From: Nathan Myhrvold
Sent: Saturday, June 21, 2014 3:43 PM
To: jeffrey E.
Subject: FW: Innovation article in the new yorker
From: Nathan M=hrvold
Sent: Saturday, June 21, 2014 8:41 AM
To: Bill Gates ( ; Larry Cohe ;=Casey Tegreene; 'Lowell Wood'; Edward Jung;
• Cam =yhrvold cky Wood; Yuki Ishikawa; Peter Detkin; Gre=
Gorder; Adriane Brown; Russ Stein; David Kris; Scott Heimendinger; Chris Alliegro; Maurizio Vecchione
Subject: Innovation article in the new yorker
In 1997 Clayton Christensen came out with a book cal=ed the "The Innovator's Dilemma". It told a compel=ing story of
how new technology could be disruptive to existing markets an= competitors. The book became wildly popular within
the tech industry. Everybody wanted their new technology to be vie=ed as "disruptive", and advocates started seeing
"disrup=ive" threats everywhere. At Microsoft there was a consta=t stream of discussion about which projects were
disruptive and which were not, or which companies were going to disrupt us, and who we could disrupt= In the years
since 1997 this book and the vocabulary it intro=uced have been part of the holy writ of Silicon Valley.
At the time I thought that the book had some value, =ut was dismayed at the extremes to which it was adopted. This
sort o= business book is rarely what science would call a theory — i.e. som=thing with predictive value. Instead they tend
to provide some nouns and verbs that one could use as a language to discus= a situation or company. The difference is
crucial - an after =he fact the story isn't much use to guiding decisions. T=e stock and trade of most business theorists
is that they tell very compelling stories which then tempt people into using them like =heories — to guide decisions.
Indeed that is why people buys b=siness books, and pay speaking fees to the author. A vocabular= for story telling isn't
the same as a predictive theory.=/p>
Here is an illustration of the difference that happe=ed to me in Africa on safari. A huge elephant charged the
open=vehicle I was in. In previous cases the guide had honked the horn, o= even put the vehicle in reverse and driven
away. This time the guide was calm, and made no effort to do anything. The=elephant stopped about 10 feet from us
(way too close for the lens I had o= the camera), trumpeted loudly and stomped off. The guid= said "you can always
tell a bluff charge". I aske= how — what were the signs that let him know that this would be OK?&n=sp; He smiled and
said "they stop".
EFTA_R1_00135407
EFTA01797931
Unfortunately a lot of popular business books have t=at property — they provide a language for telling stories after the
=act. Unlike my elephant story they are not as transparent abou= the lack of predictive power; they provide elaborate
descriptions that are full of pseudo-causation but without the rigor that =ould let them really be predictive theories.
Anyway, Jill Lepore has a long article in the curren= issue of the New Yorker that takes apart Christensen's books and
ar=ues that it was essentially all made up. The examples that he uses w=re cherry picked to make his point. Worse, if
you look closely at the examples many of them could be used to prove the o=posite point. I think that it a great article, I
just wish it had be=n published in 1998 rather than 2014. Then again, Lepore is a histor=an, so I suppose their sense of
urgency isn't the same as mine.
http://www.newyorker.com/reporti=g12014/06/23/140623fa_fact_lepore?currentPage=1
chttp://www.newyorker.com/reporting/2014/0=/23/140623fa_fact_lepore?currentPage=1>
Of course taking his examples apart doesn't me=n that Christensen is utterly wrong about everything. There can be
some va=ue to a descriptive language. Over time, further work can elab=rate on the language to make it into a
predictive theory. Medical science is a good example of that. =lt was once a catalog of names for conditions and
ailments without much id=a of the causation. Today that is still true for a distressing=number of conditions, but others
have been fully figured out. Malaria was once thought to be due to fumes emanating from swamps the (mal=aria
means "bad air" in Italian). We now kno= all about the disease and can cure it completely (at least in places with=an
adequate health care system). Meanwhile the condition that Dr Alois Alzheimer first described in 1906 is in flux. We
know = lot more about it than he did, but its precise cause and treatments still=eludes us.
In addition to taking Christensen apart, I was amuse= by the part of the article that says that the word "innovation=;
once had a largely negative context.
Nathan Myhrvold=/span>
Founder, CEO
E
T
F
2
EFTA_R1_00135408
EFTA01797932
www.intellectualventures.com <http://www.intellectualven=ures.com/>
This message m=y contain confidential information which may also be legally privileg=d information. If you are not an
intended recipient of the message, please delete it and notify the sender via reply=email. Any unauthorized
dissemination, distribution or copying of th= material in this message, and any attachments to the message, is
strictly=forbidden.
3
EFTA_R1_00135409
EFTA01797933