Apple and Intel scoop up artificial-intelligence experts, the
startups where they work
By Asa Fitch
This article is being republished as part of our daily
reproduction of WSJ.com articles that also appeared in the U.S.
print edition of The Wall Street Journal (December 27, 2019).
Technology giants are opening their checkbooks to keep pace in
the hot field of artificial intelligence, where some of the best
ideas and most sought-after talent reside in small startups and
academia.
Intel Corp.'s purchase of Israeli startup Habana Labs for $2
billion this month is the biggest AI deal in a year when those
transactions have already hit a record. Other tech giants,
including Amazon.com Inc. and Microsoft Corp., have made
acquisitions in recent years to strengthen their ability to apply
complex algorithms to improve the performance of everything from
delivery robots to self-driving cars. Apple Inc. has snapped up 17
AI startups since 2013, according to market-data firm
PitchBook.
"AI is strategically important to many industries, and getting
it right or wrong can mean the difference between a company
thriving or becoming irrelevant," said Andrew Ng, the founder of a
number of AI startups, including Landing.ai, in which Intel's
venture-capital arm is an investor.
Many of the deals are driven as much by buyers wanting to land
talent as to acquire a specific product, said Oren Etzioni, who
heads the Allen Institute for AI, a Seattle research outfit started
by late Microsoft co-founder Paul Allen.
"Often, half of what you're buying is a team that's jelled
together, that's showed they can do stuff," Mr. Etzioni said.
When Apple, in June, said it had bought the struggling
self-driving startup Drive.ai for an undisclosed amount, it was
widely seen as a bid to snag its engineers, a common justification
for deals across the tech world. Apple didn't respond to a request
for comment.
Securing the human brain power can also be a rationale for hefty
premium prices. Intel paid up for Habana even though the unit only
recently began selling its first products and likely has minimal
sales, said Mitch Steves, an analyst at RBC Capital Markets. Intel
and Habana haven't disclosed the startup's revenue.
Habana Chairman Avigdor Willenz will continue to serve as an
adviser to the business and to Intel, the chip maker said.
"We're super-committed to investing to win in this market," said
Navin Shenoy, who oversees Intel's data center and AI business.
Buying Habana, which specializes in chips suited for
machine-learning applications, is part of Intel's bet on rapidly
rising demand for AI computing needs, Mr. Shenoy said. The company
estimates industrywide sales of chips made for AI computing will
reach $25 billion within the next five years.
David Dahan, Habana's chief executive, praised Intel's strategy
and said the companies would deliver more AI innovation faster
together.
Intel has been bulking up on AI for some time. It paid $400
million in 2016 for Nervana, an AI chip maker based in San Diego.
Then it shelled out $15.3 billion in 2017 for Mobileye NV, an
Israeli startup specializing in automotive-camera technology with
self-driving-car applications.
Even before Intel's latest purchase, AI-related deals globally
had surged to $35 billion in value through early November, topping
the previous high of $32 billion two years ago and $11 billion in
transactions last year, according to a report from Stanford
University's Institute for Human-Centered Artificial
Intelligence.
Salaries reflect the high demand. AI engineers, most with
advanced degrees, make $224,000 a year on average, according to MMC
Ventures Ltd., a U.K.-based venture-capital fund. That is more than
double the $104,480 average salary for software developers in the
U.S., according to 2018 data, the most recent available from the
U.S. Bureau of Labor Statistics.
Companies have long sought to lure AI experts away from
academia.
Ride-hailing giant Uber Technologies Inc. hired away many of
Carnegie Mellon University's robotics professors in 2015 and built
a self-driving lab near the Pittsburgh institution to bolster its
autonomous-driving efforts. The company sought to build mutually
beneficial partnerships with academia, an Uber spokeswoman
said.
A University of Rochester assessment in August found that 221 AI
professors in North America decamped for tech companies in the past
decade and a half. They either took full-time jobs or split their
time between teaching and private industry. The pace has
accelerated, with nearly 40 professors going into private industry
last year alone, the study found.
Last year, private industry also lured away 60% of AI doctoral
graduates, compared with just 20% in 2004, according to the study
by the institute at Stanford.
Facebook Inc., which has struck several AI deals in recent
years, has set up AI labs near universities to help attract talent.
Last year, it attempted to hire a number of professors at the
University of Washington, according to university leaders and some
of those who were approached. Facebook offered higher compensation
-- potentially double or more what a professor might make when
including stock options and other perks, two of the people Facebook
reached out to said.
"Anyone in the department who could spell AI was approached,"
said Ed Lazowska, who holds the Bill and Melinda Gates chair in
computer science at the Seattle-based university. Mr. Lazowska
wasn't one of the people Facebook recruited.
Only one professor ultimately took a part-time role at
Facebook.
Jerome Pesenti, a Facebook executive who led the recruiting
drive, said the company allowed academics it hired to maintain
positions at their universities so that they could continue to
develop new talent.
"We also work closely with university administrations to make
sure we don't impede their research," he said.
Write to Asa Fitch at asa.fitch@wsj.com
(END) Dow Jones Newswires
December 27, 2019 02:47 ET (07:47 GMT)
Copyright (c) 2019 Dow Jones & Company, Inc.
Intel (NASDAQ:INTC)
Historical Stock Chart
From Aug 2024 to Sep 2024
Intel (NASDAQ:INTC)
Historical Stock Chart
From Sep 2023 to Sep 2024