Beruflich Dokumente
Kultur Dokumente
Innovation,
&
Solutions
Vinod Khosla
Khosla Ventures
1
May 2010
agenda
policy approach
2
“All progress depends on
the unreasonable man”
- George Bernard Shaw
3
experts reality check?
4
“when the train of history hits a
curve, the intellectuals fall off.”
- Karl Marx
5
“Heavier-than-air
flying machines are
impossible”
- Lord Kelvin, President, Royal Society, 1895
6
“There is no reason for any
individuals to have a computer in
their home”
- Ken Olsen, President, Chairman and Founder of DEC, 1977
7
Extrapolating the past…
“TCP / IP Technology will never replace ATM”
- various telecommunication co’s (1995) when Juniper was
started
The reality…
Today, the world runs on TCP / IP!
8
forecasting
9
oil price forecasts (1985-2005)
$51
5 year
forecast
error
Forecast
10 year
forecast
error
Actual
$15
1985 1990 1995 2000
10
Data/Source: World Oil Prices Barrel) (current $ / - EIA Office of Integration Analysis and Forecasting
EIA Oil Price Forecasts: 2005 through 2010
2009 forecast
2010 forecast
2006 forecast
2005 2030
11
Source: EIA, data for Imported Low-Sulfur Light Crude Oil (2006-2010), 2005 forecast is “world oil price”
telecommunications: actual vs. forecast demand
Actual
1,000 GB/s
1995 projection
1987 projection
1983 projection
1980 projection
1978 projection
1 GB/s
12
Source: internal forecasts for major telecommunications company
Mckinsey : US mobile subscribers
1980 forecast for 2000
forecast actual
13
Source: American Heritage Magazine - http://www.americanheritage.com/articles/magazine/it/2007/3/2007_3_8.shtml
yesterday’s technology, tomorrow’s
forecast
14
uantitative modeling flaws
15
the folly of predictions:
tetlock study
undreds of experts.
0,000+ “expert” forecasts & 20+ years
16
Source: http://www.newyorker.com/archive/2005/12/05/051205crbo_books1
why?
17
Source: http://www.newyorker.com/archive/2005/12/05/051205crbo_books1
How many predicted…
“How many in 1945 would have predicted
Japan and Germany would be in 35 years be
the 2nd and 3rd most powerful economies?”
-
Tetlock
18
… experts were asked to
predict about
• Soviet Union (in 1988) - by 1992, Communist Party control will
• strengthen
• weaken
• remain unchanged
19
… with unexpected results
20
so how do we
pick?
21
“extrapolation of the
past”
vs.
23
“black swan”
…rarity, extreme
impact, and
retrospective
(though not
prospective)
predictability
24
Source: Nassim Nicholas Taleb, author of “The Black Swan”
“what if…”
PRODUCTION
CONSUMPTION
}
• “lighting was 80% more efficient?”
25
kior
26
%
Millions Crud
of e
years oil
Re-
Crud
KiOR e
secon
ds
28
OPOC engine technology…new power system
29
Attitude: everyone told me…
1.It is not possible to convert a gasoline engine to a Diesel, using the same transfer
line. I did it and it is the most successful Diesel in the world and it was copied by
everybody.
2.The combustion for a high speed Diesel is not possible. I started production with a
5000 rpm Diesel with 2000 engines/day on the gasoline engine transfer line.
3. You cannot use a rubber toothed belt to drive the camshaft and the injection
pump. I did it and it is the standard solution today.
4.It is not possible to use an aluminum radiator because the corrosion will destroy the
engine. I did it and it is the standard solution today.
5.It is not possible to create an “emission free” natural gas burner. I did it. It is in mass
production at VIESSMANN. BUDERUS sued VIESSMANN about “emission free” and lost.
31
…soraa
33
…reinventing the HVAC
35
artificial leaves to produce energ
36
Source: http://blog.greenenergytv.com/blog/define-solar-energy/0/0/the-green-invention-maverick-of-solar-energy-dr-nate-lewis
silicon nanotube batteries?
37
Source: http://www.research.philips.com/newscenter/pictures/intsol.html
LENR ?
38
Source: http://www.cafeterra.info/2009/03/cold-fusion.html
…the power of ideas &
entrepreneurship
39
…the sources of innovation
Amazon : Walmart
Cree: GE
DNA Sequencing
40
“relevant scale” solutions for
... oil
... coal
... materials
... (efficiency of oil & coal use)
41
“...relevant cost”
“...relevant scale”
“...relevant adoption”
42
…the “cost” chindia test
only scalable if competitive unsubsidized
43
…relevant scale “resources”
44
resource multipliers
nanostructured materials
non-chemistry batteries
synthetic biology
……
45
one billion acres…
46
e: Chris Somerville, citing Campbell et al., Env. Sci. Technol. (2008) 42,5791
…Brazilian Cerrado – evolution of a
bread basket
48
Source: Ceres Inc
the perennial advantage
Perennial crops:
•Diversity protects
against diseases
49
Source: Wes Jackson, Land Institute
the polyculture advantage
50
51
another billion acres?
deserts as solar farms
3000 km
52
Source: Gerhard Knies, CSP 2008 Barcelona
yet another billion
acres...
Geothermal?
53
negawatt energy
savings!
54
negabarrel energy
efficiencies!
55
path to black swans...
56
…technology expands the ‘art
of the possible.’
…today’s ‘unimaginable’ is
tomorrow’s ‘conventional wisdom.’
57
…the “relevant adoption” test
adoption risk/trajectory matters
58
key criteria
• Scalability Trajectory
• Adoption Risk
• Capital Formation
• Optionality
59
reality /
perspective
60
goals, not solutions
electric cars?
62
62
reality: $2,500 nano
64
… the right & wrong battles: framing
environmentalists or pragmentalists? 65
…carbon reduction capacity
66
May want to
move
carbon reduction capacity: 10X increase in carbon
productivity!
10
8
Carbon Productivity = GDP /
7
Carbon Productivity Growth Required =
6
5.6%/yr
(2008= 1)
=1)
greater capacity to
Index
WorldEmissions
3
respond in the future?
2
67
Source: “The Carbon Productivity Challenge”, McKinsey – Original GDP projection from Global Insight through 2037
… the takeaway
don’t believe experts, econometrics, big companies
68
Takeaway: entrepreneurial
engine is powerful!
69
exponential innovation…
assuming ~2% rate of change…
= 1.8X
units
takeaway: to predict 2040 in
2010, we would need to predict
2010 in 1926!
= 1.8X
units
192 201 204
0 70
6 0
as surely as...
71
to predict the future
invent it!
72
... or get to
work.
vk@khoslaventures.com
khoslaventures.com/resources.html
73