@iki.fi, http://decoy.iki.fi/front
+358-50-5756111, 025E D175 ABE5 027C 9494 EEB0 E090 8BA9 0509 85C2

Received on Wednesday, 17 August 2011 00:16:05 UTC