NOVA scienceNOW | Wisdom of the Crowds | PBS

Fun …

Towards the Memetics of Design

Excerpt from page 13 ..

Darwinian Design:The Memetic Evolution of Design Ideas
John Z. Langrish

“Towards the Memetics of Design …

Dawkins’s memes which, in this context, are design ideas that can be
replicated do not have to wait very long for replication to take place.
They speed up the old genetic form of Darwinian change, but the
evolution of design ideas is still Darwinian because ideas about what
to strive for are in competition for scarce resources to turn them into
manufactured realities. There are no basic principles telling us how
one group of designed objects is superseded by another. The process
essentially is unpredictable. There is no law of selection “to propel
things in the direction of progress.” Selection is blind because there is
no way of knowing what happens next. Nonetheless, we keep trying.
If we stop striving for improvement, we have stopped being human,
but we should not be surprised if our efforts sometimes fail. Once
this apparently gloomy view is absorbed, it can be put to work.”

Looking for Dr. John Langrish.

Incentive to contribute unevenly distributed information

cdling: what’s the next big thing

I have been continuing to have conversations around this idea of using a prediction market to assist in seed stage investment decisions.

What's the next big thing?

I just read this post by Fred Wilson, one of New York’s hottest VCs.  It looks like they are exploring the predictive power of crowd sourcing and platforms for harvesting this insight.

USV is in the midst of raising a new fund.  It’s their third.  They have been pointed to as a model for new sized venture capital, raising relatively smaller funds.  Investing smaller amounts, etc.   Now there is speculation that they will raise a much bigger fund and if this breaks their more intimate, boutique model.  Whether they do a bigger fund or not is another issue.  In any event, Fred and his partners are bumping into the same problem of scale described below.  I.e. how to keep your finger on the pulse of more than 25 companies?

UPDATE, Jan. 29, 2011: Yuri Milner and Ron Conway’s SV Angel give us more evidence of the trend towards investors doing more, smaller deals with their commitment to lend all 40 2011 Y Combinator companies $150,000 in convertible debt. With no cap and no discount.   Like investors everywhere, Y Combinator and SV Angel need to make choices about which companies to fund much earlier and then they need better ways to track bigger portfolios full of higher risk companies.  In this case, 40 this year and if they stay the course, 80 next year and so on …

Ron & SV Angel have already invested in more than 228 companies since 2005!

UPDATE, Feb. 4, 2011: Jason Calacanis has a good round up of reaction to Milner & Conway’s announcement in “Here’s What Insiders Have to Say …”

At the end of the day, the execution may not conform exactly to a prediction market model but I am still advocating that we explore employing collective intelligence to make better bets on seed stage companies and to monitor their progress.

Crowdsourcing.  Smaller start up costs.  Trends toward integration of social analytics and predictive models.

Why can’t we lead on these fronts from Canada?

Prediction markets have been applied successfully in many ways. Most famously in horse racing (though these are strictly speaking not prediction markets) but more recently to everything from politics to Hollywood box offices. And of course there are many corporations using them internally like Best Buy, HP, Motorola, most recently announced Ford and others.

There are a few problems and opportunities in the venture capital market that make me believe that it could use a shot of innovation, particularly locally.

1. VCs complain about a lack of quality deal flow, i.e. companies that have successfully past the seed stage of development and are ready for a $5-15 million venture round.

2. Start ups are essential to the success of the local venture capitalist but they can not find seed investors.

3. Tech start up and investing sometimes suffers because it is subject to the influence of small, tightly bound networks of people and this can make it difficult to identify, embrace and take risks on game changing innovations.

4. Recent changes in tax law make it a lot easier for US VCs to invest in Canadian companies.  This creates a couple of problems for these US VCs. We represent 10% of the North America market.  The opening up of opportunities here calls for a re-balancing of large investment portfolios to increase exposure to Canada. How then do US VCs develop, screen and manage deals?  Set up an office in Canada or some other resource intensive method like a partnership? Perhaps collaborating in a seed fund that employs collective intelligence offers an attractive alternative approach (cheaper, faster, better results?).

5. In general, it costs a lot less money to start up a potentially game changing web or mobile business than a few years ago. Micro-VCs have emerged, with some success, to take advantage of this.  They have small funds ($10-20 million) and make high touch investments in more than a dozen companies. How do you scale this?  How does a $50 or $100-million fund make investment decisions and manage more than 25 companies?  Actually, big funds like Andreessen Horowitz are also running into these issues and exploring new methods to cope.

6. While we often think that these problems are unique to Toronto, many places have the same dynamics due to the structure of innovation and the venture capital investment model.  If a venture investing model could be developed that only led to 2 out of 10 deals being very successful rather than the current anticipated rate of 1 in 10, it would change everything.  Prediction markets have proven to be accurate when used properly.  For example, Google has been using them internally for more than five years.

So I am putting my money where my mouth is.  I have incorporated Cdling Capital Services Inc. and we are:

  • Working to determine exactly how a prediction market or alternative collective intelligence model can be applied to assist decision making and monitoring of seed stage companies.
  • Working with Redesign, Inc., Peter Jones’ firm to develop the concept and UI.  Peter is Faculty at OCADU’s Master’s in Design, Strategic Futures and Innovation program and he has been able to help me introduce some great minds to this effort.
  • Working with Dr. Wendy Cukier, Associate Dean (Academic) of the Ted Rogers School of Management, Ryerson University (updated May 15, 2011 – Dr. Cukier has been appointed VP, Research and Innovation for all of Ryerson University, congratulations!) and co-investigator Dr. Charles Davis, Edward S. Rogers Sr. Research Chair in Media Management & Entrepreneurship, also with TRSM@Ryerson. Together we have scoped a study to map the Ontario Cross-Border Technology Innovation Ecosystem (OCTIE).  Details on that are available at the OCTIE Study Blog.
  • Working with DFAIT. Cdling has been invited into the Canadian Accelerator Program, established by the Canadian Trade Service and the Canadian Canadian Consulate General of San Francisco and Palo Alto.
  • Working with some great Advisors, like Olav Sorenson who teaches venture capital at the Yale School of Management and has studied extensively the relationship between networks, distance investing and venture success.

I am convinced that there is better way.  I have arrived at this point of view through a series of related experiences.  For example, designing the market (i.e. the valuable IP, in my opinion) would require many of the considerations that we routinely employed while I was VP of where we delivered a platform for providing crowd sourced marketing research through a six figure subscription model for companies like Best Buy, Gateway and Johnson & Johnson.

Though not necessary for the model to work, if we could establish a real money prediction market for these purposes in Ontario, it could be the analytical equivalent of permitting stem cell research while our neighbors choose polarization of opinion versus methods to embrace complexity.