Subject: Re: idea futures markets
From: "Ben Tilly" <btilly@gmail.com>
Date: Thu, 7 Sep 2006 16:30:19 -0700

On 9/7/06, Kragen Javier Sitaker <kragen@pobox.com> wrote:
[...]
> What about the dominant-assurance-contract idea?  I've always been a
> little worried by the implicit disclaimer in the paper that it's only
> designed for public goods with increasing returns.
>
> (In case you're not familiar with the idea, one way to explain it is
> as follows: you make a public offer of $10 for an option, redeemable
> in 60 days, to sell some not-yet-written piece of free software to the
> option seller for $100.  If you succeed in buying more than 30 of
> these options at the end of 30 days, you write the software in the
> next 30 days and collect your $3000 or $4500 or whatever.  Anybody who
> thinks that nobody else will sell you the option has an incentive to
> sell it to you, since it won't be redeemed and they get a free $10;
> anybody who thinks that hundreds of people will sign on has an
> incentive to not sell it to you, since then they get the software for
> free; and anybody who thinks that exactly 29 other people will sign on
> will sell you the option iff the software is worth $90 to them.
> Derate as appropriate for uncertainties and time-value discount rate;
> all numbers above are arbitrary.)

I think I like it, but I'm not sure.  The main caveat that I see is
what happens if 20 people sign on and the developer decides, "I'd
prefer to get $2000 rather than pay $200, I'll write this anyways"?
What happens if people decide that this is likely to happen?

[...]
> Ben Tilly wrote:
[...]
> > ...
> > But if Transmeta funds themselves like you suggested, they get no
> > money up front but lots of money after they succeed.  Which cash
> > profile does not lend itself to succeeding.
>
> The real-world Transmeta was funded by patents and mask works, or
> rather by VCs chasing the promise of future revenues from same.  This
> is the same cash profile, isn't it?

Nope.

The real-world Transmeta was funded by selling part ownership in the
company for money now, with the promise that if they made it, they'd
give lots of money later.  This is the opposite cash profile.

The *VCs* had the cash profile that you describe.  But the founders
had the opposite one.  And it is critical that they had that cash
profile, else the founders would never have had the money to turn
their idea into a product.

An incidental note.  It is a common misunderstanding that Transmeta
was about selling low-power CPUs.  That is the business they found
themselves in, but it wasn't the business they wanted to be in.  The
business they wanted to be in was selling high-performance chips.  One
of the big barriers to that is the heat generated, so they selected up
front designs that reduced heat production to a minimum.  When they
got working chips, the speed wasn't what they hoped.  But they had low
power CPUS which they sold to make money.  Their hope was that their
simpler chip design would let them iterate development cycles faster
than the big boys, resulting in them beating Moore's law and
eventually getting to higher performance.  Unfortunately they had
simpler chips, but Intel has more engineers, and iterated designs
faster than Transmeta could.

[...]

Cheers,
Ben