Anything you can do

… a computer can do just as well – at least, if you believe the
predictions, it soon will, thanks to advances in artificial intelligence. And
it could have a profound and surprising impact on the role of HR. Jane Lewis

AI – artificial intelligence as opposed to insemination – is suddenly back
as a talking point. In fact, it probably has not loomed so large in the public
consciousness since the days of Metal Mickey and K9. Even if you managed to
bypass the recent children’s craze for robotic pets (including a particularly
gruesome, "intelligent" giant scorpion as well as more
run-of-the-mill dogs and cats) you’re unlikely to escape the influence of this
autumn’s $90m blockbuster: Spielberg’s A.I: Artificial Intelligence.

Best described as a kind of futurist’s Pinocchio, AI recounts the adventures
of a robot child that has been programmed to love. What he wants most, of
course, is to become a proper human child. Equally predictably, the baddies of
the piece are also human – so paranoid about being usurped by more intelligent
"mechas" or robots that their only mission is to destroy him.

Preoccupations in fiction often reflect what is going on in real life, and
AI is no exception. The tech market may be in the grips of the worst downturn
it has ever seen but a substantial chunk of any spare capital still floating
around is now being channelled in this direction. The industry is building on
old foundations: the development of AI (aka "intelligent software
agents", "softbots" or "expert systems") was once a
hotbed of development talent, generally considered the next Big Thing in the
industry. But it got shoved on the back-burner following the sudden onslaught
of the communications revolution and the Internet.

Now a new generation of systems is evolving and futurologists claim we will
shortly witness a new wave of AI applications – this time heavily boosted by
the leap forward in communications technology. Intelligent systems will no
longer be standalone entities, so much as networked "agents", capable
of intelligently interacting with other agents and people.

The momentum is such that even Japan, a nation currently ground down by
economic misery, continues to lead the field in robot "pets", capable
of understanding simple verbal instructions, running households and monitoring
security. BT’s leading futurologist, Dr Ian Pearson thinks it likely that these
mobile "pets" will feature radio links to a smart computer elsewhere
in the house and that these devices will become common by about 2010.

Systems are certainly getting cleverer in leaps and bounds. We are
periodically startled by their growing prowess in different fields – most
notably, perhaps, in the ongoing battle of wits between Kasparov and IBM’s
chess wizard Deep Blue. Futurologists predict that computers will have matched
human intelligence by about 2015.

But it is clear that in some fields they are already steaming ahead. Earlier
this summer, developers at Hewlett-Packard’s Bristol research centre surprised
even themselves by creating a robot that outperformed six of the City’s best
stock-market experts in a head-to-head joust.

The most galling thing about the exercise is that they were not even really
trying. According to Dave Cliff, who devised the experiment, the
"bots" were programmed to be the simplest possible example of a robot
trader, with the smallest amount of intelligence. He was highly sceptical about
their chances of coming out on top. "I never planned for them to
outperform humans," he says.

In fact, it is becoming obvious that computers are leading the way
throughout the money markets as a whole – both increasing the volume of trade
and the speed with which transactions are undertaken. We may be languishing in
a bear market but figures show that the number of stocks and shares exchanged
on Nasdaq one supposedly sluggish Friday afternoon this July exceeded the daily
record of the boom two years ago.

In many ways this is down to a phenomenon of supply and demand that
environmental engineers call "induced traffic" – the more capacity
you build to ease congestion, the higher the overall rate of traffic becomes.
And the faster computers become, the more they rule out the possibility of
effective human competition. No human – not even a super-trader – can respond
quickly enough in an environment in which heavily traded stocks such as
Microsoft and Intel routinely change value several times a second. But a
computer, programmed to follow certain rules on risk and price, can clean up.

The human consequences of this are already apparent: once the programming
has been accomplished, the job is done. "A lot of what we’re doing is just
watching machines talk to other machines," says one trader. Two others
spent the afternoon desultorily tossing a football back and forth. Back at
Nasdaq’s computer nerve centre, meanwhile, a solitary technician sat back and
supervised the movement of millions.

If the outlook looks gloomy for City traders, the same is true across the
spectrum of white-collar work as the next phase of AI – the automation of
"mental" work – begins to kick in. Given that this is how 80 per cent
of us spend our gainful working hours, future prospects look interesting, to
say the least.

We have already seen the impact that non-intelligent systems – typically
those handling routine administrative tasks – has had on the workplace.
"We can administrate faster and more effectively using electronic tools
than we ever could using clerical intermediaries," says BT’s Pearson.

And the same is true of the entire product cycle. According to US think-tank
the World Future Society, the entire design and marketing cycle – idea,
invention, innovation and imitation – in companies is continuing to shrink
steadily. As late as the 1940s, the product cycle stretched to 30-40 years.
Today it seldom lasts 30-40 weeks.

If automating routine tasks alone can achieve such widespread change on the
corporate landscape, imagine the impact that systems designed to mimic human
intelligence – in all its myriad forms – will have. The clear implication, says
Pearson, is that large numbers of managerial and professional jobs are next in
the line of fire. And if your job is not actually eliminated completely, you
can bet that some of the "unique" skills and experience you bring to
it will be.

Who is likely to be affected first? According to the World Future Society,
by 2005 expert systems incorporating robotics, machine vision, voice
recognition, speech synthesis and electronic data processing will be common
across all industries, incorporating every sector from health services through
insurance underwriting, law enforcement, travel, energy prospecting and whatever
aspects of manufacturing and design that have not already been taken over by

But the greatest change of all will come in the structure of future
commercial institutions, which – thanks to AI – will bear very little
resemblance to their original forms, says Pearson. "We will move from
traditional companies with management and workers, to virtual companies with
core management and project teams only."

Smart technology will ensure that these new-look, sparsely populated
companies will be lithe and highly adaptive – capable of being re-jigged and
disbanded at the drop of a hat to take advantage of the emerging market niches
their super-sensitive expert systems have identified for them. The overall
effect will be to "generally disintermediate" the entire management
infrastructure of a company" says Pearson. Some companies won’t be needed
at all.

It will be an environment in which the élite will thrive – often working for
several companies at a time for high rewards. But the bulk of people will in
effect become commodity items, with a global supplier base.

Nonetheless, the democratising aspect of systems means there will also be
room for some plodders – at least in the early days. Smart systems mean that a
semi-educated clerk will be able to perform many of the functions previously
reserved for professionals. And because budding entrepreneurs will not have to
worry about all the administrative functions entailed in managing staff, it
will also be much easier to set up companies, says London Business School
research fellow, David Cannon. "AI means computers can learn very quickly
how you are and how you operate, and adopt systems to match," he says

Almost certainly AI will also lead to enormous boosts in productivity. In
fact, says Pearson, "It will go through the roof – so we will all be much
wealthier". But the question hanging in the air, especially if you are an
HR professional, is what on earth will happen to all those large numbers of
humans who will no longer be needed to operate the productive side of the
economy? More to the point, how do you decide which individuals and roles to
retain, and which to consign to the scrapheap?

Clearly in the first stages of the cycle, demand for engineers and
technicians – as well as for computer-literate managers – is likely to grow
exponentially. But it is easy to predict a scenario in the not-too-distant
future where even these "knowledge workers" in effect program them-
selves into redundancy by creating electronic Frankensteins capable of
programming, regenerating and improving themselves indefinitely.

In the long term, therefore, the only skills worth honing if you are
determined to stay in the workplace (rather than sloping off to enjoy yourself
on the back of the predicted AI-inspired productivity boom) will be those least
capable of being replicated by computer. "We’ll need a few ‘ideas
creators’ and a few ‘assimilators’ to package their ideas into useful and
desirable products," says Pearson airily. And it’s difficult to envisage a
scenario in which expert human marketers will not be necessary – until, that
is, computers learn to experience and convey emotion.

But the really irreplaceable jobs will be those majoring on intensely human,
interpersonal attributes which involve caring and empathy. In the barren landscape
of a grey electro-world, these kind of skills will be like gold dust, says
Pearson. And we will be prepared to pay a premium for them. "We will value
the human as a human, not as a cog in the machine," he says.

Machines will liberate people to cope with much higher value transactions,
aimed primarily at making life more pleasant. But we will pay through the nose
for such luxuries. According to one pundit, the time will come when dining in a
restaurant, on food hand-cooked and served by humans, will be the ultimate in
high living.

Pearson believes that as technology gradually automates much of the
intellectual side of work, people will find alternative channels for their
brain power and competitive instincts. One prediction is that community activity
will become a much more important facet of life as people try to impose some
order on all those long, leisure-strewn hours.

A possible result, he says, is that people may begin to form their sense of
identity from their place and relationships in the community rather than from
their job – a process, one might argue, that is already well underway in the
celebrity industry.

Taken to its logical conclusion, then, the shift into the AI era will
clearly have a profound impact on almost every aspect of society and the
economy – and on how our institutions are structured and run. Be prepared to
turn current ideas about the status quo upside down.

A good example Pearson cites is healthcare. At present the group at the top
of the pecking order – both in terms of status and reward – are those who can
boast the greatest expertise and experience, typically consultants and
surgeons. "But expert judgement and high precision accuracy will
eventually become redundant skills, because they will lie firmly in the machine
domain", he says.

So who will be the new high earners? Those currently at the bottom of the
pile – the nurses and carers whose intuitive actions and tender brow-moppings
cannot be machine-replicated. From this perspective, as Pearson points out,
"The current practice of sending nurses on degree courses is completely
the wrong direction to take."

The HR profession could be said to be equally short-sighted, as it tries to
shed its role as hand-holding people in favour of areas such as general
strategy and business management. But the ability to get the best out of people
will be huge in the new economy, says Cannon at London Business School.

A highly symbolic change will be the way people interact with each other, he
says. "When people in this highly automated, virtual world actually meet,
it will be something really big. That’s going to force greater pressure on

"There will be a new kind of networking – not about information
swapping so much as "whether I can trust you or whether we have some kind
of mutual benefit," he adds. "The pressure on leadership will change
to emphasise areas like this ability to inspire trust."

Pearson is the first to admit that the new world order he outlines –
particularly the elimination of all but a few people from most of the business
cycle – may seem difficult to get to grips with. But similarly strange things
have happened before. As he points out, "If you’d told someone in the
1920s that 80 per cent of the entire economy would be based on services by the
end of the century, they would not have believed you."

So what sort of timescale are we looking at? The replacement of the
information economy by the "care economy" will begin in earnest
between 2015 and 2020, predicts Pearson – just at the point when AI broadly
approaches the human equivalent. That should give you enough time to dust off
those lucrative tea-and-sympathy skills. Just make sure you don’t forget a
spare clean hanky.

Comments are closed.