[p2p-research] Building Alliances (basic income and entrepreneurship)

[J. Andrew Rogers]

On Fri, Nov 6, 2009 at 1:53 PM, Paul D. Fernhout
<pdfernhout at kurtz-fernhout.com> wrote:
> So, do we need to structure our *entire* global economy a certain way
> because Princetonians and others strongly control 0.06% of the money flow?


No one has suggested this that I am aware of. I merely stated that
this is highly productive capital compared to many other uses, but
there is a saturation point.

It is almost always the case that there is far more venture capital
than there are innovative startups to invest it in. The real problem
is that innovators (not people that think they are innovators, but the
tiny minority that produce real results) are in short supply. Also it
requires discipline and motivation as well to accomplish anything
useful, which is also frequently in short supply.


> There's truth to that, but a basic income bypasses most of that. Most
> inventors are just struggling to pay living expenses. And if they don't have
> to have a standard job to do that, they can live where rents are cheap and
> land wide open and unzoned.
>
> Very few big ideas need much funding these days beyond living expenses for
> the developers. Some do, but not most.


This really depends. Adequate hardware often is not cheap. For many
areas you need to hire domain specialists which may be very expensive
and hard to come by. Sure, you *could* do it as you describe above,
but it would be extremely inefficient for many types of innovation.


> Can you name any such "non trival" innovations?


Software is often only inexpensive if you are doing something many
people have done before (and therefore "trivial"). Think FaceBook and
Twitter.

Many theoretical computer science innovations that are developed these
days are exactly this type of problem, even though on the surface it
seems like it is software and should be cheap. This is part of the
reason it takes so many years for it to percolate into useful
software. Everything has to scale out these days (read: massively
parallel/distributable), which requires a perfectly matched dev/test
environment -- expensive. Algorithms derived from new theoretical
models mean you basically have to rewrite every piece of software that
might use said algorithm -- a major pain.  Often, you can count the
number of people who even grok the theory of the software on one hand
due to various bits of theoretical esoterica that you can't just read
up on over the weekend.  The cost of producing the first version of
something completely new and innovative is very high. Subsequent
copies, not so much.

Here is the really interesting thing: the cost of implementing new
computing *hardware* is converging on the cost of implementing new
computing *software* in terms of implementing fully functional
prototypes.


> Very few innovations require much materials. Maybe making new desktop CPU
> foundries?


No, designing and prototyping a new computing architecture is
surprisingly cheap these days, a couple million if you are frugal. No
need to make your own foundry, there are companies that specialize in
that kind of thing if you like how things work in reconfigurable
silicon first.

Manpower is the primary expense because for bleeding edge work you
often need extremely talented people to do work that those people have
no significant personal interest in doing. Hardware usually comes in a
distant second even for large-scale computing apps.

Some other industries (like pharma) have different cost structures.


> How much is work on solar panel design really costing? In general, our
> society has favored "big science" over "little science". But it is ofter
> from little science where the really new ideas come.


I know next to nothing about solar panel R&D cost structures, so I couldn't say.


> And those ideas that do require big resources can generally gain occasional
> access through places like university labs or incubators. Not *all*. But by
> far the vast majority. Anything software these days.


Generally true on the software side, we do not own any large-scale
hardware (yet?) even though we use it a lot. However, I wonder how
many other people would have the access that we do as a practical
matter. These are limited and expensive resources. And obviously you
could not build a production application on these resources.

The big resources required are extremely smart people that can
reliably produce successful results. Even when just talking about
programmers, the vast majority are pretty useless for non-vanilla
development, which is a major bottleneck when you need to write
100kLoC of complex, semi-exotic code.


> I just don't see it. Almost all those "millions" you talk about are going to
> salary, or they are paying for fancy equipment instead of hiring more people
> or having more patience. So, a loose confederation of free individuals all
> with a basic income could do the same.


You aren't quite getting it, and at least in our case, we aren't
paying for fancy equipment so it is almost all people money. Supply
and demand is not magically solved. Let's suppose that you need a
particular expert for a very important and valuable R&D program, and
there are only three plausible known candidates in the entire world
*and* they have no particular interest in working on this program, how
do you build the necessary team?  This kind of thing actually happens,
and people at the edge of the envelope where significant innovation
actually occurs are in short supply.

Also, time is costly even if we discount the financial aspect. Slowing
things down unnecessarily slows the aggregate rate of innovation, a
net loss. All innovations that are dependent on the current
innovation, say some lifesaving innovation, gets kicked into the
future indefinitely.

This planet is desperately short of people capable of usefully
contributing to non-trivial R&D efforts. A million average people with
no particular expertise cannot replace one extremely brilliant person
with deep expertise. If this was not true, "design by committee" would
not be pejorative.


-- 
J. Andrew Rogers
realityminer.blogspot.com