Are machines really intelligent, or are they as dumb as a toaster?


Exclusive interview with Gerd LeonhardFuturistHumanistAuthor of “Technology vs. Humanity and CEO of the Futures Agency

We all see how technology is impacting society, business, and service divisions.  And how exponential it is growing.  But what should we really look out for in the near future?  Are machines really intelligent, or are they as dumb as a toaster?  Are 3D printed steaks any good?  Is AI life-threatening, or just job-threatening? Read on more as I chat with Gerd Leonhard, Service Mastery Day host at this year’s 12th Aftermarket Business Platform in Berlin, October 17-19.

Thomas Igou (Content Director at Aftermarket): Gerd, what’s the biggest trend impacting manufacturing organizations today, from a Service perspective?

G.L: Data and intelligence is the biggest trend happening right now.  Every process is completely tracked and monitored, made intelligent.  As Kevin Kelley likes to say, “first we digitize, now we cognify.”

The problem is that there is too much data but not enough precise information in aftersales, no linking of online and offline.  Intelligent machine (misnomer, intelligent, not smart), can look at a trillion facts of data and give you on-demand answer in 14 seconds.  Humans can’t do that; for huge amounts of disparate and unstructured data, you need artificial intelligence to structure this.  And it’s important to remember that these machines are not thinking machines, they are giant neuro applications like super software, and that is what AI is bringing to us, and I sometimes call it IA, Intelligent Assistants rather than AI.

It creates a huge change.  You have to train these machines, put all the good data in; only then can you get machines to create a huge amount of value by recognizing patterns and suggesting other ways of doing things.

A problem that will arise is that it’ll be widely used because AI is superior to humans in analyzing data.  This will create the black box problem: you don’t know why a machine will decide this because it’s beyond human understanding, you have to trust the machine or not.  But I think it’s important to keep humans in the loop to question the machines because they will not understand human components and reasoning; it’s important to have a counter balance.  It’s a bit like the trip advisor problem: it can be really great, but you still need to check it against reality.

When you talk about aftersales, a lot of this is manual work shoveling data and facts, and that is going to dramatically change.  This is like the doctor looking at IBM Watson who must scan a million MRIs to figure out if there’s cancer; the doctor can just push a button and ask Watson to take a look at this picture, what do you think?  And 14 seconds later he gets an analysis of millions of facts.  That’s going to be the same in Service.

Also, the cloud means everything is becoming virtual, paper is on its way out, everything is connected.  Of course, that leads to another problem: security and surveillance.  When everything is in the cloud, it becomes attackable.  And that becomes an issue, not only in terms of technology but also regulation like GDPR, keeping data safe, that is quite an effort.  You get the benefit of connectivity but also the responsibility.

One of the key challenge, when you’re in technology, is that you make ethical decisions all the time.  In 10 years from now, we’ll be in a world where machines will be able to do things unthinkable for us, then we’ll have to decide what are the guidelines, what’s fair use, what is the authority of users, do I have permission… if something goes wrong, it could kill the bank, it will be much bigger than what’s happening with Facebook right now

You have amazing technology, but you need to have the right guidelines and processes in place.  Accountability is a big deal.

We used to have wars on oils, now we’ll have wars on information.

If you look at the food chain complex, once you’ve sold something, who owns the data and who gets to do what?  You need to be hyper collaborative, not compete, because in the end it won’t be on ownership of the data because we are moving to a world of APIs where everything is connected, it will be about the license to operate, to use the data, not locking it up.  For example, in the healthcare system, it will be about the health cloud which puts all people’s data in the cloud, so the question is what kind of permission do I have to read? 

The data we’re going to get will be mindboggling; it will be bigger than oil.

T.I:  Should we fear the rise of the machines?

G.L:  We’re far away from machine becoming conscious or sentient, for machines to have this capability they would need to go way beyond the current binary operations.  One of the biggest problem is that machines will become so narrowly intelligent, we’ll give them too much power because we think they are so intelligent; we’ll give them too much authority in terms of analytics.

The biggest issue in the foreseeable future is displacement of work.  Any work that is automatable will be lost.  Bu that’s not life threatening, it’s just threatening our jobs!

To change our jobs, we need to move up the food chain to valuable things humans do, like understanding, negotiations, planning, innovation.

We’re 30-50 years to any potential existential threat of AI, where AI can supplement or surpass humans.  The most important thing will be regulation and safety. It’s like nuclear power, if you’re not doing it safely, it’s very bad.

Elon Musk and Stephen Hawkins are on the right track, if we’re going to have intelligent machines like we are, we’ll need to contain them.  But it’s not an issue today, these machines are as dumb as a toaster, compared to human brain and human thinking, because they don’t “think” at all. 

It’s a danger, but not next week.

T.I:  What are your thoughts on Blockchain?

G.L:  Blockchain is amazing but primarily a super fancy database that replaces central control and the cost of the central control.  It’ll be a big deal in transportation, logistics and smart contracts by creating a cheap way to do agreements and exchange of information.

In financial services, it might work for low end transactions for peer to peer services, but on a large scale it won’t be accepted by governments. Money is a public matter, and governments need to hold on to the accountability for money, because someone needs to be there in the end to cover it.  So Blockchain is huge but I don’t see it happening in banking anytime soon, except maybe for swift and back end systems of credit cards transaction.


T.I:  We talked about work displacement.  What can humans do about it?

G.L:  Computers don’t do relationships, they don’t do engagements, they don’t do soft factors.  If they could, they would in any case be fake.  Computers are for facts.  I think it’s Kevin Kelly who said, “computers are for answers, humans are for questions”.  So, it’ll be perfectly fine, if a snowstorm hits an airport and flights are cancelled and 40 000 people need to rebook their flights: that should be a chatbot because AI can do rebook all 40 000 people in 40 seconds.

There’ll be jobs machines can do, and others that only humans can do and that we shouldn’t touch.  We shouldn’t let a machine decide if a couple should have a baby or not.  The machine can give feedback, just like they do now, but we shouldn’t always make efficiency the top priority.

In some cases, efficiency is so important that nothing else really matters, like rebooking … or driving.  There’s no moral concerns to driving the right way.

Other areas where efficiency is not important, like dating, or hiring and firing, human resources, where a computer will not understand what you’re trying to say with relationships and engagements.  They may eventually be able to understand what we’re trying to say but won’t be in a position to feel it.

We think too much about all the great things machines can do and take away our own skills.  It’s very important in customer service.  It all comes down to engagement and trust, not efficiency.  Peter Drucker said that strategy eats technology for breakfast, I like to say that culture eats technology for breakfast.

T.I:  What about 3D Printing?

G.L:  Prototyping is everywhere.  What’s happening is we need more composite materials and more advances in material sciences to print complicated structures.  We can print fake wood, but not wood.  We can print fake earlobe, but not an earlobe. 

We’re going to make major breakthroughs so that in the next couple of years, an industrial printer can print 150 different composite materials and new kinds of plastics.  3D printed parts will be one tenth of the cost and all you need will be the license to print the authentic piece.  That will hugely change shipping, manufacturing, customer service…  I put a timeframe of 10 years, when you will have giant 3D printers driving around in trucks, driving from one place to the next to print on-demand. 

Like airports, instead of having an airplane transporting goods, you’ll have 3D printing facilities that’ll be highly secured and licensed with the right products, and the costs will be like spotify, 90% less.  That will happen for sure in many commodity jobs for printing like in construction, you will be able to print the beams and some of the structures on ships, you can print that on location.

There are many things we can’t quite print yet, but for example food, you’ll be able to print that soon.  Lab growing meat already exists, it’s completely organic, it’s actual meat, you can take that substance and print from it. 

People shouldn’t underestimate 3D Printing because it hasn’t really worked or is too expensive.  Right now, such a 3D Printer will cost 2 million €, it will be 100 000 in less than 10 years. It will be everywhere.  If we monitor all progress of science and nanomaterials, there are breakthroughs every day.