[p2p-research] Building Alliances
stanleyrhodes at gmail.com
Sat Nov 7 01:23:28 CET 2009
J. Andrew Rogers,
I agree with most of what you say, and think Michel is being a bit too free
in making assumptions about your position. I do have a few points of
contention with your reply. Forgive the length, I will try to be concise.
First, the economics of R&D from basic to applied, along with gains in
productivity and earnings, do not seem to be well understood or agreed-upon
at all. I recall a recent Science magazine podcast interview with Julia
Lane, Program Director of the Science of Science and Innovation Policy
(SciSIP) of the NSF. The June 5th show. Here is an an excerpt from the
Robert Frederick: Is there any consensus on how to calculate the return on
investment when the investment is in scientific research?
Julia Lane: No.
Robert Frederick: Does that go for basic, translational, and applied
Julia Lane: Yes. In other words, there is no consensus. I think there’s an
evolving set of approaches, but it’s still very much at the early stages. I
imagine it’s kind of at the stage – if you were a life scientist we’re kind
of at the Charles Darwin stage.
As the program director, I suspect she has a pretty good view of the big
picture, so I'm inclined to believe her basic conclusion of uncertainty
about it all. In attempting to review the literature I came to the same
Discussing "value," as in research that creates "relatively little value,"
you are mostly referring to "monetized value." How else does one measure
the aggregate output results? There's no easier measuring stick, but
obviously the inability to measure the positive externalities causes huge
omissions when assessing "value," versus directly "monetized value" via
business. You later say that some government-funded research is "valuable"
because it wouldn't otherwise be done. I agree, but you can see the problem
with different uses of the term "value."
The problem with "value" highlights a problem in your use of phrases like
"highly productive." Overall, the productivity is unclear. Obviously, a
company finding a solution--or likely solution--from the research pool is
likely quite productive for them specifically. That provides no insight
into the pool, however, unless we assume companies can somehow "drive" R&D
progress toward solutions to some extent. I am skeptical of that view (not
saying it's yours), and don't know of any solid evidence backs it up. I
suppose it's a sort of seed-versus-nuture argument. A sprout may appear,
and grow vigorously with water and sun, but that doesn't mean a lot of water
and sun on a spot with no seed will make a sprout appear.
The second major point of contention is related to the above. You mention
the lack of innovation in Web 2.0, and that it's mostly a phenomenon of
cheap bandwith. However, I don't see how using cheap bandwidth in new ways
isn't innovation. Perhaps what "innovation" means needs to be clarified,
but I only want to submit that you are being too exclusionary in your
analysis. If new networks of communication sprout from general creativity
and cheap bandwith, and are used by thousands, or even millions, of people,
we can reasonably assume the socioeconomic impact is tremendous. We may not
know what that impact is, or how to measure it, or even if it's all
"positive," but it's there.
To clarify with some adhoc terms, technological innovation may be the
solving of major technical problems, but what about sociotechnological
innovation that enables users to communicate more, or more efficiently, or
lowers the barriers to entry for new users? These effects cannot be left
out, even if they are hard to measure. Is making a technology more
"adoptable" applied R&D? Is it just "development?" Semantics, but
liberally snipped original post below...
On Fri, Nov 6, 2009 at 2:44 PM, J. Andrew Rogers <reality.miner at gmail.com>wrote:
> R&D is hopefully about maximizing useful output for the money spent in
> aggregate, not R&D theater. Productivity of R&D varies widely by
> country and source. China spends less on R&D than Japan, but an
> argument can be made that it is rapidly becoming more productive.
> You are misunderstanding that metric. R&D productivity has nothing to
> do with the riskiness or failure rate, but the aggregate output of
> useful results for the amount of money spent. In that regard,
> privately funded R&D is extremely productive, primarily because it
> does not have the "grant mill" phenomenon that you see with government
> funding that spends a lot of money on research activity but produces
> relatively little value in many cases. However, the government R&D
> funding is valuable insofar as it funds some topics that are unlikely
> to see much private research funding. That said, a rapidly growing
> phenomenon in the US is private research foundations which take on
> "basic science" type research but which are also more results
> The truth is more complicated than this. Government R&D funding is
> almost entirely captured by large corporate interests, so they have in
> effect "out-sourced" their R&D budgets to the taxpayer but still do
> the R&D work internally. Since it is not their money per se, they can
> afford to be more wasteful with it.
> That said, the vast majority of applied R&D is still privately funded
> and highly productive. It has to be, since most of the small
> innovative companies do not have the ability to obtain government R&D
> funding. Small companies *are* where most of the innovation occurs,
> but those companies are forced to use private sources of R&D funding
> so it is difficult to make the argument that government R&D is paying
> for most real innovation in many market sectors unless only big
> corporations innovate.
> I would agree with the assertion that there are certain classes of
> objectively sound venture investments that almost never get made in
> the VC market. These are almost universally areas that are extremely
> technical such that no one in the finance community can understand
> them, but there is little R&D funding for extremely technical topics
> generally outside of academic environments or very large companies.
> While it is true that Web 2.0 is not capital intensive, very little of
> Web 2.0 is "innovative" by the usual definition of the word.
> If any of the Web 2.0 companies had solved one of the major technical
> problems underlying many use cases for Web 2.0 infrastructure, they
> *would* have been innovative but they also would have been a lot more
> capital intensive.
> Indeed, Web 2.0 is very limited because of this lack of innovation.
> Twitter has not solved the constraint-indexing problem required for
> useful, scalable real-time search; FaceBook has not solved the
> self-join problem of graph databases; Amazon has not solved the
> relational analytics problem of distributed key-value stores; Google
> has not solved the problem of indexing spatial data types; etc. These
> would be landscape altering innovations because you could build Web
> 2.0 applications that are impossible today. Web 2.0 is marked by a
> complete lack of significant technological innovation, being
> essentially slick versions of very old ideas.
> Putting a lot of very old social behaviors on the web may be
> convenient and even alter the dynamics of society, but that it did not
> require any recent innovation to occur, just cheap bandwidth.
> J. Andrew Rogers
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the p2presearch