By Amir Mizroch
When the University of Washington's computer-science department
wanted to poach artificial-intelligence expert Carlos Guestrin from
Carnegie Mellon, it turned to Amazon.com Inc.
The Seattle-based tech giant ponied up $2 million to fund two
professorships: one for Mr. Guestrin, and another for his wife, who
also works in the field. To seal the deal, Amazon Chief Executive
Jeff Bezos met the academic during a campus visit.
"[Mr. Bezo] is a very smart guy. He has a crazy laugh," said Mr.
Guestrin, now UW's Amazon Professor of Machine Learning. "We got
quickly into technical things: What was I working on in large-scale
machine learning? How could I impact Amazon? What could this mean
for the business of data?"
Google Inc., Facebook Inc., Amazon and other technology
companies are scrambling to push the bounds of artificial
intelligence, or AI, and in that effort they are stocking their own
research centers with big-name academics and aspiring Ph.D.
candidates.
Tech companies also are pouring funds into universities with
expertise in the once-obscure field. University of Washington,
based in the same state as Microsoft Corp. and Amazon, has long
been a center of excellence for computer science, including
artificial intelligence. Microsoft, Intel Corp. and Google, as well
as Amazon, all fund some of UW's AI research.
UW also has become a Silicon Valley hunting ground. Before it
recruited Mr. Guestrin--who earned his reputation creating
artificial-intelligence-related tools for developers--the
university lost seven AI-related professors to Google.
"There's a massive battle under way for talent," said Oren
Etzioni, on leave from UW's computer-science faculty and now
heading up the Seattle-based Allen Institute for Artificial
Intelligence, a nonprofit set up by Microsoft co-founder Paul
Allen. "Virtually every professor at the UW computer-science
department has been called many times to work at these companies,
and frankly it's a very compelling pitch."
Companies are on the prowl not just for big names in the field,
but for newly minted Ph.D.s. Amazon is advertising for more than 50
AI positions in the U.S. and Europe, hunting for doctorate-holders
in fields like machine learning, information science and
statistics.
Last year, Google bought DeepMind, a startup founded by
Cambridge University graduates. After the Google deal, DeepMind
absorbed two Oxford University spin-offs specializing in AI. As
part of the transaction, Google agreed to a research partnership
with Oxford's computer-science program.
Google and Amazon declined to comment about their AI
ambitions.
AI is a broad academic field, encompassing techniques aimed at
giving computers the ability to make decisions that a human might,
based on data analysis. Machine learning and other subsets are a
more-targeted discipline inside the broader AI field.
Commercial uses for AI are still limited. Predictive text and
Siri, the iPhone's voice-recognition feature, are early
manifestations. But AI's potential has exploded as the cost of
computing power drops and as the ability to collect and process
data soars. Big tech companies like Facebook and Google now vacuum
up the huge amount of data that needs to be processed to help
machines make "intelligent" decisions.
"AI has become 'like wow,' in Silicon Valley today," said Akli
Adjaoute, founder and CEO of Brighterion, a software company that
uses machine learning techniques to spot financial fraud for credit
card customers.
Microsoft is working on understanding context in human
interaction. The company has been awarded a patent for
Internet-connected glasses that can detect and interpret the
emotions of people within their field of vision in real time and
provide feedback to the wearer. The patent for "a wearable emotion
detection feedback system," was filed in October 2012, and awarded
this Tuesday.
Asked about Google's top priorities at a conference last week
Executive Chairman Eric Schmidt said the "core thing" his company
is working on these days is machine learning. He cited progress in
image and speech recognition. Regarding the latter, he said it is a
"sore point" that Apple Inc.'s Siri "gets all the credit."
The relationship between tech giants and academia can be
difficult to navigate. Some faculty members complain tech companies
aren't doing enough in the many collaborative efforts now under
way. One big gripe: Companies aren't willing to share the vast data
they are able to collect.
"The high value of this work encourages companies like Google to
keep their progress more secret," said Tom Mitchell, a department
head at Carnegie Mellon's computer-science program.
Those who embrace the relationship say it can provide real-world
incentive for scientific advances. Hank Levy, head of UW's
computer-science program, said he isn't bitter about the poaching
from Google over the years.
"Often, people go off for a year or two and then they come back
and bring new experiences that expand both their teaching and
research," he said.
In late 2013, Facebook hired Yann LeCun, one of the world's most
prominent AI academics, from New York University. As an AT&T
engineer in the 1980s and '90s, he helped pioneer
handwriting-recognition processing used by banks to authenticate
checks. He is now Facebook's chief of artificial intelligence.
As part of the courtship, Facebook let him keep his post at NYU,
a block up Broadway from Facebook headquarters. He still works for
the university part time. Facebook partnered with the university on
a new center dedicated to data science, a key element of AI
research. Facebook scientists lecture at NYU, and NYU Ph.D.
students can apply for long-term internships at Facebook's AI
lab.
Facebook Chief Executive Mark Zuckerberg read some of Mr.
LeCun's papers before meeting him during the recruitment process.
"That completely floored me," says Mr. LeCun.
Rolfe Winkler in San Francisco contributed to this article.
Write to Amir Mizroch at amir.mizroch@wsj.com
Access Investor Kit for Amazon.com, Inc.
Visit
http://www.companyspotlight.com/partner?cp_code=P479&isin=US0231351067
Access Investor Kit for Facebook, Inc.
Visit
http://www.companyspotlight.com/partner?cp_code=P479&isin=US30303M1027
Access Investor Kit for Google, Inc.
Visit
http://www.companyspotlight.com/partner?cp_code=P479&isin=US38259P5089
Access Investor Kit for Google, Inc.
Visit
http://www.companyspotlight.com/partner?cp_code=P479&isin=US38259P7069
Subscribe to WSJ: http://online.wsj.com?mod=djnwires