Re: [kmgov] Big Data event announcement on-line

Greetings,

Please forgive the abrupt interruption of threads-- hope it's taken with the 
good intention it is offered.

Rob Neilson forwarded the Big Data webcast to me but I couldn't make it --  
will try to catch archive in next few days-- it was on one hand good to see 
and on another dissapointing that so little innovation has been achieved in 
improving R&D methods-- especially diffusion. We seem to call for 
collaboration without considering the needs of the would-be collaborators. 
As I shared with a cyber security lead for a major bank this week

-- "We haven't modernized the R&D structure to current millennium, for 
example, but expect different outcomes"  -- much the same could be said 
about the semantic web. Perhaps if shared publicly some good might accompany 
any arrows.

In reviewing archives of list thought I would join again as I see some 
issues that are obvious to me that may not be to others given three decades 
on the adoption side of tech-transfer, but also frequently discussing policy 
behind the scene. That we are still calling for agencies to make data public 
after all these years that isn't a security or legal concern is fairly 
amazing to me, and speaks to some of the macro economic challenges we are 
facing as a culture. I fear many are too blinded by our own passion and 
interests.

"A criticism voiced by detractors of Linked Data suggest that Linked Data 
modeling is too hard or time consuming."

I thought an exceptionally rare quote on this issue came from James Hendler 
recently that is I think worth investing a bit more time on -- primary 
reason for posting today -- 

"Yet, we don't really understand it (web) or know about it scientifically. 
We do not know its economics. It's still hard to guess which things will 
work on which scale and which won't. There are underlined principles of 
confrontation and social concepts that we need to understand better to make 
it grow." 
http://www.thehindubusinessline.com/features/eworld/article2883222.ece?ref=wl_features

With my many shortcomings, I do have one of the better track records in 
forecasting future successful technologies that scale earlier than my peers, 
especially since the commercialization of the web, although isn't apparent 
by measuring assets--the education may have some value here. Most of this 
knowledge does not reside within institutions -- at least for quite some 
time, which leads to let's call it poor data quality in that assumptions are 
quite often wrong and then scaled widely. The incentives to share are 
similar to whistleblowers prior to reform and reward.

The important summary I'd like to share which does not see enough discussion 
in public, is that those economic issues involved are more complex than the 
technology or they would likely have been resolved already. In semantics 
actually the technology is far simpler than the economics IMO. For example 
we identified in my small incubator and lab way back in 1998 that data 
standards for provenance were necessary to create the functionality required 
for most of the economically sustainable products and services, yet those 
standards are just now maturing. First we must achieve the min level of 
complexity to incentivize and sustain economics before the difficult task of 
simplification in commercialization can do its essential task. Exceptionally 
challenging given the conflicting business model of the web and the goal of 
most of the semantic web community.

Due to similar patience and persistant efforts on display here, the issue of 
alignment of interests in IT, and more recently in neural network 
economics-- alignment is slowly but surely becoming better understood. 
Suffice to say that it's no accident that linked data and associated tools 
are "too hard and time consuming" (such a general statement doesn't deal 
with the economic conflicts of course -- LD is perhaps not difficult for 
those compensated well for the heavy lifting, but rather almost everyone 
else who must pay for the pleasure). I can speak directly to part of this in 
a considerable recent effort in tool building -- insufficient economic 
incentives have existed to compensate the relatively few people who have 
demonstrated the ability to build such tools, and they are otherwise quite 
busy and in demand of course-- the talent wars are real and strongly favor 
those with some conflict. On many occasions I and others would have liked to 
have help solve this problem but we have faced massive disincentives to do 
so, especially on the consumer web (realize some don't like seperating 
consumer versus enterprise in speaking about the web, but one must if 
speaking with scientific credibility on economic incentives and modeling, 
which influences adoption).

What many advocates don't understand, apparently, is that when we insist on 
free and open data for everything our actions directly conflict with our 
passion for adoption of a more intelligent web. We live in a world of finite 
resources-- indeed shrinking in much of the world, and all incumbents have 
some economic conflict and misalignment with any innovation-- including 
government, academia, and business. Of course that's why most technical 
progress is considered disruptive -- any innovation of importance in a 
mature society threatens important, entrenched, and powerful entities. I 
can't overstate how critical this is in private conversations, some 
protected by NDA.  Indeed I am sometimes surprised by the progress given the 
perception of the threat as it has been communicated to me--speaks in part 
to soft power and diplomacy-- perhaps threat of regulation of some kind even 
if not direct.

It might surprise some to hear from one of the sources that quite a bit of 
the business community in SV did not want an advertising model during the 
commercialization of the web -- simply because it was fully understood by 
some that it would be limiting to what it could support in terms of economic 
activity and indeed functionality.  Michael Dell was recently quoted for 
example that the IT industry is a $3 trillion industry (annual revs), but 
even though most of the focus and hype is on the consumer market the 
consumer market is only about 1/10th the total. I haven't seen the same 
research so can't confirm, but for many years I have warned about the 
limitations of free and the macro negative impact it will have on jobs, 
economics, and perhaps more direct to this topic -- data quality. Some of 
the negative economic impact is degrading the ability of sponsors to fund 
solutions. The consequences of free data that represents increasing amounts 
of knowledge also represents an enormous number of jobs and a significant 
portion of especially service economies like the U.S. -- some have guilds, 
others do not, and in some cases ultimately it may not matter if the sponsor 
is illiquid, but the profound economic impact and therefore limitations are 
clearly not understood. I believe that it's the responsibility of any 
advocate to fully understand the impact of their actions -- so do many NGOs 
that have evolved their thinking on sustainable economics relative to their 
mission, becoming leading experts on dissincentives and rebalancing 
disequalibrium -- worth consideration -- we are only recently seeing signs 
of similar maturity in computing.

I submit that it's not necessary to compromise much if at all on data 
standards if some informed comprimise is made on economic modeling and 
behavior, but we must first understand the impact of our own behavior and 
ideology--and then negotiate from a position of enlightenment -- that's 
where the semantic web community quite often appears self-destructive from 
close observation (less so in these archives than elsewhere). One of the 
reasons I don't engage more in groups and conferences is to maintain some 
perspective-- another is frankly at times it has been too painful to 
observe.

So in our case we had little choice but to focus on the enterprise where a 
sustainable model exists, but even with fairly powerful economic incentives 
inside many organizations, adoption has been longer and more difficult 
journey of evolution than previous generations of technology. For what it's 
worth I think we are seeing a bit of a reversal of the consumerization trend 
for semantics that is more similar to three decades ago. That is to say that 
we may see more advanced tools developed in the enterprise market that may 
help overcome ease of use and modeling issues on the unrestricted web.

Not intending or even inviting a debate, but rather contributing part of 
what has been very expensive education and considerable sacrifice by those 
around me, although welcome constructive private discussion.

To those old friends and colleagues who have continued all these years to 
work towards a more functional global economy through computing standards--  
and my old friends in KMGov (esp. volunteers)-- thank you and continued best 
wishes. -- MM

Received on Saturday, 31 March 2012 20:57:06 UTC