volves formidable challenges, such
as how to scale up algorithms like
Eureqa to handle more complex models. While Langley considers Lipson’s
work to be technically solid, it deals
with only a seven-dimensional model.
What happens if the computer has
to deal with hundreds of equations,
all interacting with one another? The
computer’s runtime could grow exponentially with the complexity of the
model. Computer scientists need to
develop methods to deal with much
more complex models, Langley says.
A Revolution in science?
It is not difficult to envision how
computers could speed up scientific
discovery by plowing through vast
arrays of data and “spitting out hypotheses” for testing, suggests Bruce
Buchanan, professor of computer
science, philosophy, and medicine
at the University of Pittsburgh. In
the 1960s, Buchanan was involved in
Dendral, an early scientific discovery
computer program that used mass
spectrometry to identify unknown organic molecules. Simply sorting out
the interesting molecules from vast
arrays of information, the way Google
identifies the most popular results
of a search query, would be useful.
“There are so many cases where discoveries could have been made sooner if people were looking at different
data, asking better questions,” Buchanan says.
The larger question is whether computers could ever go beyond what phi-
Robots will be
able to conduct
a wet biology lab
in about 20 years,
says Ross King.
losopher Thomas Kuhn called “normal
science” and achieve a real scientific
revolution. Elihu Abrahams, a condensed matter physicist at Rutgers
University, and Princeton University
physicist Philip Anderson, expressed
their doubts in a letter to Science after
its publication of separate computational discovery papers by Lipson and
King in 2009. “Even if machines did
contribute to normal science,” they
wrote, “we see no mechanism by which
they could create a Kuhnian revolution
and thereby establish new physical
law.” Abrahams says he has seen nothing since then to change his mind.
King admits the field has not yet
reached that point. Adam “knows far
more facts than any human, but it
doesn’t have a deep understanding of
biology,” he says. “I think you need a
deep understanding to actually manip-
ulate concepts in novel ways.”
Computational discovery can cer-
tainly make science more efficient and
cost effective, and free up scientists to
do more deep thinking, he argues. But
for machines, King says, “the things
which the great scientists do are still a
long way off.”
Bridewell, W. and Langley, P.
Two Kinds of Knowledge in Scientific
Discovery, Topics in Computer Science 2, 1,
Dzeroski, S. and Todorovski, L. (Eds.)
Computational Discovery of Communicable
Scientific Knowledge. Springer, Berlin,
King, R., Liakata, M., Lu, C.,
Oliver, S., and Soldatova, L.
On the formulation and reuse of scientific
research, Journal of the Royal Society
Interface 8, 57, April 2011.
The computational support of scientific
discovery, International Journal of Human-Computer Studies 53, 3, Sept. 2000.
Schmidt, M., Vallabhajosyula, R., Jenkins, J.,
Hood, J., Soni, A., Wikswo, J., and Lipson, H.
Automated refinement and inference of
analytical models for metabolic networks,
Physical Biology 8, 5, Oct. 2011.
Waltz, D. and Buchanan, B.G.
Automating science, Science 324, 5923,
April 3, 2009.
neil Savage is a science and technology writer based in
© 2012 acM 0001-0782/12/05 $10.00
Instead of causing massive
computing will generate 13. 8
million jobs worldwide by the
end of 2015, according to a new
study from IDC.
Given its potential as an
outsourcing catalyst, the
cloud has often been linked
to future staffing cuts. But
the study, which is sponsored
by Microsoft, contends
that the cloud will fuel
growth initiatives by greatly
reducing “legacy drag”—the
maintenance of legacy systems
and routine upgrades—which
now accounts for 75% of
“Most [CIos] look at
migration to cloud computing
as a way to free up existing
resources to work on more
innovative projects, not to
do away with the positions
entirely,” according to the
study, which was written by the
IDC research team of John f.
Gantz, Stephen Minton, and
IDC came up with its
projections by analyzing
cloud-spending patterns in
more than 40 nations and
combining this with data
such as available country
work force, unemployment
rates, and other IT spending
Additional study results:
˲ Cloud-related business
revenue could increase to
$1.1 trillion by 2015.
˲ Small and medium-sized
companies will establish the
most jobs, with 7. 5 million new
jobs within this time. Why? Because they are less impacted by
legacy drag and are more likely
to invest in public cloud services due to capital constraints.
˲ China and India will generate 6. 75 million jobs by the end
of 2015. In contrast, less than
1. 2 million positions will be
created in North America.
˲ Public-sector opportunities
will grow, but at a much
slower pace with just more
than 500,000 jobs. Global
governments, the authors
write, will lag behind because
of their “slow pace [in] adopting the new computing style.”
— Dennis McCafferty