The Washington Post - 01.08.2019

(Axel Boer) #1

A16 EZ RE THE WASHINGTON POST.THURSDAY, AUGUST 1 , 2019


Economy & Business


MEDIA


Networks sue free


streaming service


The four major U.S. television
networks on Wednesday sued a
nonprofit group backed by AT&T
that streams TV channels over
the Internet free without
permission.
The lawsuit was filed in U.S.
District Court in New York by
CBS, Comcast’s NBCUniversal
Media, Fox Television Stations
and Walt Disney’s ABC against
Sports Fans Coalition NY and its
founder, David Goodfriend, a
former executive at satellite-TV
provider Dish.
Major broadcasters
traditionally charge cable and
satellite pay television providers
such as AT&T and Dish a fee to
distribute their content. By
giving viewers that content free,
the free service, Locast, is
threatening these fees.
“Locast has been operating for
more than a year and a half, and
no broadcaster has filed until
now,” said David Hosp, an
attorney representing
Goodfriend. Hosp said the
group’s nonprofit status and the
fact it was not charging for the
service put it on “sound legal
footing.”
Broadcasters disagreed.
“Locast is not the Robin Hood of
television; instead, Locast’s
founding, funding, and
operations reveal its decidedly
commercial purposes,” according
to the lawsuit.
— Reuters


LABOR

Survey: Small-business
hiring edged up in July

Small-business employment
edged up during July as owners
gingerly took on new hires.
That report came Wednesday

from payroll provider ADP,
which counted 11,000 new jobs at
its small-business customers,
those with up to 49 employees.
ADP’s smallest customers, those
with up to 19 employees, cut
18,000 jobs.
The slim overall increase in
hiring followed two months of

job cuts — small companies cut
13,000 jobs in June and 34,000 in
May, ADP said. That followed a
gain of 66,000 in April.
Small-business hiring has
been constrained by a confluence
of factors. Owners have been
conservative about adding to
their payrolls since the recession.

They have also struggled along
with larger companies to find
qualified workers partly because
they generally cannot offer the
salaries and benefits big
corporations do.
ADP said all of its business
customers created 156,000 jobs
in July.
— Associated Press

FOOD INDUSTRY

Impossible Burger gets
okay on color additive

The Impossible Burger, so far
only available at restaurants,
could finally be making its way to
U.S. grocery store shelves. It also
announced on Wednesday plans
to produce more of the meat-free
patties through a new
collaboration.
In response to a petition
submitted by Impossible Foods,
the Silicon Valley-based maker of
the burger, the Food and Drug
Administration has amended its
rules to call the use of soy
leghemoglobin safe as a color
additive in imitation beef,
clearing a key hurdle in the
company’s push to sell raw
product inside grocery stores.
The rule change is effective
Sept. 4, though petitioners still
have a chance to file objections.
The original petition filed in
December specified heme could
“not exceed 0.8 percent by
weight” of the final product.
Separately, the company said
it had signed a deal with global
food producer OSI Group to
expand production. OSI will

begin making the Impossible
Burger starting next month,
adding short-term capacity to
Impossible’s own Oakland, Calif.,
factory.
— Bloomberg News

ALSO IN BUSINESS
The annual wages and benefits
for U.S. workers rose in the
second quarter at a slightly slower
pace than the first, suggesting the
lowest unemployment levels in a
half-century have not been
triggering rapid gains in worker
compensation. The Labor
Department said Wednesday that
pay and benefits for all U.S.
workers increased 2.7 percent in
the April-June quarter from a
year earlier, down from a
2.8 percent rise in the first quarter
compared with a year ago. The 12-
month peak so far in this
expansion for wages and salaries
was a 2.9 percent gain for the
period ending in December 2018.

Pagani gave video-game fans a
first look Wednesday at the
multimillion-dollar hypercar it
will debut during Monterey Car
Week in California next month.
The Italian brand unveiled its
Huayra Roadster BC in the
popular Zynga game CSR Racing
2, which lets players drag race a
stable of over 150 meticulously
rendered vehicles. It’s the top-
grossing mobile racing game in
the United States, with a daily
haul of $60,068 and 12,593 new
installs, according to
ThinkGaming.com.

— From news reports

DIGEST

CHAIDEER MAHYUDDIN/AGENCE FRANCE-PRESSE/GETTY IMAGES
Workers load yellowfin tuna into a truck at a fishing port in Banda Aceh, Indonesia, on Wednesday.




DOW 26,864.
DOWN 333.75, 1.2% 

NASDAQ 8,175.
DOWN 98.19, 1.2% 

S&P 500 2,980.
DOWN 32.80, 1.1% 

GOLD $1,437.
DOWN $4.00, 0.3% 

CRUDE OIL $58.
UP $0.53, 0.9% 

10-YEAR TREASURY
UP $4.50 PER $1,000, 2.01% YIELD

CURRENCIES
$1=108.80 YEN; EURO=$1.

BY TAYLOR TELFORD

In just a handful of years, the
business of emotion detection —
using artificial intelligence to
identify how people are feeling —
has moved beyond the stuff of
science fiction to a $20 billion
industry. Companies such as IBM
and Microsoft tout software that
can analyze facial expressions
and match them to certain emo-
tions, a would-be superpower
that companies could use to tell
how customers respond to a new
product or how a job candidate is
feeling during an interview. But a
far-reaching review of emotion
research finds that the science
underlying these technologies is
deeply flawed.
The problem? You can’t reli-
ably judge how someone feels
from what their face is doing.
A group of scientists brought
together by the Association for
Psychological Science spent two
years exploring this idea. After
reviewing more than 1,000 stud-
ies, the five researchers conclud-


ed that the relationship between
facial expression and emotion is
nebulous, convoluted and far
from universal.
“About 20 to 30 percent of the
time, people make the expected
facial expression,” such as smiling
when happy, said Lisa Feldman
Barrett, a professor of psychology
at Northeastern University, who
worked on the report published
earlier this month. But the rest of
the time, they don’t. “They’re not
moving their faces in random
ways. They’re moving them in
ways that are specific to the situa-
tion.”
It’s not surprising that some-
thing as complex and internal-
ized as human emotion defies
easy classification. Humans tend
to instinctively draw on other
factors, such as body language or
tone of voice, to complete their
emotional assessments. But the
majority of emotion-detection AI
makes inferences purely on map-
ping facial positioning, a concept
that stems from the work of Paul
Ekman, a psychology professor at
the University of California at San
Francisco. He posited that six
emotions — happiness, sadness,
disgust, fear, anger and surprise
— are represented by universal
facial expressions across all cul-
tures.
When Microsoft rolled out its

emotion detection technology in
2015, it said its algorithms could
“recognize eight core emotional
states — anger, contempt, fear,
disgust, happiness, neutral, sad-
ness or surprise — based on uni-
versal facial expressions that re-
flect those feelings.”
It’s a common justification for
such technology, and exactly
what Barrett and her colleagues
are pushing back against. The
companies are not trying to be
misleading, she said, but they
need to dramatically change their
approach to emotion detection to
get the kind of accuracy that
many already purport to have.
(Microsoft declined to comment
on how or if the review would
influence its approach to emotion
detection.)
“We now have the tools and the
analytic capability to learn what
we need to learn about facial
expressions in context and what
they mean,” Barrett said. “But it
requires asking different ques-
tions with that technology and
using different analytic strategies
than what are currently being
used.”
Such technological limitations
come with risks, especially as
facial recognition becomes more
widespread. In 2007, the Trans-
portation Security Administra-
tion introduced a program

(which Ekman consulted on but
later distanced himself from) that
trained officers to identify poten-
tial terrorists via facial expres-
sion and behavior. A review of the
program in 2013 by the U.S. Gov-
ernment Accountability Office
found that the TSA hadn’t estab-
lished a scientific basis for it, and
that the program didn’t translate
to arrests. In 2017, a study by the

American Civil Liberties Union
concluded that the program fu-
eled racial profiling.
To get on the right track, Bar-
rett said, companies should be
working with far more data,
training their programs to con-
sider body language and vocal
characterization just as a human
would. At least one company says
it’s embracing the more multi-
faceted approach: Affectiva, the
first company to market “emotion
AI.” The company, which claims
to hold the largest collection of
emotion data in the world, works
with naturalistic video rather

than static images and is trying to
integrate such factors as a per-
son’s tone or gait into its analyses.
Rana el Kaliouby, Affectiva’s
co-founder and chief executive,
said she welcomes the review’s
findings, adding that they mirror
issues she has been trying to
tackle since she was a doctoral
candidate at Cambridge Univer-
sity in the early 2000s.
“I’ve tried to solve the same
issues all these years, but we’re
not there yet as an industry,”
Kaliouby said. “I liken it to the
emotional repertoire of a toddler:
A toddler will understand sim-
plistic states, but they won’t have
the language or sophisticated
sensing to recognize complex
emotions.”
Affectiva has encountered
these limitations organically in
the course of business and tried to
adjust to them, she said. Several
years ago, Kaliouby said, a client
complained that Affectiva’s tech-
nology wasn’t producing results
in China when it tried to analyze
responses to ads. It turned out
that the program had trouble
recognizing a subtle “politeness
smile” that many subjects dis-
played because of a lack of train-
ing specific to that region.
This helped highlight how dif-
ferent cultures display facial ex-
pressions, Kaliouby said, leading

Affectiva to incorporate “cultural-
ly specific benchmarks” for facial
movement. The company now
trains its systems with more than
8 million faces from 87 countries.
The industry will evolve as it
acquires more data, Kaliouby
said. In the meantime, the tech-
nology is already being put to use.
“The naive mapping that peo-
ple in academia and, unfortu-
nately, in the industry are doing is
quite dangerous, because you’re
getting wrong results,” Kaliouby
said. Often, she added, clients
aren’t interested in the more
comprehensive approach, asking
instead for analyses based on the
six basic emotions defined by
Ekman’s research decades ago.
Kaliouby said she hopes the
review will educate consumers
and those in the industry about
the limits of emotion recognition,
and reinforce how far the indus-
try has to go. The theoretical aim
for much artificial intelligence is
to create machines that are able
to operate just as well as, or better
than, a human.
But even creating a machine
with the perceptiveness of a per-
son wouldn’t be enough to solve
the problem of emotion detec-
tion, Kaliouby said.
“Humans get it wrong all the
time.”
[email protected]

‘Emotion detection’ AI has serious flaws, researchers say


Most systems rely on


facial expressions, ignore


tone and body language


BY LAURA REILEY

The Trump administration last
week revealed details of a $16 bil-
lion aid package for farmers hit in
the U.S.-China trade war, with key
provisions meant to avoid large
corporations scooping up big pay-
outs at the expense of small farm-
ers.
According to a report by the
nonprofit Environmental Work-
ing Group (EWG), most of the
$8.4 billion given out so far in last
year’s farm bailout went to
wealthy operations, exacerbating
the economic disparity with
smaller farmers.
An EWG analysis found that
the top one-tenth of recipients
received 54 percent of all pay-
ments. Eighty-two farmers each
received more than $500,000 in
trade relief.
DeLine Farm Partnership of
Charleston, Mo., has so far re-
ceived $2.8 million.
The top 1 percent of recipients
of trade relief received, on aver-


age, $183,331. The bottom 80 per-
cent received, on average, less
than $5,000, EWG said.
The Agriculture Department
said in a statement that the pro-
gram is designed to provide a
level of support proportionate to
a farm’s size and success. Pay-
ments were based on production.
The more acres they farm and
bushels per acre they produce,
the more assistance farmers re-
ceive.
“To our knowledge, USDA’s
payments have all been made in
accordance with our published
regulations and existing pro-
cedures,” the statement said.
“USDA follows producer eligibili-
ty protocols that are established
by Congress for other farm safety
net programs.”
In a year of trade tariffs, natu-
ral disasters and weather prob-
lems, and depressed commodity
prices, EWG senior analyst Anne
Schechinger said, some of these
farmers could be quadruple-dip-
ping from federal aid programs.
They may be receiving money
from either or both rounds of the
trade relief, agricultural-risk cov-
erage and price-loss coverage, as
well as crop insurance (especially
for those farmers who were un-
able to plant and took “prevent
plant” insurance payouts) and

finally money from the disaster
relief bill.
“These programs are designed
to give the most subsidy money to
the biggest farms; we don’t dis-
pute that,” Schechinger said. “But
for these huge farms that are
making tons of money each year,
is that really something taxpayers
should be subsidizing while
smaller farmers struggle?”
In 2018, payouts to individual
farmers hurt by the trade war
with China were capped at
$125,000; this time, the cap has
been raised to $250,000 per per-
son or legal entity, with a cap of
$500,000. Still, Schechinger said,
farmers can get around this easi-
ly.
“If your farm partnership has
20 people, each person could take
$125,000. You only have to be
‘actively engaged’ in farming, and
that is a super vague qualifica-
tion. It could mean that you call in
once a year to a shareholder
meeting.”
Some agricultural economists
say that changes to the second
round of trade relief payments
will further favor the largest
farmers by linking payments to
the number of acres, not the
number of bushels or bales pro-
duced.
In his weekly agricultural con-

ference call Tuesday, Sen. Charles
E. Grassley (R-Iowa) said the
EWG findings were in line with
his own views about farm pro-
grams.
“There should be a cap on them
because 10 percent of the farmers
get 70 percent of the benefits of

the farm program. The farm pro-
gram is meant to help people over
humps beyond their own con-
trol,” he said. “Some large farmers
do have the benefit of having
resources to get over those humps
without government help.”
According to EWG, nearly

28,000 farms have received pay-
ments from farm subsidy pro-
grams for 32 years straight. Total
subsidy payments across all pro-
grams in 2018, including the
trade relief package, totaled more
than $18 billion.
[email protected]

Richest farmers get most of bailout


DANIEL ACKER/BLOOMBERG NEWS
A soybean farm near Tiskilwa, Ill. The top one-tenth of recipients received 54 percent of payments so
far in the $16 billion package for farmers hurt by the U.S.-China trade war, a nonprofit found.

Aid package prompted by


trade war largely benefits


top producers, study says


“We’re not there yet


as an industry.”
Rana el Kaliouby,
CEO of Affectiva
Free download pdf