AI has the potential to significantly impact jobs and the economy. Many roles could be automated, displacing millions of workers worldwide over the coming decades. However, AI can also boost productivity and free up time for higher-value work. The document discusses how AI is being applied in human resources to simplify processes, provide insights from large datasets, and help align HR and business strategies to drive growth. Concerns around bias, lack of emotions, and costs of developing AI systems are also addressed.
2. BusinessCase
Commercial Context
CHANGING TIMES. CHANGING METHODS
Inequality has grown in most advanced economies. Some
800 million jobs worldwide are thought to be at risk and in
the US half of all roles could be displaced inside the next 50
years. AI is largely abstract & experimental but as business
users find it easier to code & program (e.g. Ludwig Python,
Azure ML drag & drop) – this could happen much sooner
Autonomous driving vehicles could impact 3% of American’s
who rely on driving trucks or taxis for a living. Pizza Hut
automation technology can make 360 pizza’s an hour
generating £47K in revenue in that time. And to meet UK
annual consumption it would take just 390 bots doing away
with more than 35,000 pizza production jobs (IBIS DATA)
And it’s not just blue-collar roles that AI can do. They’ve
crept into being tax, legal advisors and secretaries. In
radiology AI computes 10,000 times faster than an average
human - 2 days of scanning and prognosis are done in less
than a minute. Diagnosis could be done much quickly to
positively impact health. With the cost of technology falling it
could go some way to fixing ‘third of US healthcare costs are
wasted’ (Don Berwick) & more affordable healthcare
There’s been lots of talk about machines taking
over jobs. Some have even pointed to a job’s
apocalypse not dissimilar to the what the film the
“Matrix’ portrays with people living underground,
and struggling to survive above ground
3. Brains process 16 million bits of information a
second. Ears distinguish between hundreds of
thousands of sounds. Eyes distinguish between 10
million different colours. A hundred pain sensors on
every cm² of skin tell us when things are wrong.
Pressure, fatigue & emotions make poor decisions
much more likely. Patience, time & planning help
avoid. Choice blindness is when the thoughts are
distorted through fake news or bias delimiting the
right decisions. AI helps in all these instances
THE HUMAN
& BE THE NORM
03
FOR GROWTH
02
AI’S POTENTIAL
01 "Ethical AI is a win-win
proposition that can become a
competitive advantage for
Europe: being a leader of human-
centric AI that people can trust"
EU Commission Vice President
- ANDRUS ANSIP
4. AI DOES MUCH MORE WITH MUCH LESS
McKinsey suggests AI could boost productivity 10% by 2030 and
improve future generation living standards. McKinsey finds that just
under a third of activities in 60% of all roles could be automated freeing
up time for higher value work focus and outcomes
AI’s capabilities in data analytics and surfacing improvement helps
build lean agile businesses. Technology has grown at five times the rate
of management methods. Growth in engagement and productivity have
been “sluggish” (Gallup). AI could help to close this gap, improve on
capacity & capabilities, discover more about how things are done,
analyze the 99% of worlds data that hasn’t been touched and how to
make improvements
Managerial decision making could also be better and a third of
employees think robots would make better decisions than their
manager. 80% of job leavers tend to do so because of conflicts or
disdain with their line managers rather than the company. 60% of UK
employees suggest they have a workplace nemesis (Totaljobs)
AI could help see if specific managers or employees are sacrificing
company values over personal biases or prejudices. It could pinpoint
territorial protectionism or political silos. More than 80% of companies
don’t do exit interviews – AI helps cut flight risks and establish a
retention safety net where people fail to intervene. A problem shared is
a problem halved after all and having a vent for frustrations is better
than bottling up concerns to see them explode out of proportion
“Managers today have to
do more with less, and get
better results from
limited resources, more
than ever before”
BRIAN TRACY
5. AI ALIGNS & ACCELERATES GROWTH
AI enabled ubiquitous computing helps build more frictionless
operating environments cutting structural & operating complexity.
Doing that saves 10% in annual earnings (Global Simplicity Index)
Employees and customers want simpler experiences. Customers are
even prepared to pay more. AI helps enhance features and shape
product development strategies as well as simplifying HR processes
60% of CEO’s are rethinking their HR function (PWC). 94% of HR leaders
feel that ‘agility and collaboration’ are critical to success (Deloitte). But
Accenture finds less than a quarter had an HR target model that
responds to changes externally and only a fifth say it’s aligned with
specific growth initiatives
Measuring more across company value chains and linking that to roles
would improve effectiveness & efficiency. Customers and employees
are the two most important parts of a business.
At last years CIPD annual event a quarter of HR delegates wanted
simplified HR processes. A fifth said they’d like a sophisticated all-in-
one solution which AI could comfortably deliver on
A couple of years ago Gartner reported ‘7% of top-performing
companies ranked AI & ML as game-changing’ – just a year later that
figure went to 40%. The results are astounding and undeniable
“Growth begins when
we begin to accept our
own weakness”
JEAN VANIER
“Growth and comfort
do not coexist”
GINNI ROMETTY
6. ECONOMIES RACE FOR #1POSITION
Europe lags behind the US and China in AI development. More
precarious employment, living standard uncertainty and polarization
could easily lead to civil unrest. Some 22 million people will be chasing
less skilled UK roles in the next two years – most of which could easily
be taken over by robots, automation and AI
Governments are looking for ways to keep societies civil. In the last fifty
years, $5 trillion was spent on development aid. AI could fuel $13 trillion
in new economic activity by 2030 (McKinsey)
AI could become a platform for giving citizens a universal basic income
UBI to support the projected 76 million UK inhabitants in 2045. And, when
some 46% will be looking to retire in the following year
Stress is the biggest culprit of ill health and Governments often have to
pick up the bill later. At 65 most are likely to have one illness costing 2.5
times that of a 30-year-old; at 75 they’re likely to have two illnesses, and
85-year-old costs 5 times that of 30-year-old
In the workplace ageing populations and greater reliance on older
workers to stay healthier for longer has ramifications on workforce
planning and recruiting. The average worker is likely to change jobs ten
times before they reach 40
“Creating an economy that
harnesses AI and big data is
one of the great
opportunities of our age”
Secretary of State for Business,
Energy and Industrial Strategy,
GREG CLARK
7. CHANGES WILL FUEL ADOPTION
British workers averaged 42 hours in a working week (2 hours more
than the EU average) working an extra two and half weeks a year.
Danish employees work four hours less a week but are 23% more
productive
One in six employees are subject to stress, anxiety and emotional fears.
Women are twice as likely to have common mental health issues at
work. These negatively impact engagement and productivity, lower
confidence, communication skills and fuel mental blocks. 44% know a
colleague unable to work due to stress (Capita Employee Benefit’s
survey in 2017)
Companies that promote health and wellbeing are 3.5 times more likely
to have more innovative and creative employees (EU figures). CEB
reported an 80% growth in employee workloads and 42% of
accomplishments go 'unnoticed'. Besides communication is only 30%
words said and “they may forget what you said but not how you made
them feel” is important
Having a system that’s embedded helps alleviate unsubstantiated fears,
makes for more inclusive and engaging workplaces and installs a
process of recognition that goes onto grow confidence and trust in
leaders. AI could easily help serve as a vent for employee frustrations,
pinpoint wellbeing challenges, cut non-sensical conflicts, merit-based
rewarding and sustainable resolutions
“Jobs are no longer secure…
because of downsizing and the
changing of the psychological
contract that took place during
the recession”
CARY COOPER, Professor of
Organisational Psychology & Health MBS
8.
9. THE PROS & CONS OF USING AI
There are pros and cons of using AI at work. Some are clear and
compelling. Others are less clear with hidden costs. In HR and
marketing areas the greatest ones are in computation and making
sense of fast-changing and vast volumes of data. Deep learning helps
to come up with the right decisions use multidimensional inputs and
respond to triggers like a customer visiting a website in a timely and
responsive fashion
AI, automation and bots are non-emotional and therefore able to see
through bias. That said bias programmed into AI will only output
similar bias. But deep learning and correction can help alleviate this
In contrast, people can find emotional luggage, fatigue and emotions
can get the better of them and supersede decisions that may be right
for the business. AI systems able to work around the clock without
regular pay but do need people to help reprogram, train and correct
algorithms which often rely on technical scarce skills at a premium. It
takes people to ask questions, scrutinize the systems response and
fine tune the computation engine
AI for decision making is proactive in that they process and mine into
data in real time. Three-quarters of CEO time is spent on interacting
with teams, strategy design and improving customer experience. AI
deep learning means leaders and CEO’s are able to get answers to
questions early. This can mean less wastage, clutter & complexity
Having this capability does rely on people investing time and
resources into the AI build up front. This means having well-
coordinated expensive multidisciplined project management skills
High failure rates in change programs mean it could take quite a few
iterations to come up with the right system and experimentation could
mean costs mount up, too many starts and stops, evaporating project
appetites and delivery momentum
10. COPYRIGHT 2019. STRING BUSINESS LIMITED
UNIT 34 | 67-68 HATTON GARDEN | LONDON EC2A 4NE | +44 (0) 203 879 4628
12. BusinessCase
Commercial Context
THE TWO MOST IMPORTANT PARTS
The two most important parts are customers & employees.
It’s often been said that “employees are just as important
as customers” but management varies considerably.
People are only capable of personally managing some 100
relationships. At company level, HRMS plays a critical role
in managing much greater numbers
In customer areas measuring & monitoring the quality of
relationships is much more advanced. In HR it’s often just
about administering databases – logging requests and
processing them with a few often overstretched people.
Innovative HR has so much more to do with culture, work
processes, practices and engagement
Best case practices are easier to see in unified systems
with frictionless settings with automation giving timely
answers and resolutions. Doing it this way simplifies
management, cuts friction and payoffs are handsome
saving at least 10% in earnings each year. Recruiting is also
easier and cheaper when processes are simplified and the
value proposition more compelling than competitors
A couple of years ago Gartner reported ‘7% of top-
performing companies ranked AI & ML as game-
changing’ – just a year later that figure went to 40%.
The results are astounding and undeniable
13. Three areas of AI when used in human resources
Reporting is to do with getting hold of contextual data and
getting it into a format for computation and ultimately
learn from people to predict what could happen
These are areas that fall into ‘machine learning’
The intersection is where ‘deep learning’ sits and the
place where people teach or train AI in how make the best
possible decisions given their personal experiences
WHAT’S MACHINE LEARNING & DEEP
LEARNING IN AI BASED HR
14. CLEAR PURPOSEFUL GOALS SIMPLIFY
AlphaGo took 11 years less than it was expected to do so at mastering
GO the board game. To play it - the goal is to win territory and there are
infinitely more moves to play than we are confronted with in improving
on customer or employee relationships
In management the goal is clear and, more often than not, to grow
profitability – by cutting costs and growing revenues. Translating this
into a vision that’s easy to understand helps define the starting blocks,
pitch terrain and where the goal posts are
The vision behind our platform works across every profit generating
company and is one of fostering an environment that grows client
experience, satisfaction and spending. In employee management
similarly to grow experience, retention, engagement and performance
in bespoke terms. And without too much experimentation because
people dislike personal changes. And to build on human capital by
developing a new class of managers and leaders with exceptional
people & soft skills. Doing that well could put £150 billion back into the
UK economy each year
“92% weren’t organized to
succeed and only 14%
knew how to do that”
DELOITTE HUMAN CAPITAL
TRENDS 2016
15. GRADUAL MOVES IN THE RIGHT DIRECTION
Employees and customers want simpler experiences. Customers are
even prepared to pay more. But low structural and operating
complexity calls for a process and operating model that delivers on the
vision. And, one that people can relate to easily strategically and
operationally. This goes a long way to helping the 80% of companies
don’t know what data is needed for AI (IBM 2018)
In 2016 IDC found that HR leaders were less concerned with metrics are
more focused on making management process simpler. 60% of CEO’s
were also rethinking their HR function (PWC) with ‘agility and
collaboration’ deemed critical (Deloitte)
But Accenture found less than a quarter had an HR target model that
responded to changes externally and only a fifth found it was aligned to
specific growth initiatives. At last years CIPD annual event a quarter of
HR delegates wanted simplified HR processes. A fifth said they’d like a
sophisticated all-in-one solution and companies appear to have gone
back to the drawing board to decide on what really matters and glove fit
processes around that
Half of internet traffic is over mobile devices but fewer than 20% of HR
management systems to work off mobile. A shortage of skills, a billion
job searches a month over mobile devices and a 40% abandonment rate
on non-mobile devices means it pays to have a robust digital HR
strategy that works across multiple devices and channels
“Progress lies not in enhancing
what is, but in advancing
toward what will best”
KHALIL GIBRAN
"Predicting rain doesn't count.
Building arks does"
WARREN BUFFETT
16. FUTURE PROOF, MERIT BASED & UNIVERSAL
A good vision works for many years and deciding on a target model that
supports it helps make managing future proof
Its often been said that in HR ‘one size doesn’t fit all’ but we’d argue that if
the vision is clear, purposeful and applies to all – why can a central
model not be used across the whole business applying to all?
Getting this right also means HR isn’t rudimentary, is standardized,
inclusive and with least reporting burdens. At a globally dispersed
organization shared services also works much better. Deciding on the
things that matter the most and measuring against those simply helps
users grow sophistication bottom up
AI has feature rich computational capabilities and able to track in
multidimensional terms. It beats studying and trying to make sense of
spreadsheets and tables. The bigger the datasets the harder people find
working with using standard applications like excel. Analyzing the
wealth of behavioural and transactional data is easy for AI and
multidimensional visualization helps users intuitively understand
what’s needed and improve on discovery process in concept clustering,
social network analysis, engagement and sentiment analysis. These
help foster a culture of higher performance
“The key is not the will to
win... everybody has that. It
is the will to prepare to win
that is important”
BOBBY KNIGHT
17. FAIR, INCLUSIVE & ENTERPRISE WIDE
Digital AI based HR needs a model and one that gives direction aka
‘delineated model’. A clear ‘win-win’ scenario helps navigate. Analyzing
and exploring data to identify trends and issues has more precise
contexts. Uptake and effectiveness will be impacted by things like
company culture, inertia for legacy systems
Monitoring progress helps show what worked best and that discovery
process helps formulate best practices HR doesn’t need lots of data just
enough to show how things are, where they could be and seeing things
on an enterprise scale
Establishing this process as a closed loop system means AI sequencing
and training is much easier. Knowing the anatomy of this process makes
it easier to know what data types to work with – textual or numerical
Tracking what worked in past interventions forms part of the training
algorithm and prediction engine. Using segments or some other
classification mechanism helps to manage large amounts of data. It also
makes the journey between data, knowledge and wisdom a less arduous
one. Greater involvement from business leaders and autonomous
machine learning would go some way to mitigating digital skills
shortages and opportunity losses of $90+ billion (IDC)
“Inclusion and fairness in the
workplace is not simply the
right thing to do; it's the
smart thing to do”
ALEXIS HARMAN
18. Segmentation is an established tool in marketing but only a few
companies like BP make use of it for employee engagement and
management. Just like no two customers are the same - no two
employees are either
With five generations at work, today complexity can creep in easily.
Bain & Co found that segmentation at a telecommunications and utility
grew performance, productivity and cut costs considerably. Combine
this with methodical employee experience management, at scale and
agility – success is much more likely
It also helps to grow psychological security and going some way to
alleviating unsubstantiated fears given a quarter of employees don't
feel confident enough that they perform well in their roles.
Confidence helps cut conflicts and improve on soft skills. This is
valuable in that $5 out of every $6 is not explained on the balance
sheet. Slow productivity growth means Millennials expect to have a
lower living standard than their grandparents
Deciding on what data is needed, how algorithms deal with it and what
applications support the data life cycle are critical. If HR is a means to
an end then performance is a good lens. It’s also an approach that
Costco found made biggest difference in growing income per store by
$100K each year. The diagram on the next page looks at best practice
elements of performance management
Its broken down into different sections. The first looking at four broad
components of measuring and managing performance – purpose,
process, systems and reporting
The larger or more dispersed or varied the workforce the greater the
relevance of scalability. The flower describes what’s important there
Four areas of strategic planning consideration that help articulate
what’s needed in operational terms. The things that matter the most to
employees from an engagement perspective
Beneath it we look at the shape and characteristics needed in data –
goal led, responsible decision making, data quality and goal relevance
19. COPYRIGHT 2019. STRING BUSINESS LIMITED
UNIT 34 | 67-68 HATTON GARDEN | LONDON EC2A 4NE | +44 (0) 203 879 4628