Worrying about the potential downside of smart machines can blind us to the exciting possibilities they may open up for us in the future, argues Darl Kolb

By Darl Kolb*

Dire warnings about the threats posed by the rise of intelligent machines are nothing new. But their growing impact on our daily lives is giving such warnings a new urgency. In the face of creeping automation, the question everyone wants answered is: will robots take my job?

"If your job is routine, no matter how interesting and intellectually demanding, you are at risk of being replaced with an intelligent machine," says Darl Kolb, Professor of Connectivity in the Business School's Graduate School of Management.

"Planes have flown with autopilot for nearly 100 years, and the list of professions under threat is rapidly expanding. If you think a computer can do your job, it probably can – and should."

In any case, for Kolb that is not the point. Intelligent machines are intrinsically better than us at certain things: they land planes more consistently, they make more accurate medical diagnoses, they outperform humans at complex board games, and they would be unlikely to attempt to enter the Auckland property market.

But, rather than see them as a threat, he says, we should view them as an opportunity for us to figure out what makes us human, and what humans are good at.

High on that list, says Kolb, is collaboration.

"Most of us find maths difficult because our cognitive brains are less developed than our social brains. As one neuroscientist put it, our comparatively large brains were developed to solve complex problems – but not the problems of thermodynamics. They were social problems, such as who was in charge of the tribe, and how to build inter-tribal relationships."

Computers excel at providing answers, says Kolb, but humans are better at asking questions. So, rather than imagining a Darwinian future in which we are locked in competition with ever more powerful intelligent machines, perhaps we should think in terms of a partnership that plays to the strengths of each.

"In this scenario, we would not just preserve, but continue to grow and develop, our human character traits of curiosity, courage, and compassion," says Kolb.

Along with programming skills, education should focus on the attributes that make us more resilient, more resourceful, and more ready to make the hard calls when analysis has reached its limits, he says.

Of course, things could still go wrong if we choose the co-evolution path. For a start, social changes may not occur fast enough to keep up with technological ones, and massive economic dislocation may trigger a social revolution – in which case, says Kolb, all bets are off. At the opposite end of the scale, the winner-takes-all tendency in technology may see one or two companies come to dominate robotics globally, stifling innovation in the sector. Then, there is the likelihood of something going wrong due to the very nature of complex systems – sociologist Charles Perrow's notion of 'normal accidents'.

But the opportunity cost of not choosing co-evolution could be worse. For one thing, we would miss a chance to reap productivity gains. The McKinsey Global Institute has calculated the productivity gain from the steam engine at 0.3%, and information technology in the decade to 2005 at 0.6%. It puts the contribution of artificial intelligence, robotics, and machine learning over the next half century at 0.8% to 1.4%.

Then, too, there is the missed incentive to redesign work itself. Organisations have scarcely taken advantage of mobile technologies, says Kolb, and if we don't make jobs more human, machines will end up doing them, which will worsen unemployment.

He offers some pointers on the way forward:

• Don't make it an 'us versus them' debate. Humans and machines win or lose together and, more importantly, the issue quickly becomes politicised when job losses are mentioned.

• Don't take empathy, compassion, and creativity for granted. Even as we are working to make machines more human, we must foster what makes us unique.

• Expect more from machines (and the people who make them). Co-evolution requires us, as humans, to lift our game. But we shouldn't settle for software and machine intelligence that make us act like machines.

---------------------------------------

*Darl Kolb is Professor of Connectivity at the University of Auckland Business School.

This article was first published in the October 2017 issue of UABS Insights, and is reproduced with the permission of the University of Auckland Business School.

We welcome your help to improve our coverage of this issue. Any examples or experiences to relate? Any links to other news, data or research to shed more light on this? Any insight or views on what might happen next or what should happen next? Any errors to correct?

We welcome your comments below. If you are not already registered, please register to comment or click on the "Register" link below a comment.

Remember we welcome robust, respectful and insightful debate. We don't welcome abusive or defamatory comments and will de-register those repeatedly making such comments. Our current Comment policy is here.

30 Comments

If the machines give us a chance, may be we can evolve with them ?
Like we are already evolving with the Smartphone ?

Look at how computers changed work in recent times. Bookkeepers, gone, typing pools, gone, technical drawing offices, gone. Look at the most computer/AI ready jobs that involve large numbers of people. Call centres, etc resemble what has disappeared before.

The next round will be any repetitive jobs that can be done by a robot with minimal AI. There's going to be a lot of factory jobs in this category.

People will need reeducation to move into other jobs. So long as we provide it and don't follow the US example of student loans that financially destroy people while leaving them ready for no job.

In 1965 they started to worry about robots taking over - since then I spent a working life programming and have been retired a while. So after a lifetime of thinking:
(a) the experts usually get it wrong
(b) computers used to take the really simple jobs - 40 years ago I remember programming a stocktake that had been done manually - most stocktake sheets balanced with reality and the program remove the bother of adding them up and just left the interesting ones where there was no balance (mistakes in counting, mistakes in selling, breakages, etc). So my code destroyed boring work and left the interesting stuff. Now programs are doing the interesting things such as driving cars and recognising faces and translating languages. So all the interesting jobs will disappear leaving the jobs where people prefer person to person contact: hair-dressers, funerals, nursing.
Conclusion - don't send your kids to university, get them into jobs that require personality: politics, fashion, tourism.

Lapun, you may enjoy Gary Kasparov's recent book re AI, its role and its limitations. Deep Thinking:Where Machine Intelligence Ends and Human Creativity Begins.

Or professional jobs that actually require understanding and where decisions are very gray instead of black and white. Otherwise end up in charge of directing the AI/robots.

In the end the jobs just end up reallocated as you've pointed out and they are typically jobs that suit people more.

For some low IQ people there simply won't be a job. You can only reeducate people so much, especially when the the adult brain is slower to learn new tricks.

We already have this problem. Look at the benefits for those that reach 55 and just end up on a benefit with little or no prospects of getting another job. Some will not be able to adapt. It's one reason why I want our education system to change at high school level. It doesn't prepare people for the real world at all.

@keyo; It's not just low skilled jobs that will change. Doctor's and lawyers roles will also be effected but this should help to remove the mundane tasks that should make job roles more interesting and engaging.

So it's not all doom and gloom. :)

BBC; Will Your Children Have A Job?
http://www.bbc.co.uk/programmes/p04rq0px

There will always be a place for the low IQ people in our society , a special place for those needy simple souls ...

... just so long as robotics doesn't supplant politics ...

I wonder if working with AI will be much different to working with bureaucrats, or corporations, in that they don't come with morals.

For those that are interested two giant robots are going to fight (human piloted) and it starts in about 10 minutes.
https://www.twitch.tv/megabotsinc

Thanks, that was so much fun. The chainsaw was a little scary, I was almost sitting on the edge of my seat biting my fingernails.

Have you seen this Boston Dynamics robot?

https://www.youtube.com/watch?v=EP_NCB3KkiY

and the cheetah

https://www.youtube.com/watch?v=M8YjvHYbZ9w

The Boston Dynamics robots are quite interesting. I always found the speed at which the one in the first video moves to be impressive. Still a long way to go but at the same time they've advanced a lot.

"massive economic dislocation may trigger a social revolution – in which case, says Kolb, all bets are off"

Which is most likely.

Most jobs are completely non essential & designed for us to burn through as much fossil fuel / biodiversity as possible in the name of growth. Ultimately, the more you automate, the less need (demand) there is for the output of the automation.

AI and robots are just the latest in the long list of tools we have developed since hammerstones, spears and wheels. All tools allow us to achieve what we have imagined more efficiently, leaving us free to imagine more things to achieve. The tools themselves cannot do this as they have no imagination.

However, the vast majority of the human population whose current tasks are to create or achieve what the few have imagined will become redundant. The real question is how do we otherwise entertain the unimaginative dullards once they cost more to motivate than the cost of the tool to replace them. Then they will be truly worthless as they are both unproductive and unable to be targeted for profit having no assets.

Many groups around the world are now in this position, they are dispossessed, stateless and cannot earn enough to survive. They eke out an existence from rubbish dumps and foraging where land holders cannot be bothered removing them. That looks to me to be the future for many to follow them.

Does the list of unimaginative dullards include property investors?

They'll stick weapons on that robot a few comments up and use it to exterminate the extras. Or set the killer drones to 'civilian'. But look on the bright side, antibiotic resistant infections and catastrophic insect die-offs will probably get us first.

Jobs are a minor worry when human survival may be at stake. If we cannot co-evolve then our extinction becomes more probable unless we ensure AI has no incentive to improve itself while replicating, and is programmed to love us. However, predicting possible futures are so complex that even if there was such an entity as GOD, she, he, or it would only be guessing. Of course if God loves us she/he/it must step in and save us.
Science fiction has many scenarios on this subject. Take your pick. My own opinion for what it is worth is that we humans are on a course to self destruction. Ai may save us from ourselves.

and then there is the small problem of what actually powers the AI?
No industry = no AI

https://srsroccoreport.com/worlds-largest-oil-companies-deep-trouble-as-...

AI will continue to fascinate people as long as they benefit from it. If it only means some people will get very rich and the most very poor then there will be a revolution. As money no longer relates to productivity then I suggest that as long as people receive enough money to live well we won’t have a revolution. You only have to look at NZ’s Universal superannuation to see that an UBI is the answer.

Man, this sort of stuff boils my blood. Here we are, talking about "evolving" with bloody machines for crying out loud, all the while as we are causing the sixth great extinction. I would rather we reconnected with nature than embark on this utter idiocy. We will eventually be left alone with bits of tin and plastic, with no real company of other species. We separate ourselves from the rest of the world at our peril, but tell me, how many of today's young kids will ever get to ride a horse. Seriously sad.

It is where I would differ from the author, we won't see the level of complexity he envisions because the answer to a shortage of resources is not more complexity.

Jobs? Who cares, there are already five (or more) times as many people as could live a modern lifestyle sustainably, and most of them already don't do anything useful. Mankind is a plague upon the earth, and when there's nothing left alive but us and rats and cockroaches, and nothing left to destroy but us, we will. Thank goodness there are probably myriad other worlds out there. I am a cheery chap!

In order for both sides to grow you would need to add ethics to AI. There's a very interesting radio article from the BBC on this subject that discuss everything from driverless cars to autonomous weapons.

Certainly worth a listen; BBC The Inquiry: Can We Teach Robots Ethics?
http://www.bbc.co.uk/programmes/w3csv1c2

I cannot wait to replace the Government, Banking and other wastrels with simple computers, programmed to never lie, never skim off the top, only one bank, an honest one, no opposition needed as no over paid, over indulged people printing and leveraging people to death and beyond. And replacing them every 3 years with more of the same dorks.

No need for Land Banking, no need for building a ladder to climb. All done by robots in factory, not in house.

Cos all we need is a simple fact programmed into the memory Banks....live within ones means...not someone else they conjured up from here to eternity. with flippers and agents adding to the plot. 4% percent a flipping time.

Computers are fine, no need to have any overheads, that add to the cost of everything, plus GST, of course..

Just a little oil, for the wheels of commerce...and of course free contraceptives, so we may never screw it all up....again.

The human race could die out...happy as Microsoft...and an Apple a Day would allow.

End of mankind.....how good for the Planet....and not a plastic bag in site. Just a memory chip, off the old block. Self destruct...last one out, close the door....make a good movie too....don't ya thunk.

Just dreaming....Just idling, used to work in computers......never been put to good use....YET.

The AI developed by good folks will have it's risks but the AI developed by desperate governments and terrorists will certainly involve some products that are not limited to preserving human life.

In the wider interests of humanity a reasonable person could develop an ethical AI product that would limit world population to 1 billion humans.

Ooops - 1 billion human persons.

I'll command my AI army to be benevolent after they've destroyed your AL army and subdued you.