Stanley, Kenneth O., and Joel Lehman. Why greatness cannot be planned: The myth of the objective . Springer, 2015.
This book was born from a radical idea about artificial intelligence (AI) that
unexpectedly grew to be about much more. At first I was thinking only about AI
algorithms, programs of concern mainly to computer scientists like me. Usually
these algorithms have explicit goals and objectives that they’re driven to achieve.
But I began to realize that the algorithms could do amazing things even if they
had no explicit objective—maybe even more amazing than the ones that did have
an objective. Testing this idea led to some surprising experimental results, some
of which are documented in this book. And that’s pretty interesting—if you’re a
computer scientist.
But then something unusual happened. I started to realize that the insight wasn’t
just about algorithms, but also about life. And not just life, but culture, society, how
we drive innovation, how we plan for achievement, our interpretation of biology—
the list just kept expanding. If you don’t understand why I say that’s unusual, just
consider how rare or bizarre it is for a computer algorithm to change how you think
about life. After all, you don’t fall into an existential crisis every time you boot up
your laptop. The unexpectedly broad implications of the idea surprised me, so at first
I kept them tucked quietly in the back of my head—but they kept getting louder.
As a professor, I’m sometimes invited to give public talks on my research. So
as an experiment, when I spoke publicly about the idea I began touching on its
relationship to our lives and society. And as I saw how people reacted and how much
passion it provoked, the message took on a new life—because I could see it had
meaning well beyond the field of its origin. So I realized someone had to write this
book at least to try to communicate the novel insight. And that’s why you’re in for a
unique experience in the pages ahead. There is a story here—a story about an idea
in AI and how it grew into something bigger—but there’s also a journey through a
dizzying set of surprisingly broad implications for everything from personal dating,
to the march of science, to the evolution of the human brain. I hope you’ll enjoy this
whirlwind tour across a landscape of once familiar concepts—but now observed
through a psychedelic new lens.
One more important detail about the history of this project—it was really
a two-person effort. From the earliest experiments to realizing the growing set of implications, my coauthor Joel Lehman was instrumental all along the way.
The ideas here are the result of years of exchange and debate between the two of us,
and the book is therefore truly a joint effort. So from the first chapter onward, we’ll
be speaking with a single unified voice as we guide you through Why Greatness
Cannot Be Planned: The Myth of the Objective.
Joel and I would both like to express gratitude to the institutions that supported
this work: the University of Central Florida, the University of Texas at Austin,
and the Santa Fe Institute, where I completed the book while on sabbatical. Every
past and current member of the Evolutionary Complexity Research Group at UCF
deserves special thanks for their input and ideas over the years. I am also indebted to
Gary Leavens, the Chair of the Computer Science Division at UCF, who encouraged
me to make this book a reality. This book also probably would not have happened
without the motivation provided by early invitations to speak from Richard Gabriel
of IBM Research (at the 2010 SPLASH conference) and Seung Chan Lim (Slim)
of the Rhode Island School of Design in 2011. For creating the experiment that
provided the initial inspiration for the ideas in this book, the entire Picbreeder team
and later Picbreeder contributors deserve special mention: Jimmy Secretan (lead),
Nick Beato, Adam Campbell, David D’Ambrosio, Adelein Rodriguez, and Jeremiah
T. Folsom-Kovarik; Nazar Khan, Peter Matthews, and Jan Prokaj later contributed
to researching Picbreeder’s color extension.
The book is organized into two parts for the convenience of different readers.
The first part of the book (the first nine chapters) constructs the main argument
against objectives and provides general evidence for their cost in a number of areas
of life and society. For readers interested in more elaborate implications of the
myth of the objective in specific scientific fields (in particular biology and artificial
intelligence), two additional case studies are included at the end of the book. That
way, while you can absorb the main ideas from just the first nine chapters, these
additional case studies provide further depth for those who desire more.
Kenneth Stanley
Santa Fe Institute, Santa Fe, NM
March 10, 2015