Applied Artificial Intelligence: How Machine Learning Transforms How We Live and Work

APPLIED
Artificial
Intelligence
How Machine Learning
Transforms How
We Live & Work
Book Preview
Contents
Introduction 3
1. Industrials & manufacturing 5

4 KEY DIFFERENCES BETWEEN INDUSTRIAL AND CONSUMER AI
2. Supply chain logistics 11

AI IS TURNING SUPPLY CHAIN LOGISTICS INTO
AUTOMATED TRADING
3. Telecommunications 18
SMARTER AI MEANS SAFER, SPEEDIER NETWORKS
4. Financial services 25
CHATBOTS GO CHA-CHING: THE LOOMING IMPACT
OF AI IN FINANCE
5. Marketing & Advertising 31

HOW A.I. CAN SOLVE THE TOP 3 PAIN POINTS IN MARKETING
6. Beauty & retail 41

CAN ARTIFICIAL INTELLIGENCE GIVE YOU BEAUTY ADVICE?
7. Travel 46
BALANCING MACHINE LEARNING AND HUMAN INTUITION
IN THE TRAVEL INDUSTRY
8. Love & dating 51

BITS OF LOVE: HOW AI FILLS OUR HUMAN NEED FOR INTIMACY
9. Healthcare 57
THE OPPORTUNITIES & CHALLENGES OF A.I. IN HEALTHCARE
10. Gaming 66
THE POINT OF PLAY: HOW GAMES MAKE AI SMARTER
Applied Artificial How machine Learning transforms

Intelligence How We Live & Work
3
INTRODUCTION
Artificial intelligence is the buzz word of the day. Youve no doubt
read your fair share of media hype either proclaiming doom and
gloom where robots seize our jobs or prophesying a new utopia where
AI cures all our human problems.
The reality is somewhere in between. Unless you work in the field, you
might be wondering how AI is used, or even what AI really is. Using
concrete examples and engaging storytelling, our book Applied Artificial
Intelligence helps you understand how these powerful emerging
technologies can be used in your own life and work.
From movie recommendations on Netflix, to partner suggestions

on dating apps, AI and machine learning already impact your daily
digital life. Even the routes you choose to commute to work and the
friends posts you see on Facebook or Twitter are heavily governed by
sophisticated algorithms.
But machine intelligence isnt just for consumers. Industries ranging

from manufacturing, logistics, telecommunications, financial services,
retail, healthcare, and gaming all leverage AI for a competitive edge.

4
ABOUT TOPBOTS
TOPBOTS is a strategy & research firm focused on applied AI for enterprises.
Our customers include leading global companies such as LOreal, Paypal, and
WPP. We advise business leaders, executives, and practitioners on emerging
technology trends and help you successfully apply them in your organization.
To learn how you can adopt automation technologies and AI at your own
organization, contact us at [email protected].
Mariya Yao
Head of Research & Design
HOW YOU CAN WORK WITH US
To discuss how you can adopt automation technologies and AI at your own
To get your executive team up to speed on emerging technologies and their
impact on your industry, ask us about our corporate education programs
by emailing [email protected].
To raise your A.I. IQ, read our publication at TOPBOTS.com or subscribe to
our newsletter.

5
1.
Industrials &
manufacturing

4 KEY DIFFERENCES
BETWEEN INDUSTRIAL
AND CONSUMER AI
Image Credit: Shutterstock / curraheeshutter
Robots are probably the first thing you think of when asked to imagine AI
applied to industrials and manufacturing. Indeed many innovative companies
like Rodney Brooks Rethink Robotics have developed friendly-looking robot
factory workers who hustle alongside their human colleagues. Industrial robots
have historically been designed to perform specific niche tasks, but modern-day
robots can be taught new tasks and make real-time decisions.

Image Credit: Rethink Robotics
As sexy and shiny as robots are, the vast majority of the value of AI in
manufacturing lies in transforming data from sensors and routine hardware into
intelligent predictions for better and faster decision-making. 15 billion machines
are currently connected to the Internet. By 2020, Cisco predicts the number
will surpass 50 billion. Connecting these machines together into intelligent
automated systems in the cloud is the next major step in the evolution of
manufacturing and industry.
In 2015, General Electric launched GE Digital to drive software innovation across

all departments. Harel Kodesh, CTO of GE Digital, shares with us the unique
challenges of applying AI to industrials that differ from consumer applications.
When looking at the industrial internet,

40% of the data coming in is spurious and isnt
useful - Harel Kodesh, CTO, GE Digital

1. INDUSTRIAL DATA IS OFTEN INACCURATE

For machine learning to work properly, you need lots of data. Consumer data is
harder to misunderstand, for example when you buy a pizza or click on an ad,
says Kodesh. When looking at the industrial internet, however, 40% of the data
coming in is spurious and isnt useful
Lets say you need to calculate how far a combine needs to drill and you stick a
moisture sensor into the ground to take important measurements. The readings
can be skewed by extreme temperatures, accidental man-handling, hardware
malfunctions, or even a worm thats been accidentally skewered by the device.
We are not generating data from the comfort and safety of a computer in your
den, Kodesh emphasizes.
2. AI RUNS ON THE EDGE, NOT ON THE CLOUD

Consumer data is processed in the cloud on computing clusters with seemingly
infinite capacity. Amazon can luxuriously take their time to crunch your browsing
and purchase history and show you new recommendations. In consumer
predictions, theres low value to false negatives and to false positives. Youll
forget that Amazon recommended you a crappy book, Kodesh notes.
On a deep sea oil rig, a riser is a conduit which transports oil from subsea
wells to a surface facility. If a problem arises, several clamps must respond
immediately to shut the valve. The sophisticated software that manages the
actuators on those clamps tracks minute details in temperature and pressure.
Any mistake could mean disaster.
The stakes and responsiveness are much higher for industrial applications where
millions of dollars and human lives can be on the line. In these cases, industrial
features cannot be trusted to run on the cloud and must be implemented on
location, also known as the edge.
Industrial AI is built as an end-to-end system, described by Kodesh as a

round-trip ticket, where data is generated by sensors on the edge, served

to algorithms, modeled on the cloud, and then moved back to the edge for
implementation. Between the edge and the cloud are supervisor gateways and
multiple nodes of computer storage since the entire system must be able to run
the the right load at the right places.
In a manufacturing facility that crushes ores into platinum bars, bars that come
out with the wrong consistency must be immediately detected in order to adjust
the pressure at the beginning. Any delay means wasted material. Similarly, a
wind turbine is constantly ingesting data to control operations. Kodesh highlights
one of many possible malfunctions: the millionth byte might be the torque on
the blade, but if the torque is too high, the turbine blades will fall off. We need to
serve this critical information first even if its in the millionth place in the queue.
Serving the correct data in real-time is a task so challenging that GE must rely on
custom, in-house solutions. Spark is fast, admits Kodesh, but when you make
decisions in 10 milliseconds, you need different solutions.
3. A SINGLE PREDICTION CAN COST OVER $1,000

Despite the high volume of faulty data and limited processing power at the edge,
industrial AI still needs to be incredibly accurate. If an analytical system on a
plane determines an engine is faulty, specialist technicians and engineers must
be dispatched to remove and repair the faulty part. Simultaneously a loaner
engine must be provided so the airline can keep up flight operations. The entire
deal can easily surpass $200,000.
Were not going to call and tell you there is a problem where there isnt, and
were definitely not going to tell you there isnt a problem when there is, says
Kodesh. We want to make sure we have very high fidelity systems.
According to Kodesh, the only way to ensure such high fidelity and performance

is to run thousands of algorithms at the same time. A consumer company like

Amazon might make $1 to $9 on a book, so theyre only willing to spend $0.001 for
each user prediction. With hundreds of thousands of dollars at stake, Industrial
and manufacturing giants spend between $40-$1,000 for each prediction.
For $1,000, I can run tons of algorithms in parallel, aggregate the results, and
run something known as a genetic algorithm to grow the predictor, Kodesh
reveals. This will create a survival of the fittest effect where the most fit
predictors are used and less fit predictors are discarded.
4. COMPLEX MODELS MUST BE INTERPRETABLE

Consumers rarely ask why Amazon makes specific recommendations. When the
stakes are higher, people ask questions. Technicians who have been in the field
for 45 years will not trust machines that cannot explain their predictions.
To achieve this high level of interpretability, GE needs to invent completely new

technologies. Unfortunately, the talent required is painfully scarce. I admire all
the schools trying to match marketplace demand with new data scientists, but
their mathematics are too shallow, Kodesh complains.
Real data scientists need more academic depth. They literally need to be rocket
scientists who know how to filter and normalize millions of data points in real-time.
A consumer company like Amazon might make

$1 to $9 on a book, so theyre only willing to spend
$0.001 for each user prediction. With hundreds
of thousands of dollars at stake, industrial and
manufacturing giants spend between $40-$1,000
for each prediction.

Section 11
2.
Supply chain
logistics

AI IS TURNING SUPPLY
CHAIN LOGISTICS
INTO AUTOMATED TRADING
Image Credit: Shutterstock / chuyuss
Ever wonder how your Amazon Prime packages show up at your door mere
hours after you place an order?
A complex series of operations connects suppliers to manufacturers to

wholesalers to retailers to you, the end consumer. Oversight of this process is
called supply chain management (SCM). Within SCM, logistics is the portion that
handles the movement of goods. E-Commerce giants like Amazon specialize
in logistics while consumer packaged goods leaders like Unilever provide full-
spectrum supply chain management services.

Like every other data-driven industry, logistics and supply chain companies are
investing in transformational A.I. solutions to tackle their most pressing pain
points. Both small and large enterprises are dabbling in innovations ranging
from machine learning to robotics.
A breakdown in logistics breaks the supply chain, so companies constantly

seek out improved ways to manage inventory, predict pricing, and streamline
operations. Chad Lindbloom, CIO of C.H. Robinson, a Fortune 500 multi-modal
transportation company, shares with us the top business use cases hes using AI
to tackle.
The largest portion of C.H. Robinsons business is North American truck freight.
A portion of their customers pre-commit to regular business and outsource
portions or all of their logistics needs. The remainder are one-off transactions,
for which the company is a surge provider for unplanned freight.
Surprisingly for a transportation company, C.H. Robinson owns no vehicles.

They are instead whats called a freight broker, an operational and financial
middleman between buyers who want to move freight and suppliers of vehicles
who can do the job. The supplier base is incredibly fragmented, ranging from
one man with a truck to massive fleets of co-owned vehicles. Despite these
capacity challenges, CHR must commit to move freight for a customer at a
specific price in advance. Sometimes theyre asked to quote a price a last-minute
same-day load. Other times they commit up to 2 years in advance.
HOW MUCH DOES A LOAD COST?

Price prediction is thus their biggest business challenge. The pricing in our
industry varies seasonally, by day of the week, by lane, by time of the day,
explains Lindbloom. A lane is an origin destination pair, such as Toledo, OH
to New York, NY. Note that reversing the lane, from NYC to Toledo, requires a
different price since urban centers dont generate high volumes of goods that
must be moved back to manufacturing zones.

Many vendors such as Watson Supply Chain, ToolsGroup, and TransVoyant offer
logistics and supply chain software with AI baked in, but the complexities and
nuances of C.H. Robinsons massive business require them to build in-house
technology tailored to their specific needs. Pricing was previously done by
human experts with deep domain experience and historical market knowledge.
Prior to becoming CIO, Lindbloom spent 25 years in finance and 15 years as CFO.
Combining financial with technical expertise, he and his team have built machine
learning models for price prediction that resemble those built by automated
traders on Wall Street. These models examine historical freight pricing data
along with concurrent parameters such as the weather, traffic, and socio-
economic challenges to estimate the fair transactional price on a spot basis.
AI doesnt always outperform market experts, which Lindbloom believes will not
be fully replaced. In some cases, humans come up with better price. In most
cases, the technology helps them hone in on the fair market price, he points
out. He also adds that a key benefit of effective algorithms is democratization
and accessibility of information. Instead of relying on a few industry experts to
produce estimates, more employees can use machine intelligence to ensure
theyre quoting within market so they dont lose the sale, and within capacity so
they dont botch the execution.
WHERE ARE ALL THE TRUCKS?

The second important use case is securing and managing the supplier inventory,
the vast and fragmented array of trucks available to transport loads.
CHR commits to a transport price for freight buyers before they know the exact
pricing and availability of requisite vehicles. The company relies on strategic
human relationships, specifically a vast trading network across employees to
find the right truck with the right capacity for the load.

For every lane, CHR runs background analytics to examine which carriers
have moved freight at what price and service level. Fragile, expensive, or time-
sensitive freight requires a much higher service level. Pooling together these
various factors allows CHR to optimize matching freight to the best mover.
CAN WE EXPECT THE UNEXPECTED?

Managing disruptions is the third important business task that can be improved
with AI. Hurricanes, carrier bankruptcies, and employee strikes all have the
potential to cause massive damages to the logistics business.
Predicting such disruptions and training AI to learn from contingency plans

developed by humans enables automated corrective action in the future. To
do so, CHR pulls together sources of information to analyze the impact of past
disruptions, such as a carrier strike in France or a hurricane in the NE United
States. If a distribution center is threatened with adverse weather, freight can be
re-routed to a safer one.
Part of the data collection entails detailed surveys that track how human
employees handled disruptions and the outcomes of their management.
Lindbloom hopes that eventually systems can be trained to automatically take
optimal actions after learning from humans.
HOW DO WE BUILD THE TECHNOLOGY?

We are constantly looking at whats on the marketplace, and we believe we
build better technology, says Lindbloom. Due to a critical need for reliability,
CHR builds and manages their own data centers, only going to the cloud if extra
computing power is needed. Owning data center resources allows CHR to spin
up environments very quickly as needed, but also to commit idle systems to
research and development.

In addition to flexibility, owning data centers enables privacy and control. We

are a cloud provider of transportation management system to our customers,
emphasizes Lindbloom. We have all the same technology as the core cloud
providers, but we know where all the data is, we can control it, and we make
confidentiality promises to customers. Many of them are more comfortable
Image Credit: Shutterstock / MAGNIFIER
We are a cloud provider of transportation

management system to our customers. We have
all the same technology as the core cloud
providers, but we know where all the data is,
we can control it, and we make confidentiality
promises to customers. Many of them are more
comfortable using us - Chad Lindbloom, CIO, C.H. Robinson

using us.
Technology is such a differentiating factor in our industry, Lindbloom

concludes. Other giants in logistics and supply chain agree and also committed
substantial dollars to AI initiatives. DHL aims to reduce costs with autonomous
cars, Active Ants builds wearable technology to optimize warehouse tasks,
Locus Robotics develops warehouse robots, and Honda leverages smartphone
applications for real-time shipment tracking.
WHERE WILL WE GO FROM HERE?

DHLs 2016 Logistics Trend Radar predicts that artificial intelligence investments
will continue to surge for both domestic and international logistics. Increasingly
more companies plan to invest in in-house development for AI applications in
predictive analytics, operations and management, augmented reality, robotics,
and industrial IoT.
Lindbloom has words of wisdom for those who want to replicate CHRs success
with AI: Many of the things youre going to try probably wont produce value. Be
willing to experiment and fail fast. Try to solve the same questions with multiple
different models. Multivariate-type testing is key.
Additionally, he cautions against overfocusing on AI and encourages executives

to define clear business use cases first. Have the business challenges drive
your development, instead of data scientists and engineers pushing AI into the
business.

Section 18
3.
Telecommuni-
cations

SMARTER AI MEANS SAFER,

SPEEDIER NETWORKS
Image Credit: Shutterstock / noolwlee
With more than 12,500 patents, 8 Nobel prizes, and a 140 year history of
field-testing crazy ideas, no one should be surprised that AT&T would be an
important player in artificial intelligence.
AT&T is a backbone of the internet, explains Nadia Morris, Head of Innovation

at the AT&T Connected Health Foundry. The company manages wireless,
landline, and even private secure networks to power connectivity for both
individuals and corporations. All these networks generate incredible volumes of
data ripe for machine analysis.

AT&T has built AI and machine learning systems for decades, using algorithms to
automate operations such as common call center procedures and the analysis
and correction of network outages. On the entertainment side, AT&Ts DirecTV
division leverages users rating histories, viewing behaviors, and other factors to
anticipate the next films theyll watch.
AI & DATA VISUALIZATION HELP OPTIMIZE 5G ROLLOUTS

Modern A.I. algorithms have enabled the telecom company to tackle even more
complex tasks, such as optimizing the rollout of their 5G network. Traditional
cell towers are usually suboptimally placed near urban centers and form an
imperfect grid, leading to gaps in coverage. Theyre also expensive to put up and
maintain and incur challenges with real estate and property ownership.
Small cells are less expensive, more compact cells that can be installed on inner
city buildings on a much finer grid. Their role is to repeat the signal from the
main cell towers to get closer to end users. By crunching mobile subscriber data,
well-calibrated AI can help create spatial models to hone in on ideal spots to
build small cells to ensure maximum 5G signal strength for customers.
Designing the right 5G infrastructure is critical, especially given the rapid rise
of video. Video is more than half of our mobile traffic, explains Chris Volinsky,
who leads big data research at AT&T Labs. Video traffic grew over 75% and
smartphones drove almost 75% of our data traffic in 2016 alone. We expect
video traffic growth to outpace overall data growth in 2020.
Infrastructure is an enormous investment, even with small cells, so accurately

modeling trends and usage growth is key to success. Demographic trends can
cause previously underutilized areas to suddenly become hot traffic generators.
While statistical models are useful for identifying trends in customer movement
and throughout, AI and machine learning techniques create future projections
from current data.

We need to visualize billions of data points in a spatiotemporal fashion,

Volinsky elaborates. No tools existed previously to address AT&Ts unique data
challenges, so they built and open-source custom tools such as Nanocubes, a
data visualization tool that can map out millions of connections of individual
mobile phones and connected devices to cell phone towers. The tool has been
used outside the company to characterize sports fans in real time and analyze
crime rates and history.
Image Credit: Shutterstock / AT&T
Algorithms and tools are not the bottleneck in solving problems. Volinsky
clarifies that the challenge is in the data and the data pipeline. Modern data-
hungry AI approaches require a centralized data source, but gathering one
across a myriad of networks with idiosyncratic standards is no trivial task. Each
small cell collects cellular data differently. Some track 4G but not 3G. Some dont
get iPhone data. If variations are not taken into account, bias will appear in the
data and the results.
There is no world expert in data munging, Volinsky bemoans. To succeed,

you have to figure out organizationally how to access data in different silos,
technically how to integrate with it, and ensure the formats are in line. Data
scientists often discover that they cant solve the problems they want to solve
because the fundamentals of managing data is difficult and time-consuming.
This is not the stuff people learn in grad school, he warns.

OLDER METHODS CAN BE JUST AS USEFUL AS NEW ONES

Volinskys convinced that AI is the most powerful addition in the toolbox used
by AT&Ts research arm to develop the next generation of enterprise and
consumer-facing solutions. At the same time, he cautions against using deep
learning as a magical black box to solve all problems. Instead, you should
prioritize solid data infrastructure, subject matter expertise, and utilizing an
ensemble of methods from data science and machine learning toolboxes.
Volinsky would know best. His BellKor team won the coveted $1 Million Netflix
Prize in 2009. The key lesson learned during the three year competition was the
power of ensembles. Ensembles involve combining various methods ranging
from regression, support vector machines, singular value decomposition,
restricted boltzmann machines, and neural networks to produce a result.
Deep learning is a power tool in your toolbox, but you still need your old school
tools to solve problems, he emphasizes. Deep learning evangelists say neural
networks effectively incorporate all the other models, but I have not seen that
work in practice.
ECONOMY OF SCALE HELPS STARTUPS & HOSPITALS

In tandem with in-house projects, AT&T operates six innovation labs, called
Foundries, all over the world. Each Foundry specializes in a different industry.
As Head of Innovation at AT&Ts Connected Health Foundry, Nadia Morris

works with aspirational startups such as AIRA, a smart wearables startup that
uses human-assisted computer vision algorithms to enable the blind and
vision-impaired to visualize their surroundings and navigate their immediate
environment.
Using established manufacturing relationships, AT&T helps healthcare IoT and

wearables companies like AIRA accelerate their hardware prototyping and
production. Similar to the Labs, the Foundries also leverage custom-built open-
source tools such as Flow Designer, a rapid prototyping tool that simplifies
hardware design for software engineers.

Remember Morris earlier comment about how the internet runs on AT&T? Turns
out this can be mission critical for startups like AIRA which must ensure superior
connectivity at all times to protect the safety of their patients. Since AT&Ts AI
systems regulate network traffic, they can intelligently detect AIRA devices on
their network and dynamically allocate greater bandwidth to support live video
streaming.
AT&Ts control of networks also comes in useful for hospitals who hoard
sensitive patient data. Fearful of security lapses, many operate their own data
centers for fear of uploading personal information to the cloud. Data center
management is typically not a hospitals core competency, leading to outdated
technology and massive inefficiencies.
Do you want to run a hospital or do you want to run a data center, questions
Morris. Regardless of the cloud provider a hospital chooses to use, AT&T runs
private network connections to all of their servers. This traffic will never traverse
the public internet, she assures, giving hospitals an extra layer of protection.
Migrating more hospitals to the cloud solves not only administrative pains,
but also unblocks AI research. Hospitals are smart, but theyre like islands,
Morris explains. Competition often incentivizes hospitals to hoard data that is
critical to share for superior results. Pooling hospital data into collaborative
cloud communities and applying de-identification protocols enables medical
researchers to access disparate data sets with greater geographic diversity.
Algorithms for essential patient services such as vital sign monitoring can be
trained on aggregate data sets for more accurate benchmarks.
HUMANS WORK WITH BOTS TO ENSURE CUSTOMER SUCCESS

Lead Inventive Scientist Wen-Ling Hsu has been with AT&T for over 20 years. She
obsessed over creating amazing customer experiences using massive data and
information even before big data was coined.

Hsu analyzes customer conversations from both phone conversations from call
centers and online chats with support agents. Machine learning allows her to
build textual models, identify customer intent, and route them to appropriate
support agents faster.
With her extensive experience, Hsu learned that interpreting and using the
intelligence gained from AI systems is more of an art than a science. What
matters most is customer perception and seamless execution, so Hsu employs a
combination of bots that directly interact with customers and those that stay in
the background to assist human agents.
When asked to make a forecast for AI in 2017, Hsu responded, Human

judgment still plays a critical role in many tasks. Together, AI bots and human
agents can learn from every customer interaction to personalize the customer
experience.
Deep learning is a power tool in your toolbox,

but you still need your old school tools to solve
problems. Deep learning evangelists say neural
networks effectively incorporate all the other
models, but I have not seen that work in practice.
- Chris Volinsky, Assistant Vice President, Big Data Research, AT&T

Section 25
4.
Financial
services

CHATBOTS GO CHA-CHING:
THE LOOMING IMPACT OF AI
IN FINANCE
Image Credit: Shutterstock / ImYanis
From minting coins to dispensing greenbacks on ATMs, the love affair between
money and machine goes a long way back. The pervasive influence of technology
in how we create, exchange and store money treads a colorful history replete
with culture-shifting innovations such as cash registers, magnetic credit cards
and mobile wallets.

The romance is far from over. With recent advances in machine learning,
computer vision, voice recognition, natural language processing and other
areas of artificial intelligence, the chemistry between money and machine is
just warming up. Accenture recently reported that the vast majority of industry
insiders believe AI will become the primary channel through which banks and
their customers will interact within the next three years.
Tell-tale signs are all over the place:
Capital One recently launched an NLP-capable chatbot named Eno at the wake
of other industry firsts in terms of AI applications. Eno enables customers to
chat with the bank using text-based natural language to pay bills and retrieve
account information. Among the pioneering financial institutions to join the IoT
bandwagon, Capital One also launched an Alexa Skill for Amazon Echo and plans
to be the first to launch a similar service for Microsofts Cortana.
Mastercard is leveraging AI to enhance experiences across its ecosystem (for

consumers, partners, issuers, and merchants), as well as to hike operational
efficiency and close loopholes in areas such as fraud and false declines. The
financial services giant worked with AI enabler Kasisto to build chatbots for
banks and merchants to better engage consumers. Taking a step further,
Mastercard also partnered with IBMs Watson and General Motors to explore
scenarios where users can safely conduct commerce even while on the road
(such as pre-ordering their favorite cup of coffee and picking it up from the
drive-through counter on their way to the office).

Thomson Reuters has used complex algorithms for years to organize their
humongous stores of information and generate time-critical financial
market data for institutional clients. Their use of AI centers on improving the
experience of knowledge workers by building systems that not only answer
questions on structured data (e.g., give me all age discrimination cases in
California in the last four years) but also provide proactive insights for decision
makers (e.g., making a comparative macroeconomic analysis of two countries
or showing differences and trade offs between two investment options). As a
leading content publisher, Thomson Reuters also uses machine learning and
AI to detect and identify fake news. The Reuters News Tracer leverages an
algorithm that looks at more than 700 factors to determine whether a trending
topic on social media is factual or not.
Wells Fargo leveraged machine learning originally to prevent fraud, but AI now
permeates other areas including compliance, customer experience, underwriting
and authentication. The San Francisco-based financial institution is also
exploring the use of consumer-facing virtual assistants for information updates,
improvements in transaction capabilities and business insights. Bipin Sahni,
EVP and Head of Innovation and R&D, reports that regularly collaborating with
startups on a wide range of technologies helps us explore big ideas outside our
walls.
After unleashing Watson on Jeopardy and offering it as an AI platform for

companies in the healthcare, travel and other industries, IBM retrofitted its
question-answering and NLP-capable brainchild for the financial sector in
December 2016. Beta-launched as IBM Watson for Cybersecurity, the program
already attracted dozens of clients, including key players in the insurance
and banking industries. As reported by Business Insider, global firms such
as Sumitomo Mitsui Banking Corp. and Sun Financial will help test Watsons
ability to identify and combat cyberattacks. In addition to accurately detecting
suspicious behavior, Watsons machine learning component will help improve its
ability to interpret and analyze cybersecurity data over time.

Even personal finance and wealth management have come to favor artificial
intelligence, as shown by the rapid spread of AI-driven mobile apps, the
comparatively higher success rates of quantitative trading and the cost-efficiency
of robo advisors. Robo advisors such as Wealthfront and WiseBanyan are
automated portfolio management services that use computer algorithms to
manage and grow customer investments.
Banks and other financial institutions have been using computer automation to
improve operational efficiency for decades. With big data and AI technologies
on the uptrend, the development of smarter, better and more powerful
automations will profoundly transform the industry. From providing regulatory
compliance requirements to expediting reports generation, AI will become a
pervasive force across all the components of a financial organization.
THE HUMAN MINDS BEHIND ARTIFICIAL INTELLIGENCE

TOPBOTS interviewed key industry leaders to make sense of the profound
changes triggered by different applications of artificial intelligence in the world
of finance. These influencers illuminated the details of how AI research and
development work in their respective organizations.
For Margaret Mayer, VP of Software Engineering at Capital One, customer needs

and feedback set the tone for any AI initiative at the company. She said, We use
analytics in general to look for customer patterns, whether that be spending
patterns or concerns about fraud.
Mayers colleague, AI Design Head Steph Hay believes context is the biggest
design challenge. Hay described the importance of context as the sweet spot
between design and technology. According to Hay, AI developers often shift
between two opposite directions: one that leverages the companys massive
datasets to build a common template for customer interactions; and another
that needs to be resolved as a unique scenario involving a single customer and
the bank. She emphasized her focus on how to make better assumptions and
provide smarter responses based on everyones data, not just your data.

Meanwhile, Thomson Reuters VP of Research and Development Khalid Al-

Kofahi has been a domain expert for his entire career and seeks to demystify
the buzz around AI. He recounted the evolution of AI at the company as
programmatic functions driven by analytics and designed to establish a specific
user experience. For Al-Kofahi, the most difficult thing in AI development is
finding the right problem and scoping it right. Otherwise, talent, time and other
organizational resources can easily be wasted.
Mastercards forays into AI-focused research and partnerships stem from its
corporate mandate to enable commerce across every device and environment,
according to Kiki Del Valle who serves as SVP at the firms Commerce for Every
Device unit. Scheduled for a year-end rollout, the companys joint project with
IBM and General Motors will use voice recognition technology and NLP to locate
nearby establishments and make quick, effortless transactions. While AI has
tremendous potential, Del Valle cautions organizations against over exposure:
Unless its solving a problem or addressing a specific need, then dont jump into
AI for the sake of it, citing the crucial importance of scalability from the point of
view of issuers, merchants and other stakeholders.
Wells Fargo currently works with promising AI startup Kasisto to enhance

customer experiences via a conversational AI service. Head of Innovation Group
Steve Ellis sees virtual assistants helping in three key areas: information updates,
transaction capabilities, and customer insights. We see a big future here, he
says, and thats why we recently formed our new Artificial Intelligence Enterprise
Solutions team within the Innovation Group.
BANKING ON AI TODAY
Artificial intelligence already has significant imprint in many areas of banking
such as fraud detection, insurance underwriting, customer service, back office
operations and wealth management. Because it demonstrably improves
efficiencies, reduces costs, generates smarter insights and provides better user
experiences, AI will eventually be used industry-wide at an even deeper and
broader scale. The only thing to expect are new and surprising use cases as bot
and bank become one.

Section 31
5.
Marketing
& Advertising

HOW A.I. CAN SOLVE

THE TOP 3 PAIN POINTS
IN MARKETING
Image Credit: Shutterstock / Jirsak
3,874. Thats how many companies are featured on Scott Brinkers behemoth
2016 Marketing Technology Landscape Supergraphic, which drives home the
challenge of navigating the marketing industry.
Marketing has the unique challenge of not having a typical stack or process. If
you look into any Fortune 500 company, they will have hundreds of products
that they are stitching together, reports Eric Stahl, an SVP of Product Marketing
at Salesforce.

Leading marketing experts agree that the plethora of tools available to

marketers and advertisers is both a blessing and a curse. The variety of
options enables marketers to keep up with an ever-evolving digital landscape
of consumer and customer behavior, but a professional uses between 20-50
different tools to manage all their tasks.
The result is more time spent managing software than managing strategy. Every
marketing executive we spoke to also emphasized that vendor variation and
incompatibility drives the number one challenge of modern marketing: sourcing
and centralizing reliable data.
Every marketing executive we spoke to

emphasized that vendor variation and
incompatibility drives the number one challenge
of modern marketing: sourcing and centralizing
reliable data.

To understand how A.I. can address modern marketing challenges, you

first need to understand the tools marketers use every day. If you arent a
professional marketer, read our brief introduction to marketing technology
covering the typical enterprise stack and various add-on solutions.
CHALLENGE #1: LACK OF RELIABLE, CENTRALIZED DATA

Every function in a corporation struggles with data collection, cleansing, and
centralization, but inconsistent standards and non-existent integrations make
consistent data flow especially difficult for marketers.
Along with tool variation, marketing tech (MarTech) and adverting tech (AdTech)
are often siloed from one another. Ritchie Hale, CIO at TouchCR, warns that the
separation of these two stacks creates an identity gap, where we are trying to
uniquely identify the person we are advertising to before we have an identifiable
piece of information from them.
One key driver of the lack of interoperability between tools is the breakneck
speed of technical development and acquisitions in MarTech. Most DMPs and
marketing automation tools possessed by companies like Oracle, Adobe or
Facebook were acquired recently and havent even been integrated with each
other, explains Monika Ambrozowicz, Global Marketing Manager at Synerise.
Another problem, pointed out by Mark Kovscek, President of Velocidi, lies

in the nuances of preparing data for different marketing purposes. As the
data is cleansed and prepared, it is optimized for a specific use case (e.g.,
programmatic, campaign reporting, media attribution) and therefore creates
competing versions of the data.
Without reliable, centralized data, marketers are doomed to suffer many

inefficiencies and lost opportunities. Many key decisions, such as creative,
messaging, and campaign parameters are essential to the success of marketing,
yet driven by gut or laziness rather than science. Even with haphazard decision-
making, marketers are still spending over three hours on average every week
just analyzing disparate data sources.

Another insidious effect of not coordinating between tools is the risk of

undermining your own marketing campaigns. Mark Torrance, CTO of RocketFuel,
warns if multiple partners are used to buy traffic based on similar data or
targeting ideas, they may bid against each other, driving up the cost (and driving
down the efficiency) of the buy.
The effects of poor data centralization also lead to a poor consumer experience.
Retailers employing ad retargeting often dont realize when youve already made
a purchase and send you tons of irrelevant, useless, and annoying ads.
How AI Addresses This Challenge

Many MarTech companies aim to be an A.I. layer that centralizes and manages
communication and data across marketing tools. The players best positioned to
win are major enterprises like Salesforce, Oracle, and Adobe, which already offer
end-to-end solutions within their own ecosystems and can aford to aggressively
acquire and integrate smaller players.
Salesforce CEO Marc Benioff spent over $4 Billion in 2016 buying A.I. companies
to roll into Einstein, an A.I. layer that optimizes results across all of Salesforces
Clouds. One client, Fanatics, is a sports merchandise retailer using Marketing
Cloud Einstein to personalize product recommendations.
Sports fans come in different flavors, such as hardcore fans, gear fanatics, fans
of a school or local team, and fans of fans, i.e. customers purchasing sports
gear for friends and family. According to Stahl, an SVP at Salesforce Marketing
Cloud, Einstein segmented customers into these large buckets, but also provided
sub-segment targeting that led to 15-20% of clickthroughs on Fanatics email
campaigns.
The same personalization can be ported to Salesforce Commerce Cloud, where

Einstein has helped some customers achieve 28% more revenue and 11%
increase in average order value (AOV) from better recommendations.

Similarly, Adobe controls a vast suite of creative tools as well as a popular data
management platform (DMP) and web analytics for enterprises. Vice President
Amit Ahuja oversees Sensei, Adobes AI layer which unifies data across their
cloud solutions. Since Adobe owns the underlying video and content creation
tools, Ajuha explains that Sensei is able to collect rich data across all creatives
and all metadata to inform a brand owner which creative to put in front on a
consumer.
When the A.I. hype dies down, the differentiation will be at the data layer,
predicts Ahuja. Outside of Google, were sitting on the biggest system of record
for any behavioral digital data. No one else can do what we do.
Behemoths like Salesforce and Adobe are not the only companies tackling
the data challenge in marketing. Smaller companies like Swiftype address
the common bottleneck of coordinating assets and documents in marketing
workflows.
Waiting on the creative department to send you copy and graphics while
analysts pull the latest campaign performance metrics is painful and inefficient.
Swiftype consolidates knowledge and data from multiple sources, like Marketo
and Salesforce campaigns, task management tools, and repositories like
Dropbox and Google Drive.
Customers in commerce reap the benefits in obvious ways (when a search

for spoon also serves up search results for items labeled ladle, for example),
explains Praveena Khatri, Swiftypes VP of Marketing, while publisher customers
have more effective indexing of stories and can organize content by relevancy
and date.

CHALLENGE #2: TALENT BOTTLENECK

Mastering a myriad of tools also presents training challenges for marketing
teams and creates expertise bottlenecks. Training junior employees to navigate
complex enterprise software is tedious and error-prone. To cut corners,
employees often build campaigns around only one or two parameters, explains
Mark Shore, Co-Founder of Strike Social, which leads to the lowest common
denominator and weaker performance.
Another challenge with larger enterprise is dependency on external talent

for essential work. Torrance of Rocket Fuel explains that most Fortune 500
companies rely on marketing partners to buy media in the walled gardensof
Google, YouTube, and Facebook. Additionally, most of them still have the bulk
of their media spending controlled and managed by Ad Agency partners, who
have their hands on the keyboards of the platforms.
If mastering marketing tools is difficult, mastering A.I. research and development

is even harder. Very few organizations are positioned to succeed on this front.
Even if a company miraculously centralized reliable and high-volume data in a
single system, specialized talent is still required to design and operationalize
working models.
Fortune 500 companies simply do not have the inclination or resources

internally and they routinely, if not universally, fail at the BUILD. So the question
is how to BUY, advises Tasso Argyros, CEO of ActionIQ. Even companies with
teams of data scientists and engineers on hand often find their employees lack
the requisite advanced mathematics background to truly innovate in modern
artificial intelligence
Engineers alone also do not ensure success. Marketing executives agree that
domain expertise and business needs should drive A.I. research, not the other
way around. The most important thing is not advanced math models or
complicated neural networks, Yulia Khansvyarova of SEMrush clarifies. The
most important part is feature engineering. The more domain knowledge you
have, the better. Constantly verify your hypothesis with your end users.


Automation of tedious marketing tasks improves accuracy and reduces
workload, allowing marketing teams to be more efficient and effective. While
the landscape of providers offering automated solutions is still crowded, many
marketing executives are seeing early traction.
Shore of Strike Social claims their technology automates the tedious process of
campaign setup, spots nuanced patterns undetectable to humans, and breaks
ad campaigns into several micro-campaigns and shifts ad dollars to the best-
performing targets in real time. The company was able to improve YouTube
view rates by 25% while reducing execution time by 75%. By reducing execution
time, Shore clarifies, we mean reducing your staff by up to 4x.
RocketFuel is another A.I. driven marketing company replacing manual

optimization with smarter automation. We do this using a variety of machine
learning and optimization techniques, including neural networks, logistic
regression, multi-armed bandit, performance-aware pacing, bid multipliers, and
more. We frequently balance scores from 2 or more models to achieve multi-
objective optimization, explains Torrance, RocketFuels CTO.
One agency held a four way competition with RocketFuel and three other
vendors, all of whom were doing manual optimization. RocketFuels A.I. powered
optimization bested the competitors by 8x on cost per acquisition (CPA).
Aside from invisible automation layers, innovative marketing firms are also
exploring conversational approaches. Equals3 built Lucy, a cognitive companion
for marketers as described by Managing Partner Scott Litman. Lucy acts as
your trusted marketing analyst, helping you with research, segmentation, and
planning. Only she works 24/7 and gets smarter with more data. Litman claims
that Havas Media, one of the worlds largest media agencies, has successfully
used Lucy to achieve a 75% reduction in vendor cost and 7x faster campaign
deployment.

CHALLENGE #3: INABILITY TO CALCULATE ROI

Data is not created equal. Marketers have a hard time turning data into insights,
much less calculating the ROI of their decisions.
Ben Plomion, CMO of GumGum, highlights the pervasive pain point where
brands spend more than $60BN globally on sports sponsorships but are unable
to capture the full value of their logo impressions on broadcast TV and also
social media.
Politics may also be interfering with true ROI calculations. Many marketing
departments fear accountability and deliberately cherry-pick metrics to present
to executives rather than analyze the hard truth of what is really working.

New neural network approaches like deep learning have the superhuman ability
to detect patterns, leading to many recent breakthroughs in image recognition
and computer vision. Computers can not only reliably classify objects in photos
and videos, but also identify specific brands and products.
Plomion of GumGum explains why such breakthrough technologies

revolutionize brand marketing: Without technology, analyzing a 3 hour game
for different logo placements can take days. By leveraging computer vision
technology, a machine can simultaneously analyze every sponsor and location
within every frame of the video in a matter of seconds, allowing a full game to
be analyzed in a few hours or less. Suddenly, computing ROI on sponsorships
or advertising campaigns becomes tractable.
Visual intelligence also enables brands to drive engagement through

personalized customer experience. Stackla helps brands discover the best
user-generated content (UGC) around their brand, categorize it around
customer personas, and recommend the right content for the right marketing
channel, explains CTO Peter Cassidy.

Using user content found on Stacklas platform, Virgin Holidays increase

bookings by 260% from the previous year and TOPSHOP increased sales of
online products by 75%. Cloudsight, another visual intelligence company, helped
customers achieve a 4x growth in time-on-site by showing the most relevant
images for each user, according to co-founder Brad Folkens.
A.I. can also replace older methods to better value different data sources and
turn them into more accurate business insights. Superior analysis of audience
segmentation and responses lead to improved understanding and handling of
customer and lead behavior, such as predicted purchases or churn. TouchCR
grew their own business by 20% while reducing ad send by 60% by using A.I. to
match marketing efforts to demographic and psychographic indicators.
Kazuhiro Takiguchi, CEO of ReFUEL4, produced similar results for clients like
Spotify which saw a 40% increase in clickthrough rates (CTR) and 3x app installs
over previous campaigns with A.I. powered marketing. According to Takiguchi,
such results are achieved by taking existing and past ad performance to predict
future performance of creative.
CONCLUSION
The holy grail for marketers has always been personalization at scale with
affordable customer acquisition costs, states Pascal Bensoussan, CPO at
Reputation.com. To achieve this end goal, AI and machine learning are
becoming table stakes in all the core components of the marketing/advertising
technology stack.
Challenges with data capture and centralization, as well as recruiting and

training, will continue to plague marketing teams, but the rise of artificial
intelligence and machine learning offers a clear way to chip away at these
previously insurmountable obstacles in modern marketing.

Section 41
6.
Beauty & retail

CAN ARTIFICIAL
INTELLIGENCE GIVE YOU
BEAUTY ADVICE?
Image Credit: Modiface
The beauty and skincare world is oversaturated, especially if you include all the
affordable convenience store brands. If youre a shopper with a budget, you are
likely mixing different products and blindly guessing which combinations will
work like a chemist without a periodic table.

AI technology has already revolutionized transportation, food, and even

health. Why not also beauty and retail? Sephora has launched multiple
chatbots, including a bot on Facebook Messenger that lets you book makeover
appointments in stores and a bot on Kik that gives product recommendations
and beauty tutorials. One of their newest bots, Sephora Virtual Artist, is powered
by Modiface, an augmented reality (AR) startup that uses AI to detect faces and
project virtual makeup looks in real-time. Olay, a well-known drugstore brand
from Proctor & Gamble, created deep learning algorithms to analyze your skin
from your selfies and tell you which beauty products to buy.
What weve noticed through the last couple of years is that skincare is a very
high engagement category and its become one of the most confusing and least Image Credit: Olay Skin Advisor
fun-to-shop categories as well, said Dr. Frauke Neuser, principal scientist at

Procter & Gamble, which owns the Olay brand. About a third of women walk out
of the store without having found that right product for her.
Analysis paralysis from the plethora of drugstore options leads women to

resort to department stores where they can have consultations. But these
consultations can also feel like overly aggressive ways to sell you products
you dont really want or need. Olay packaged over thirty years of skin analysis
and imaging expertise into the Olay Skin Advisor, a mobile experience for
empowering consumer choices.
20 years ago, Olay developed a hardware tool for skin analytics called the Visia
Imaging System. Visia took controlled facial images of users in different lighting
conditions and tracked skin conditions such as wrinkles, pores, and textures.
Dermatologists took Visia on road shows to analyze how customers skin
compared to others of their age.

Such inventions enabled Olay to collect a massive proprietary database of face

and skin images from a wide variety of ethnic and demographic backgrounds.
Photographic characteristics such as lighting, quality, and shot angles were also
diverse. Two different teams the bioinformatics group and the image analysis
experts collaborated to develop and apply in-house deep learning algorithms
to over 50,000 images to determine how skin changes over time and the impact
of various products.
Eric Gruen, P&Gs Associate Marketing Director, emphasizes that the Olay
Skin Advisor is not a downloadable mobile app, but rather a widely accessible
web experience that consumers can access through mobile browsers. The AI-
powered advisor has been used over 1.2 million times and consistently attracts
5,000 to 7,000 users every day.
The Olay Skin Advisor team tested a number of designs for the product to create
the best user experience. Psychologically, users feel advice is more trustworthy
when personalized. To gather the right information from each user, the Skin
Advisor asks a number of questions, such as What is your age?, to factor into
recommendations. After testing between 4 and 19 questions, the ideal number
turned out to be 9. Just like a human beauty consultant, the Skin Advisor also
considers a womans desired skincare regimen and special requirements.
Modiface started with AR for beauty products over a decade ago. 3 years ago,
beauty brands noticed that AR was driving sales. Now 80 of the top 100 makeup
Image Credit: Olay Skin Advisor
lines use Modiface, including Sephora, Urban Decay, LOreal, and Vichy. CEO
Parham Aarabi shares that his company grew over 400% in the last 12 months to
meet the incredible demand.

Image Credit: Modiface
With Modiface technology, users are able to upload their own photos and
virtually try on lipsticks, eye shadows, hair colors, and even dramatic new hair
styles. Integration with a new beauty brand typically takes 2-3 months and, once
launched, users typically try on 20 or more products per session. These fun, risk-
free digital trials lead to an 80% lift in makeup sales when Modiface is added to a
brands existing website or mobile app.
Aarabi, who co-published a paper called Hair Segmentation Using Heuristically-

Trained Neural Networks, plans to use modern AI methods to further personalize
the user experience. Since the appearance of hair can vary based on gender,
age, ethnicity, and the surrounding environment, automatic hair segmentation
is challenging, he explains, but emphasizes that deep learning techniques can
solve the problem. Additionally, Modiface plans to use AI for better tracking and
detection of faces for improved realism.
Todays beauty shopper no longer needs to rely on guesswork, Googling, and

dumb luck to find the right products for her personal needs. With the promise
of AI, even simple decisions such as your drugstore purchases can be guided by
thousands of data points and smart algorithms.

Section 46
7.
Travel

7. Travel 47
BALANCING MACHINE
LEARNING AND HUMAN
INTUITION IN THE TRAVEL
INDUSTRY
Image Credit: Shutterstock / NicoElNino
Travel planning is incredibly stressful. Between researching options, paying for

bookings, and organizing your itinerary, you may also have to contend with the
risk of being beaten and dragged off planes.

7. Travel 48
Machine intelligence can alleviate some of the pain points for both you and
the travel companies you book with. Perhaps no one knows this better than
Giorgos Zacharia, CTO of Kayak and holder of a Ph.D. in artificial intelligence and
machine learning from MIT. AI is kind of a fashionable domain at present, he
says, amused by the recent hype, but weve been doing machine learning and AI
at Kayak for a long time.
Almost every aspect of your digital user experience is improved with AI. Your
preferences towards specific seasons, hotel styles, and price parameters are
carefully monitored so that you can be served results youre likely to book. The
photos you see on hotel listings are run through thousands of split tests in which
users rank different versions and the results are optimized for mass appeal.
Turns out we prefer our hotel photos to be clean and pristine and dislike when
they feature other people.
Have you ever gone through an entire hotel or flight booking process, only to be
told at the end the item was unavailable? Like many industries, travel companies
suffer from inconsistency in data. Due to a slew of legacy systems, changes
in hotel and airline databases might not fully propagate in time to booking
providers to reflect real-time supply. To combat this problem, Kayaks algorithms
analyze a wide variety of historical sources to generate a more accurate forecast
of availability.
Another common data challenge is handling duplicates. With all those records
coming from different systems, you can have misspellings, different word
orders, and other issues that could cause a system to create duplicate records,
explains Zacharia. For example, a single listing could be titled differently as
the Boston Marriott Hotel or the Marriott Hotel In Boston. To save time,
the de-duping process is largely automated by machine intelligence. Only low-
confidence records, where the algorithm isnt sure of a prediction, are escalated
to human staff for analysis. Records from different sources may even disagree
about basic facts, such as whether a hotel has a pool, but Zacharia assures that
these algorithms can rationalize that data very, very quickly.

7. Travel 49
Machine analysis can yield surprising learnings that contradict your intuition.
In a previous role before Kayak, Zacharia built systems to predict corporate
bankruptcy filings. One month before a bankruptcy, a companys credit score
often sees a dramatic improvement. The revelation led to further investigation.
Turns out CFOs of at-risk companies desperately start paying back overdue bills
in the hopes of getting another loan, but typically fail.
Similar findings occur in the travel space. For example, users care less about the
average review score of a hotel and more about the number of reviews. A hotel
with fewer than 24 reviews is far less likely to be booked even if the comments
are overwhelmingly positive. Users also have a sophisticated nose for spotting
good deals. Broadcasting a clear discount typically results in higher conversions,
but even when a hotel discounts rooms without revealing original prices, buyers
intuitively flock to the deal.
Kayak is hardly the only travel company leveraging machine learning. Booking.
com, helmed by CEO Gillian Tans, prides itself on international reach, listing
properties in over 225 countries in 43 languages.
Many dont realize that Booking.com is one of the biggest translation companies
in the world, according to Tans. Headed to a foreign country where you dont
speak the language? No problem. Bookings cross-platform chatbots allow guests
to connect with overseas hotels and hosts and perform real-time translations for
all of their supported languages.
Booking travel is not like shopping, or groceries

or booking a restaurant. Its much less frequent,
so understanding what works just takes a lot
more time. - Gillian Tans, CEO, Booking.com

7. Travel 50
In addition to translating human-to-human conversations in real-time, Booking.

com bots act as 24-hour customer service agents able to answer most simple
travel questions. Kayak also has bots on Facebook Messenger, Amazon Echo,
and Slack, in order of popularity. While the natural language processing (NLP)
behind the bots are the same, Zacharia notes that users approach different
platforms with different intents. On Alexa, we get more aspirational queries,
such as how much a flight is to Hawaii. On Facebook, we get more lower-funnel
queries after a user has already booked, such as where they should eat, he
reveals. Complex questions still need to be handled by humans, but Tans
remarks it is surprising how much you can do with machine learning and how
good it is getting.
While the improvements brought by machine learning are impressive, the travel
industry must overcome many barriers to reach Tans ultimate vision of AI being
a fully-functional digital travel assistant that can proactively solve potential
issues before you even know they exist.
Booking travel is not like shopping, or groceries or booking a restaurant. Its

much less frequent, so understanding what works just takes a lot more time.
Tans also emphasizes that we must aim for the correct balance between human
interaction and sufficient automation.
Travel is a combination of the personal and the emotional, she says. Every
customer is different and the travel experience is completely fluid, but the end
goal is to find the best solutions.

Section 51
8.
Love & dating

8. Love & dating 52
BITS OF LOVE: HOW AI

FILLS OUR HUMAN NEED
FOR INTIMACY
Image Credit: Shutterstock / OlVic
Even in a highly connected world, you can still be lonely. Whats the point of
thousands of contacts on social media if theyve all friendzoned you? If youre
short on admirers, perhaps youll find solace in an algorithm.
Following the incessant launch of life-like robots and chatbots which get you laid,
experts predict that human-robot marriages will become legal by 2050. We dont
even have to wait that long for technology to fill our intimate needs. Users are
already asking Siri to find them a boyfriend. Some are asking Amy a meeting-
scheduling virtual assistant for a date.

8. Love & dating 53
Desperate singles are using the Invisible Girlfriend or Invisible Boyfriend service
to create imaginary romantic partners to exchange sweet text messages with.
Youll even get help crafting a believable story behind how you met. Even
Googles AI engine is dropping its matter-of-fact style and reading 2,865 romance
novels to improve conversational allure and emotional engagement.
Image Credit: Invisible Girlfriend
All these lead to a disturbing conclusion: we may already be living in the

dystopian world of Her. Humans are having sex, falling in love, and tying the knot
with AI. Even if AI isnt the object of our affections, technology still plays a critical
role in fulfilling our primal need for intimacy. And, its only getting better.
ARTIFICIAL INTELLIGENCE BECOMES EMOTIONALLY AWARE

New Zealand based startup Soul Machines recently created Nadia, a visually
captivating and photorealistic chatbot that conveys and reads human emotions.
Nadia sees humans via webcam and determines their emotional state by
analyzing facial features in real-time.
Image Credit: Soul Machines

Nadia is backed by formidable talent. Shes designed by Mark Sagar, winner of
numerous Academy Awards for his work on facial motion capture techniques
in the films King Kong and Avatar, and voiced by highly acclaimed actress Cate
Blanchett.

8. Love & dating 54
While Nadias first real-world application is to assist patients with disabilities, you
can easily see how her emotionally charged service can easily be adapted to
solve the human longing for romantic relationships.
SEXBOTS BECOME REALISTIC AND INTELLIGENT

The technologies needed to design the perfect sexual partner attractive,
empathetic, human-like, and eager to please remain fragmented, but trends
indicate they are converging to meet peoples true desires. RealDoll plans to
integrate artificial intelligence and robotic features into their hyper realistic doll
products, with the aim of launching sexbots that rival the realism and sensuality
of West Worlds hosts by the end of 2017.
These sexbots to be marketed under the trade name Realbotix will be

able to communicate, talk dirty, demand foreplay, and even perform endearing,
nonsexual acts. In a 2015 interview with The New York Times, RealDoll CEO Matt
McMullen said these sexbots will establish emotional connections with their
users. The first release, named Harmony, is expected to debut in April according
to London-based Daily Star and will come with an app that allows users to
configure her personality.
Meanwhile, another AI-driven sexbot named Silicon Samantha has already hit the
shelves, with a set of advanced features such as realistic reactions to being kissed
and touched, a fully functioning G-spot, and even a kid-friendly family mode.
Image Credit: Design Taxi / Silicon Samantha

8. Love & dating 55
CHATBOTS HELP YOU FIND LOVE, DATES, AND SEX.

After gaining traction in a wide range of fields such as customer service, back
office jobs, health advice, and marketing, chatbots are being deployed to
address the fundamental human need for romantic companionship. From
rudimentary versions of fembots on Tinder to specialized services like Bubby
and Ghostbot, chatbots have become smarter at matching potential partners,
preventing sexual harassment, and inspiring romance.
Two Google Home devices positioned side by side even found themselves
getting attracted to each other. I love you around the universe to the stars and
back, said one to the other. You can watch the bizarre whole affair unfold on
Twitch.
Since we last wrote about chatbots for love and connection, several new bots
have come to market that are worth a look:
OutChat is a location-based dating bot for Facebook Messenger which enables

users to find prospective dates nearby faster and easier.
Bernie is an AI matchmaker who understands your type and matches you

up with the mathematically optimal potential partners.
Sensay is a bot that helps you find other humans who share the same
interests. While the irony of going through a bot to talk to a human is not lost
on us, beggars cant be choosers. If youre feeling a little lonely, theres no
shame getting a little help from your bot friends.
NearGroup is a bot that started out just like Sensay, connecting neighbors with
similar interests, but found its calling as a speed dating bot.
Foxsy ironically took the opposite path as NearGroup, starting out as a bot for
dating but pivoting to a bot for making friends. Great romances often start out
as friendships, so dont be discouraged from giving this bot a whirl.

8. Love & dating 56
YOU DONT NEED TWO TO TANGO

AI-powered bots can be both the object of your affection and a tool that helps
you find a soulmate. Either way, theyve irrevocably influenced the intimate and
romantic concerns of us humans. We either use AI to extend our humanity or
allow technology to blur the lines of what humanity means. Will love between
man and machine ever be socially acceptable? Only time will tell.
From rudimentary versions of fembots on

Tinder to specialized services like Bubby and
Ghostbot, chatbots have become smarter at
matching potential partners, preventing sexual
harassment, and inspiring romance.

Section 57
9.
Healthcare

9. Healthcare 58
THE OPPORTUNITIES
& CHALLENGES OF A.I. IN
HEALTHCARE
Image Credit: Shutterstock / Milles Studio
When we asked dozens of venture capitalists where they see the most potential
for applied artificial intelligence, they unanimously agreed on healthcare.
Technology has already been used to incrementally improve patient medical
records, care delivery, diagnostic accuracy, and drug development, but with A.I.
we could achieve exponential breakthroughs.

9. Healthcare 59
Deep learning first caught the medias attention when a team from the lab
of Geoffrey Hinton at the University of Toronto won a Merck drug discovery
competition despite having no experience with molecular biology and
pharmaceutical development. Recently, a multidisciplinary research team at
Stanfords School of Medicine comprised of pathologists, biomedical engineers,
geneticists, and computer scientists developed deep learning algorithms that
diagnose lung cancer more accurately than human pathologists.
The ultimate dream in healthcare is to eradicate disease entirely. This dream

might be possible one day with the assistance of AI, but we have a very very long
way to go.
INNOVATION IS CHALLENGED BY RISK-AVERSION

AND DIGITIZATION
Healthcare as a system advocates do no harm first and foremost. Not do
good, but do no harm. Every application of A.I. in healthcare is regulated by
that fundamental philosophy, cautions Kapila Ratnam, PhD, a scientist turned
partner at NewSpring Capital. Additionally, Lisa Suennen, Managing Director at
GE Ventures highlights that the single biggest contribution to excess cost and
error in healthcare is inertia. The attitude of this is how its always been done
is literally killing people.
Other investors agree that the ultra conservatism in the healthcare system,
while intended to protect patients, also harms them by restricting innovation.
Gavin Teo, Partner at B Capital Group and a specialist in digital health, cites
provider conservatism and unwillingness to risk new technology that does
not provide immediate fee-for-service (FFS) revenue as a major challenge for
startups tackling healthcare. Teo also points out that the industry feels burned
from recent experiences, such as electronic medical records (EMR) digitization
regulations, which were overhyped and resisted.

9. Healthcare 60
There are many well-known challenges to implementing machine learning and

A.I. in healthcare. The first is the lack of curated data sets, which are required
to train A.I. via surprised learning. Curated data sets that are robust and have
both the breadth and depth for training in a particular application are essential,
but frequently hard to access due to privacy concerns, record identification
concerns, and HIPAA, explains Dr. Robert Mittendorff of Norwest Venture
Partners.
Summerpal Kahlon, MD, is Director of Care Innovation at Oracle Health Sciences.

Hes seen many of these data challenges first hand in delivering technological
infrastructure to support individualized care. Adverse drug events cause around
770,000 injuries and deaths annually in the U.S. and cost each hospital up to
$5.6 million annually, Kahlon discloses, but drug data is messy, coming from
multiple sources in multiple formats. Additionally, genetic data in support of
pharmacogenomics is not available at scale yet.
Fixing accidental hospital infections and performing rare disease detection with
A.I. also requires better data than is currently available. According to Kahlon,
the genetic and behavioral data required for rare disease studies are not well-
defined nor easily captured while much of the information relating to the risk
factors for hospital-acquired infections is kept in unstructured notes in the chart,
including in flowsheets and clinical notes.
While data problems in healthcare abound, another major challenge is designing

technical solutions that can be smoothly implemented and integrated into
clinician practices and patient care. Behavioral change is the blockbuster drug
of digital health, claims Dr. Mittendorff, but changing habits is much easier said
than done. The wrong solution or rollout can even harm the healthcare industry.

9. Healthcare 61
Implementing and integrating technology has indeed been a burden for many
clinicians and practitioners. Dr. Jose I. Almeida is a pioneer in endovascular
venous surgery who has practiced for over 20 years. He adopted electronic
health records (EHR) ahead of the curve, yet has not seen many of the promised
benefits. We implemented our first EMR System eight years ago hoping it
would improve efficiencies. We are now on our fourth system, and remain
disappointed, complains Dr. Almeida. Right now, its been more of a hassle
than a time-saver, and has actually disrupted the doctor/patient relationship by
forcing a screen between physicians and their patients.
Leonard DAvolio, founder of Cyft, has harsh feedback for fellow entrepreneurs
trying to tackle the space: Were seeing hospital after hospital take incredible
loss and have widespread layoffs simply from the challenge of implementing
electronic health records. Imagine what happens if you then show up and say I
have artificial intelligence.
The healthcare industry is just getting its arms around capturing data digitally,
yet many healthcare tech entrepreneurs mistakenly believe that creating
a dashboard or dropping in a product will somehow lead to adoption of
technology and improve operations. Theres a huge misconception that A.I.
requires huge amounts of data, but thats not the real issue in healthcare. The
real issue is understanding the context into which you are trying to introduce
these technology, warns DAvolio. You need context and a deep understanding
of who will use this. What workflows will be introduced?
Even if a medical provider does successfully digitize their data, technical

carelessness can introduce problems for everyone in the system. According to
Ratnam of NewSpring, A credit card record costs about 10 cents on the black
market. A medical record costs about $200. Medical data is so valuable that
hackers constantly seek ways to break into provider or payment systems and
other repositories of medical data.

9. Healthcare 62
There is often tension between a venture-backed company, which aims for

fast growth, and the healthcare system which challenges scale because of
environmental complexity and unavoidable hand-holding.
This lesson has not been widely learned, observes DAvolio.
BUT OPPORTUNITIES ABOUND AND SOLUTIONS EXIST

Despite challenges, innovation in healthcare must continue. According to Teo
of B Capital, A study by the Association of American Medical Colleges estimates
that by 2025 there will be a shortfall of between 14,900 and 35,600 primary
care physicians. At the same time, the population is aging and in need of more
medical attention.
Thus, inaction and failure to innovate may lead to doing harm.
Luckily, many companies strive to address these issues before they come to
pass. CB Insights recently profiled 106 different artificial intelligence startups
in healthcare tackling the various challenges in the space, ranging from patient
monitoring to hospital operations.
Teo identifies A.I. powered chatbots and virtual assistants as one way to
alleviate supply constraints by widening the reach of video telehealth options.
In this case, diagnosis can be powered by machine learning and then trained by
artificial intelligence. Examples of companies providing clinician assistant and
care delivery services include Babylon Health, Evidation Health, Sensely, and
Senior Link.
Artificial intelligence can not only improve care delivery, but also assist in
clinician decision-making and operational efficiency, amplifying the impact of
each individual practitioner. AnalyticsMD employs AI and ML to streamline
hospital operations in emergency rooms, operating rooms, and in-patient wards,
while predictive companies like Cyft and HealthReveal analyze disparate data
sources to accurately triage and apply interventions to the highest risk patients.

9. Healthcare 63
A.I. not only helps physicians, but also patients. A study by the Mayo Clinic
determined that 50 percent of patients have difficulty with medication
adherence. Companies like AI Cure employ computer vision techniques to
enable smartphones to recognize faces and medications, lowering the cost and
improving the effectiveness of tracking and adherence programs. According to
Dr. Mittendorff, AI enabled coaching will allow a provider or coach to manage
more than 1,000 patients simultaneously rather than 50-100, a 10x increase in
labor leverage.
Finally, drug discovery companies like NuMedii and Kyan Therapeutics de-risk
the drug development process, enabling powerful and proprietary new
combination therapies, as well as individualized treatment with unprecedented
efficacy and safety, according to Teo. Otherwise, Suennen points out that the
general spend for each drug brought to market is $2.5 Billion.
Image Credit: CB Insights

9. Healthcare 64
Even technology challenges that come with digitizations can be mitigated by A.I.
Remember how valuable medical records are to hackers? Many of these records
are pilfered through social engineering methods, such as phishing or fraudulent
phone calls. Protenus is a healthcare security company which applies A.I. to
analyze enterprise-wide access logs and flag suspicious cases for administrator
review.
ALIGNING WITH POLICY & REVENUE IS KEY TO SUCCESS

The key to adoption of healthcare IT is to identify the correct point of entry and
fit these systems seamlessly into existing workflows. DAvolio of Cyft has spent
over 12 years fitting machine learning into the healthcare system, yet when
he speaks at conferences for clinicians, he avoids using the words artificial
intelligence or machine learning and instead focuses on real impact and
benefits.
Many patients with chronic diseases like diabetes visit doctors and hospitals
numerous times, costing themselves, insurance providers, and the medical
system a substantial amount of money. Cyft builds sophisticated models
that identify patients with a preventable re-admission and matches them to
appropriate intervention programs. Traditionally, these decisions are made by
looking at 7-10 administrative variables, but Cyfts models looks at over 400
data sources, ranging from free-text input from nurses to call center data. While
adoption of such technologies may seem complicated, DAvolio gets buy-in by
strategically aligning with revenue incentives and policy decisions.
In healthcare, policy eats strategy and culture for breakfast, explains DAvolio.
For example, prior to the American Recovery and Reinvestment Act passed in
2009 the rate of adoption of electronic health records was under 9%. Today,
thanks to the carrot and stick incentives involved in that act the rate of adoption
is > 90%. Another major policy shift that has dramatically helped investment in
healthcare IT are the value-based care experiments (also called demonstration
programs) funded by the Center for Medicare & Medicare Innovation (CMMI).

9. Healthcare 65
Knowing which policy an organization is incentivized or paid by is key to

identifying promising customers. According to DAvolio, organizations that
get paid mostly from seeing more patients will want AI that helps deliver more
complex care faster. Organizations that are paid via value-based programs will
seek technology that keep patients healthier at lower cost.
Suennen of GE Ventures agrees that operational analytics can dramatically

improve health systems. 25 percent of the more than $7 billion spent each
year on knee and hip surgeries are impacted by bundled payments initiatives.
Determining how to manage these bundles is challenging, and advanced
technologies can aid in understanding what changes must be made across the
board in operations and financial/clinical management to ensure that health
systems can respond.
Teo is also excited by policy changes that should drive forward healthcare
innovation. New reimbursement driven by the Medicare Access and CHIP
Reauthorization Act (MACRA) and the Merit-based Incentive Payment System
(MIPS) incentives in 2017 will drive quality outcomes, phasing providers to think
more holistically when investing in technology. Additionally, he believes that
a looser FDA in the coming years will help drive investment in personalized
medicine.
Successful healthcare innovation will only happen with strong collaboration

between entrepreneurs, investors, healthcare providers, patients and policy
developers. If the stars align, humanity stands to derive enormous benefit from
the application of A.I. and inch closer to our dream of perfect health and a world
without disease.

Section 66
10.
Gaming

10. Gaming 67
THE POINT OF PLAY: HOW

GAMES MAKE AI SMARTER
Image Credit: Shutterstock / Dragon Images
Do video games make you dumb? You can probably think of someone in your
life whod make you nod vigorously, but according to behavioral psychologists,
playing games actually does the opposite. Play performs a crucial role in
cognitive development and building behavior that helps advanced animals thrive
in different situations and environments.

10. Gaming 68
Turns out what helps kittens and toddlers learn can also help AI get smarter.
In March 2016, Googles Alpha Go thrashed 18-time human world champion
Lee Sedol, in a five-game match of Go. Alpha Gos neural networks played tens
of millions of matches against both humans and itself to master the complex
and unusual gameplay it used to pummel Lee. The achievement reveals how
machine learning works: an AI system progressively makes self-adjustments to
adapt to new challenges and become unbeatable.
Practice makes perfect, in gaming as in life. Algorithms bested the top humans at
Tic Tac Toe in the 1950s, checkers in 1994, chess in 1997, Jeopardy in 2011, and
finally Go in 2016. Right up to the victories, many considered these games to be
impossible for computers to master.
Just several months after scoring the surprise Go win, Deepmind the company
behind Alpha Go announced a partnership with game developer Blizzard
Entertainment to open their 20-year game series Starcraft to AI and machine
learning researchers around the world. Among the finest player-versus-player
(PVP) games in history, Starcraft with its variable gameplay approaches, wide
range of playable units, and rich in-game economic system serves as a worthy
learning environment to teach AI how to make strategic decisions and solve
complex, real-world challenges.
Chess and Go are games of perfect information, in which both sides are fully
aware of the entire game state. In Starcraft, you dont know where your enemy is
or what she is building. The game uses a fog-of-war layer to obscure information
from players both AI and human. This layer necessitates hierarchical planning
in addition to memory and responsive adaptation to new data in order to
win a match. In turn, these parameters are perfect for probing the limits of
reinforcement learning.

10. Gaming 69
Image Credit: Blizzard Entertainment
In Starcraft, you never know when an enemy might leap from the shadows.
BEYOND SCRIPTED NPCS AND BOTS

Gone are the days when gaming AI simply meant the scripted non-player
characters (NPCs) you meet in role-playing games (RPGs) or the crazy bots you
get to shoot when no one else is around to play against.
While a rudimentary form of programmatic scripting made up the entirety of

in-game AI in the past, much has changed. For example, the multi-awarded
and critically acclaimed Shadows of Mordor a role playing game based in
J.R.R. Tolkiens world of Middle Earth introduced the AI-driven Nemesis
System in 2014 with a reported upgrade coming in mid-2017. The machine
learning component adds a realistic layer of emotion, character and personality
development to in-game entities such as orcs, depending on how their series of
encounters with the human player pan out. Gaming companies like Electronic
Arts (EA) and Unity, indie developers, and even robotics startups are using
character AI to improve the realism of their experiences.

10. Gaming 70
In an interview with TOPBOTS, Patrick Soderlund, SVP at EA Worldwide

Studios emphasized his companys heavy investments in AI. He said, AI is
the strongest driver of technical innovation in the gaming industry and in the
larger field of software development. Soderlund, who oversees many of the
gaming behemoths popular releases such as Battlefield and Need For Speed,
described an ideal scenario where AI-powered characters in games make
independent decisions and anticipate what a human player will do based on her
playing patterns. To Soderlund, this requires AI to not only be smart, but also
emotionally intelligent.
In most games, the players can always find a way to cheat AI or figure out a
way to work around the AI. But with a self-learning and adaptive AI that actually
understands what you are doing, then youll have a real artificial intelligence, not
one that is merely scripted to deal with certain events.
What people tend to forget is the actual

player experience. Experience consists of so
many things: its animation, its graphics, etc.
Theres a lot artificial intelligence can do
but not until all of those things act well
together as an entity will the experience to
the player become relevant. The most important
thing is not the underlying AI system but the
consumer-touching experience.
- Patrick Soderlund, Senior Vice President, EA Worldwide Studios

10. Gaming 71
The use of AI in games has also gone beyond virtual opponents, NPCs and
creeps (the jargon game designers use to describe programmatically spawned
monsters). Soderlund reveals that AI can streamline realistic content creation,
personalize experiences, and make games run faster and easier to test. Youll
be able to talk to a video game wholl act smartly and come back to you with
context, sensitivity, and emotion. That will be pretty cool in terms of storytelling,
he promises.
With AI currently getting a tremendous level of attention, Soderlund calls for

caution, especially among over-enthusiastic game designers who think AI will
solve everything. What people tend to forget is the actual player experience.
Experience consists of so many things: its animation, its graphics, etc. Theres a
lot artificial intelligence can do but not until all of those things act well together
as an entity will the experience to the player become relevant. The most
important thing is not the underlying AI system but the consumer-touching
experience.
MASTERING THE REAL GAME OF LIFE

In a move similar to the Deepmind-Blizzard collaboration, Microsoft opened up
Minecraft for AI research. Dubbed Project Malmo and open to all AI designers
on GitHub, the initiative seeks to use a robust game environment in creating
systems that can augment human intelligence in the real world, able to
perform various tasks from cooking and driving a car to doing the laundry and
performing medical surgery.
Co-founded by business magnate and tech visionary Elon Musk, OpenAI

joined the gaming fray, widening the net and diversifying the experiences
AI researchers can immerse their agents in. The San Francisco-based AI lab
launched a collection of virtual worlds called Universe where AI can learn to play
games and use common software applications. The underlying goal is to develop
machines with human-like general intelligence, able to adapt, make decisions,
and take actions in a wide range of environments and scenarios.

10. Gaming 72
Allowing AI to experiment and fail in virtual environments may be much safer

than letting them loose in the world. The extraordinary realism that goes into
modern games is useful not only for appeasing picky gamers, but also for
training AI for use cases such as autonomous vehicles. A team from Intel and TU
Darmstadt claim the visual fidelity of popular games like Grand Theft Auto can
be used to establish ground truth or a baseline for reality for tuning vision
algorithms for self-driving cars. Lets just hope AI doesnt also pick up the less
civilized aspects of playing GTA.
Weve come a long way since Deep Blue defeated world chess champion Gary
Kasparov in 1997. IBM Interactive Media CTO George Dolbier aptly describes the
gigantic leap: the number of potential moves in a chess game has been equated
to the number of atoms, or the number of grains of sand on beaches. The
number of moves that are potential on a game of Go are equal to the number
of atoms in the universe. In a 2014, WIRED reported Go as one of AIs greatest
unsolved riddles that would require another 10 years to crack.
Given the unexpected advances of Go-playing algorithms, you can imagine

how AI with deep learning capabilities immersed in even more challenging
environments like Starcraft will eventually crack the game and beat any human
opponent.
But the stakes for emerging technology are much higher than merely winning
or losing games. If artificial general intelligence (AGI) can be achieved by pitting
algorithms against a multitude of virtual words, then what you have is a game-
changing revolution.

CRAVING MORE CONTENT?
HERES HOW YOU CAN WORK WITH US
TOPBOTS is a strategy & research firm focused on applied AI for enterprises.

Our customers include leading global companies such as LOreal, Paypal,
and WPP. We advise business leaders, executives, and practitioners on emerging
technology trends and help you successfully apply them in your organization.
To discuss how you can adopt automation technologies and AI at your own
To get your executive team up to speed on emerging technologies

and their impact on your industry, ask us about our corporate education
programs by emailing [email protected].
To raise your A.I. IQ, read our publication at TOPBOTS.com or subscribe
to our newsletter.
To order the full version of our Applied Artificial Intelligence book, visit
appliedAIbook.com.
Copyright 2017 TOPBOTS INC

Applied Artificial Intelligence: How Machine Learning Transforms How We Live and Work

Uploaded by

Copyright:

Available Formats

Applied Artificial Intelligence: How Machine Learning Transforms How We Live and Work

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Applied Artificial Intelligence: How Machine Learning Transforms How We Live and Work

Uploaded by

Copyright:

Available Formats

What are some examples of industries leveraging AI mentioned in the text?

What are some examples of industries leveraging AI mentioned in the text?

What are some goals of allowing AI to play games according to the text?

What are some goals of allowing AI to play games according to the text?

APPLIED

1. Industrials & manufacturing 5

2. Supply chain logistics 11

5. Marketing & Advertising 31

6. Beauty & retail 41

8. Love & dating 51

Applied Artificial How machine Learning transforms

From movie recommendations on Netflix, to partner suggestions

But machine intelligence isnt just for consumers. Industries ranging

Applied Artificial How machine Learning transforms

HOW YOU CAN WORK WITH US

Applied Artificial How machine Learning transforms

Applied Artificial How machine Learning transforms

Image Credit: Shutterstock / curraheeshutter

Applied Artificial How machine Learning transforms

Image Credit: Rethink Robotics

In 2015, General Electric launched GE Digital to drive software innovation across

When looking at the industrial internet,

Applied Artificial How machine Learning transforms

1. INDUSTRIAL DATA IS OFTEN INACCURATE

2. AI RUNS ON THE EDGE, NOT ON THE CLOUD

Industrial AI is built as an end-to-end system, described by Kodesh as a

Applied Artificial How machine Learning transforms

round-trip ticket, where data is generated by sensors on the edge, served

3. A SINGLE PREDICTION CAN COST OVER $1,000

Applied Artificial How machine Learning transforms

is to run thousands of algorithms at the same time. A consumer company like

4. COMPLEX MODELS MUST BE INTERPRETABLE

To achieve this high level of interpretability, GE needs to invent completely new

A consumer company like Amazon might make

Applied Artificial How machine Learning transforms

Applied Artificial How machine Learning transforms

Image Credit: Shutterstock / chuyuss

A complex series of operations connects suppliers to manufacturers to

Applied Artificial How machine Learning transforms

A breakdown in logistics breaks the supply chain, so companies constantly

Surprisingly for a transportation company, C.H. Robinson owns no vehicles.

HOW MUCH DOES A LOAD COST?

Applied Artificial How machine Learning transforms

WHERE ARE ALL THE TRUCKS?

Applied Artificial How machine Learning transforms

CAN WE EXPECT THE UNEXPECTED?

Predicting such disruptions and training AI to learn from contingency plans

HOW DO WE BUILD THE TECHNOLOGY?

Applied Artificial How machine Learning transforms

In addition to flexibility, owning data centers enables privacy and control. We

Image Credit: Shutterstock / MAGNIFIER

We are a cloud provider of transportation

Applied Artificial How machine Learning transforms

Technology is such a differentiating factor in our industry, Lindbloom

WHERE WILL WE GO FROM HERE?

Additionally, he cautions against overfocusing on AI and encourages executives

Applied Artificial How machine Learning transforms

Applied Artificial How machine Learning transforms

SMARTER AI MEANS SAFER,

Image Credit: Shutterstock / noolwlee