Lloyds of London report – how self driving cars will affect the insurance industry

Red flag act

When the first steam powered cars (or locomotives as they were then called) first took to the roads in the early 1860’s it was feared that engines and their trailers might endanger the safety of the public, cause fatal accidents, block narrow lanes, and disturb the locals by operating at night. Many complained the cars were scaring the horses meanwhile the farmers were shooting the cars!

To allay the concerns, restrictions and speed limits were imposed by the Locomotive Act of 1865 (more popularly known as the “Red Flag Act“) which required all road locomotives, which included automobiles, to travel at a maximum of 4 mph (6.4 km/h) in the country and 2 mph (3.2 km/h) in the city – as well as requiring a man carrying a red flag or lantern to walk in front of road vehicles. The Act also included such matters as “Damage caused by locomotives to bridges to be made good by the vehicle owners.” However it was soon discovered that the steam carriages’ brakes and their wide tyres caused less damage to the roads than horse-drawn carriages because of the absence of horses’ hooves striking the road and wheels which did not lock and drag as they did on horse drawn carriages.

In the latter part of the 1800’s the British Motor Syndicate began a public relations campaign to lobby for the repeal of the Locomotive Act, which was declared the main obstacle to wider adoption of the car in Britain. The Act of 1865 was repealed in 1896.

Almost 120 years later we stand at the precipice of a new dawn in the age of the automobile – self-driving cars. Everything, from how we move goods to how we move ourselves around, is ripe for change.

Self-driving cars will require legislators to make new amendments to the current Highways Acts in place. There is already considerable debate surrounding the legal issues of driverless cars, the ethical concerns and questions over the liability of who to sue when a driverless car is involved in an accident.

Nevertheless the case for self-driving cars is considerable, the death toll (from motor accidents) is increasing worldwide with 1.3 million fatalities and 50 million injuries every year globally and around 93% of these accidents are caused by human error, and cost a staggering US$ 871 billion in the United States alone, according to the National Transportation Safety Board (NTSB) – self-driving cars will dramatically reduce accidents significantly. Autonomous cars will also reduce the burden on law enforcement, may also make streets safer and less congested, and allow those with disability greater movement – leading to a cleaner environment and improved quality of life.

Self-Driving Cars — Not if, but when

Lloyds reportAs well as the technical obstacles that are quickly being eliminated by manufacturers of self-driving cars, there are a few potential barriers that will delay the deployment or growth of self-driving cars, just as the ‘Red-Flag Act’ did in the 19th Century.

One such obstacle is Insurance Liability.

Lloyd’s of London, the insurance underwriter, has recently produced a report about how autonomous car technology will affect the insurance industry.

According to the report, insurance companies will have to create new methodologies on how they measure risk and charge motor insurance premiums:

There could be “significant changes in motor insurance: if accidents reduce in frequency, claims will also decrease, leading to lower premiums and, in turn, tighter profit margins for insurers.”

The report also highlights declining car ownership as potentially reducing the total number of insurance premiums:

Car ownership could decline in favor of a renting model and taxi companies could become owners of rentable car fleets, leading to company-wide policies.

With respect to insurance concerns Lloyds emphasized the problems with transferring risk from the driver to the machine:

With less reliance on a human driver’s input, however, increased risk would be associated with the car technology itself. Computers can do many things that a human driver cannot: they can see in fog and the dark, and are not susceptible to fatigue or distraction. However, they can also fail, and systems are only as good as their designers and programmers. With an increased complexity of hardware and software used in cars, there will also be more that can go wrong.

A new form of driving tests and licensing will be required:

By gradually transferring responsibility from the driver to the car, there is a risk that a driver could misjudge the responsibility they currently have… To mitigate this risk it is important that drivers are well-educated about the limitations of autonomous functions, and how they can retake control of the car when it is necessary.

When we rely on the car we may ‘switch-off’ and not be attentive to what is happening in the surrounding environment when it seems that their input is not needed.

If drivers are expected by law to supervise an autonomously operating car, they may find it difficult to remain focused on this task if they feel able to trust the car.

However the report does acknowledge the higher safety standards of autonomous vehicles:

The advent of autonomous cars could revolutionize the world of motor insurance. Autonomous cars could potentially lead to a substantial reduction in motor insurance claims.

Some might argue that if cars really do become crashless, there may not even be a need for motor insurance.

In many ways driverless cars may make assessments easier because it would be possible to see data such as what speeds the vehicles were travelling at.

An accident caused by autonomous technology, would need software and hardware analysis expertise in order to understand how and why it occurred. As sensors and computers become more commonplace in cars, and some driving responsibility is devolved to the car, an increase in telematics-based policies could be an option. Premiums could be better matched to exposure rather than based on proxies, and in the event of accidents, the car’s ‘blackbox’ could be examined. Whereas at present insurers using telematics devices incur the cost of their fitting, in the future the data may already be collected, making telematics a more viable option.

Whilst some analysts indicate self-driving cars will be on our roads in greater numbers with 3 to 7 years, it may be some time before the new technology earns the public’s trust. Lloyds also commissioned a Google Consumer Survey according to the survey, which was taken by 2,016 people in the UK, 50.4% male, 49.6% female.

People are uncomfortable placing their trust in a computer and they like the process of driving too much to give it up.

Automated driving will become progressively more standard in the near future, but it will require the development of high standard autonomous navigation systems and adequate safety standards, along with the implementation of new legal and licensing procedures. With a shift towards autonomous technology and away from human operators, the landscape of insurance, especially motor insurance, may change dramatically.

Picture source: 1st picture – Red Flag Act and 2nd picture Lloyds Report

5 Reads: Drones, Robots, jobs and the economy

Drones the wall street journalDrone Dogfight: Big Defense Firms vs. Techies – The users and makers of smaller drones are focusing their frustration with the FAA onto larger drone makers, who aren’t in such a hurry for rules. (The Wall Street Journal)

Humans 1, Machines 7 — It is easy to underestimate how quickly robotics is improving. (The Economist)

The computerisation of European jobs — Who will win and who will lose from the impact of new technology onto old areas of employment? H/T Robert Went (Link to article Bruegel)

The Rise of the Robots — What impact will automation – the so-called “rise of the robots” – have on wages and employment over the coming decades? (Project Syndicate)

Even greater workplace disruption lies ahead. Labor markets may once again be entering a new era of technological turbulence and widening wage inequality. (Project Syndicate)

Me, Myself and Robots

Robot builderIn his book The Moral Consequences of Economic Growth, Benjamin M. Friedman posits that steady economic growth: “fosters greater opportunity, tolerance of diversity, social mobility, commitment to fairness and dedication to democracy.” Friedman’s hypothesis is defended with an impressive range of evidence from disciplines including economic and political history and other relevant sources. He also highlights our understanding of the relationship between economic growth and our willingness to cooperate by drawing on literature in psychology and especially behavioral economics and he emphasizes that economic growth transforms society into a “cooperative venture for mutual advantage.”

But when economic growth stagnates society has less tolerance and social unrest coupled with people less willing to cooperate and more focused on satisfying their own basic needs – a me, myself and I world. Some authors have predicted we are already in a Great Stagnation, others indicate that economic growth, in ‘advanced economies’ will continue at a nominal rate of growth of circa 1 to 2 per cent for many years, possibly decades.

The big question is – are we likely to see the growth rates of the 20th and early 21st century again?  Even though much of it was built on a house of debt!

The economic boom after the Second World War was predominately built on two sectors: housing and automobiles – which collectively accounted for 40 percent of the high growth rates. In more recent years we can see the same sectors responsible for the high growth rates in emerging markets such as Dubai, Abu-Dhabi and other Emirates and in Europe countries such as Poland which emerged from Communism some 25 years ago, a housing boom and flourishing car market quickly followed. The same is happening today in China.

Flourishing housing and automobile markets have many spin-offs: fuel, highways, transportation, associated household equipment, mechanical parts, banking, finance, etc. As history shows, housing and automobile markets do slow and the consequences are now being felt across the developed world.

Without such markets how will we get the levels of employment near to 95 or 100 percent?

The facts are; new flat screen TV’s, smartphones, tablet computers, laptops, mobile Apps, etc. will not generate anywhere near the level of economic activity that is required to grow the economy and put people back to work.

Historically, technological innovation has provided the momentum for long-term economic expansion. The area with the greatest potential for technological innovation today is advanced Robotics and the associated technology of Artificial Intelligence. But let’s not kid ourselves, there is no point in developing these technologies unless they serve a significant purpose – and the overall purpose for robots and A.I., especially within the workplace, is greater efficiencies and reduced costs. Much of these reduced costs will, in the long run, come from reduced headcount – people.

I am a strong proponent of the advances in robotics. Our lives can be greatly enhanced by exoskeletons, driverless cars, artificial assistants and so on. In the short term these sectors will create more jobs, but over the next two decades the full economic benefits of investing in robots will mean one thing – less people working.

Government investment in infrastructures, green energy and other advanced technologies will of course help spur the economy. But these will have to be paid for from somewhere and with Government debts at record levels few countries can sustain the levels of investment required to make meaningful long-term investments.

This is where I believe a new welfare state is increasingly becoming one of the only sensible options. Many are calling for a ‘living wage’ or Basic income Guarantee. Seattle in the United States and Switzerland, where approximately 45% supported such a motion in a recent referendum, could prove to be very informative test cases. The former US Secretary of Labor Robert Reich has said:

We’ve got to seriously think about how we widen the circle of prosperity, how we get shared prosperity. Otherwise, who’s going to be the customer? And a minimal guarantee with regard to income, it seems to me as almost inevitable in terms the direction that the structural changes of our economy are taking us in.

With excessive household debt, low savings, wage stagnation, almost 2 in 5 on the poverty line in advanced economies it will take something of the order of a basic income guarantee to spur economic growth or we can be sure to witness Gartner’s prediction of mass social unrest in the coming decade.

Let’s also not forget – robot’s are now good at building structures and 3D printers are capable of building houses!

More jobs in companies that employ robots

Time magazine 1955On March 28, 1955, Time magazine reported on a new generation of machinery called computers. The cover featured a drawing of IBM’s Thomas Watson, Jr. in front of a cartoonish robot drawn by Boris Artzybasheff, over a headline that read, “Clink. Clank. Think.” The story marveled at a computer built by IBM, working inside a Monsanto office building. “To IBM, it was the Model 702 Electronic Data Processing Machine,” the story reported. “To awed visitors, it was simply ‘the giant brain.’”

Time equated the IBM computer with the advance of civilization. “The prospects for mankind are truly dazzling,” the article said. “Automation of industry will mean new reaches of leisure, new wealth, new dignity for the laboring man.”

The Time magazine article has echoes of Keynes prediction from his 1930 essay Economic Possibilities for our Grandchildren, that the working week would be drastically cut, to perhaps 15 hours a week, with people choosing to have far more leisure as their material needs were satisfied.  In his essay Keynes may have been looking towards 2030 when he wrote: “What can we reasonably expect the level of our economic life to be a hundred years hence?”

Even though his essay was written in the second year of the Great Depression Keynes highlighted the advances in technology that contributed to the impressive economic growth that the West attained since the industrial revolution:

From the sixteenth century, with a cumulative crescendo after the eighteenth, the great age of science and technical inventions began, which since the beginning of the nineteenth century has been in full flood – coal, steam, electricity, petrol, steel, rubber, cotton, the chemical industries, automatic machinery and the methods of mass production, wireless, printing, Newton, Darwin, and Einstein, and thousands of other things and men too famous and familiar to catalogue.

Keynes then gave a warning of  ‘a new disease in the years to come, namely, technological unemployment.’

This means unemployment due to our discovery of means of economising the use of labour outrunning the pace at which we can find new uses for labour.

Major current debate concerns whether new technologies are creating ‘technological unemployment’ whereby many workers are displaced by new technologies and find it difficult to become employed again.

Much of this debate points to robots as the main culprit of displacing people from the workplace. Professor Robert Gordon of Northwestern University, in a Financial Times article, invites us to: “think of every employee you’ve had contact with in the last two or three days, and think, is that person going to be replaced by a robot in the next 20 years?”

Thomas Watson Junior believed the introduction of machines into the workplace would not lead us (the average workers) into technological unemployment but provide more wealth and more leisure time and also free us up for more creative activities. From the1955 Time magazine article:

President Watson hopes to mechanize hundreds of processes which require the drab, repetitive “thought” of everyday business. Thus liberated from grinding routine, man can put his own brain to work on problems requiring a function beyond the capabilities of the machine: creative thought.

This creativity aspect is something author Tyler Cowen emphasizes in his book Average is Over, and venture capitalist Mark Andreesen screams: “Robots will not eat the jobs but will unleash our creativity.”

The fact — automation is driving technological unemployment, jobs are being obliterated. According to the International Labor Organization (ILO) there are 200 million unemployed people worldwide. Many of these, such as millions of professionals that worked in banks, insurance companies, travel agents, retail, and other service industries have been displaced as software automates their jobs.

Robots in, jobs go up 

But so far companies that implement robots (a mechanical hardware and software device that incorporates movement/action) are actually adding jobs. Our research shows 76 companies that implemented industrial or factory/warehouse robots actually increased the number of employees by 294,000 over the last 3 years. Amazon famed for the acquisition of Kiva Robotics has in fact added more than 89,000 new staff to its payroll over the last 3 years. The company now employs over 117,000 people, more than four times the 28,300 employees it reported on June 30th 2010. Tesla Motors, manufacturer of electric cars and with one of the most sophisticated robot manufacturing sites has added over 6,000 new jobs.

Hundreds of thousands of new jobs are being created in drone manufacturers, industrial robot makers and other sectors of the robotic arena. The EU recently announced the world’s largest investment in robotics and a target of an additional 240,000 new jobs in the region.

It is highly probable over a million new jobs will be created in the robotics sectors in the coming 5 years. Whilst significantly less than the ILO cited numbers to solve the world’s unemployment problems, nevertheless it is a step forward – when others are screaming that robots ‘ARE’ eating jobs – let’s be realistic, software IS eating jobs; robotics is currently creating jobs.

Whilst I am optimistic about the near future for job creation in the robot sector, in the long run it is also highly probable that robots will displace people from jobs. How this will play out in 2030 is anyone’s guess – it is not too far away so I would suggest asking the same question as Professor Robert Gordon above. I also know robot manufacturers are very aware of the likely societal impact robots can have on jobs. As Melvin Kranzberg said: “Technical developments frequently have environmental, social, and human consequences that go far beyond the immediate purposes of the technical devices and practices themselves.”

It’s better to be prepared than caught out. My advice to young people — now is a good time to join the robot sector.

Update – Business Insider has a nice summary and other pointers on this article – This Settles The ‘Robots Will Take Our Jobs’ Argument Once And For All

6 reads in AI, Robots, Drones and Economics

This is an excellent response to a post by Marc Andreesen – What you left out was the essential question: who owns the robots? (By Alex Payne)

Microsoft Unveils Machine Learning for the Masses (New York Times)

AI’s dueling definition: Why my understanding of AI is different from yours. (O’Reilly Radar)

More Robots won’t mean fewer jobs (Rodney Brooks on Harvard Business Review)

Before you travel to a city why not see how it looks from the air? (Travel By Drone)

Our Work Here is Done: Visions of a Robot Economy (Free eBook by Nesta). Contributors: Ryan Avent, Frances Coppola, Frederick Guy, Nick Hawes, Izabella Kaminska, Tess Reidy, Edward Skidelsky, Noah Smith, E. R. Truitt, Jon Turney, Georgina Voss, Steve Randy Waldman and Alan Winfield.

The economics of Amazon’s delivery drones

prime-air_high-resolution01

Last December there was a lot of skepticism when Jeff Bezos, CEO and founder of Amazon, announced on the 60 minutes TV program that they were looking into using drones for delivering small packages. Many pundits called it a publicity stunt and nonsense! — timed for the biggest online shopping day of the year .

Bezos on the other hand was bemused and took pains to point out in his 2013 annual letter to shareholders that Amazon are serious about delivery by drones, writing: “The Prime Air team is already flight testing our 5th and 6th generation aerial vehicles, and we are in the design phase on generations 7 and 8.”

On their Prime Air Q and A page Amazon anticipate FAA’s rules for commercial drones will: “be in place as early as sometime in 2015.” And state: “We will be ready at that time.”

To be ready they are assembling a team said to already consist of between 45 to 50 employees and at least 10 additional personnel sought according to job openings on Amazon’s website.

Job postings for the Prime Air team range from a Patent Lawyer, to a Communications Manager, Software Engineers, Machine Learning Engineers, Executive Assistant, Project Coordinator, Research Scientist and Technical Program Manager.

To give you a taste of what the company is aiming for, the Communications Manager post indicates:

We’re looking for a communications leader for Amazon Prime Air, a new delivery system that will get packages into customers’ hands in 30 minutes or less using unmanned aerial vehicles.

So what’s driving Amazon’s Prime Air initiative?

Instant gratification from customers is clearly one element; providing outstanding service is another; as is staying ahead of the curve with innovative delivery and order fulfillment. All highly significant points in their own right to meet Amazon’s goal: “to be Earth’s most customer-centric company.”

Cost of transportation is another. Amazon’s total shipping costs in 2013 were $6.635 billion. They received shipping revenue of $3.097 billion and incurred overall losses of $3.538 billion related to shipping costs.

Amazon shipping costs

Amazon use several services for shipping, UPS, FedEx, US Postal, and others as well as developing their own City Pick Up points, delivery van service and Amazon courier cycles. Shipping is clearly a major cost factor to Amazon and one where they are focused on improving service whilst reducing cost.

Technical difficulties

In addition to the regulatory hurdles that must be overcome there are many technical difficulties.

Amazon is aiming for their drones to deliver shoebox size packages.

They probably do not want video onboard due to the extra weight on the drone and also privacy concerns, so will need another way of identification before customers can accept delivery, e.g. biometric identification or pin code to release the package provided with the consignment email – package drop off will be a challenge.

Wind will be a factor in delivery.

Sense and avoid – very few of the current breed of drones (especially the hobbyist drones) have sense and avoid capabilities and should not be flown where there are people or objects.

Amazon will want their drones to be as safe as regular manned planes. Piloted planes have 9.4 accidents per million flights, in other words statistically very safe.

CyPhy Works is possibly the leader in this drone technology with their tethered drones having already developed high payload, high wind, environmentally sealed systems.

GPS lock is an issue that has caused drones to drop out of the air and piloting inexperience is also a major issue, although one the FAA is looking very closely at and I’m sure Amazon will too – drones should never be operated without formal training and some license arrangements.

Another hurdle will be location for dispatch – most of Amazon’s Fulfillment Centers are outside major cities, although it is probable that the items Amazon will provide via Prime Air will be a vastly reduced inventory and kept at the Amazon Pick Up points or smaller Fulfillment Centers closer to major city centers.

These are just a few of the technical hurdles, not insurmountable and as Amazon state they will be ready when the FAA approve the use of commercial drones.

Will drones be more cost effective?

According to shipping-industry analysts Amazon typically pays between about $2 and $8 to ship each package, with the cheapest option through the Postal Service and the most expensive via UPS or FedEx.

Amazon may be able to get a premium price for the Prime Air delivery service – customers who want their package within 30 minutes may be prepared to pay a premium of say $15 to $20 per delivery. Irrational for a book that costs $18, but as behavioral economics shows humans do not always act rationally.

During a 15 hour window (7am to 10pm) one Prime Air drone could potentially make 30 deliveries (absolute maximum efficiency and at a significant stretch).

Assume the drones are fully in service for 360 days per year, this will be the equivalent of 10,800 deliveries per drone.

The drones will require 2 full time pilots. With drone pilot wages ranging between US$13 to US$23 per hour, plus fulfillment center costs, insurances, drone fees and service, the Prime Air Team overheads and development – the annual fees per Prime Air drone service – annual cost per drone could be a minimum of between US$105,000 to $186,000 per year.

Those numbers divided by the maximum number of deliveries per drone indicate that the Prime Air service could cost Amazon between $9.75 and $17.44 per delivery, which is okay if Amazon get a premium rate for 30 minutes delivery.

Amazon will want to get maximum number of deliveries and efficiencies out of the drone capabilities, whilst reducing the costs as close to $2 per delivery as possible and maximizing the revenue by providing the wow factor to customers. They will also be aiming to reduce the US$ 8.829 billion in cumulative shipping losses in the last three years.

At some stage don’t be surprised if Amazon seeks to move into the logistics business. The robots that they are deploying in their fulfillment centers and now with Prime Air Drones, Amazon are clearly building a high quality, high capability logistics service – and ultimately that is good news for consumers.

Photos: Amazon 2013 Annual Report and Amazon Prime Air Q&A

As Helen Greiner, CEO and founder of CyPhy Works says: “The shortest path between two points is as the drone flies.”

Updated: There is an interesting take on this article edited by Business Insider — The 4 reasons Amazon is dead serious about its Drone delivery service.

A chatbot probably did not pass the Turing Test in Artificial Intelligence

UniofReading

Recent headlines have been awash with stories of a computer program ‘disguised’ as a 13 year old Ukrainian boy Eugene Goostman has passed the Turing Test. Similar claims have been made in the past (this paper was shared with me slightly tongue in cheek by Professor Joanna Bryson), indeed a couple of years ago it was claimed that Eugene came ‘close to passing the Turing Test’ when ‘he/it’ was the overall winner in a similar tournament. We will likely here of more programs that have passed the Turing Test in the near future – how close these claims are to the real ‘spirit’ of the Turing Test is certainly very debatable.

Author and Roboticist, Professor Alan Winfield told me he was less than convinced last Saturday’s announcement resulted in the Turing Test being passed, especially as Turing “intended higher thresholds, >5mins, and >30%.” The pass rate claimed in the University of Reading announcement was 33%, just a very narrow pass if Turing really was stuck on the greater than 30% of judges threshold, which is highly unlikely.

In fact Turing does specify 5 minutes but I am in agreement with Alan, he does not indicate how many judges or what percentage. What he says is:

“An average interrogator will not have more than 70 per cent chance of making the right identification after five minutes of questioning.”

Is this the same as the University of Reading and the organizers claim?  “If a computer is mistaken for a human more than 30% of the time during a series of five minute keyboard conversations it passes the test.”

So far very little has been shared about the organization of the event, other than the claim that 30 judges took part and ‘in each five minutes a judge was communicating with both a human and a machine,’ and 33% of judges were convinced that Eugene was human.

At no time does Turing specify in his paper 30 per cent of judges, and I would posit that he would expect the 70 per cent to be nearer the threshold.

To help analyze the number of judges, Turing states:

“A number of interrogators could be used, and statistics compiled to show how often the right identification was given.”

Turing originally set out that the Imitation game should also be based on gender.

The new form of the problem can be described in terms of a game which we call the ‘imitation game.” It is played with three people, a man (A), a woman (B), and an interrogator (C) who may be of either sex. The interrogator stays in a room apart from the other two. The object of the game for the interrogator is to determine which of the other two is the man and which is the woman. He knows them by labels X and Y, and at the end of the game he says either “X is A and Y is B” or “X is B and Y is A.”

“What will happen when a machine takes the part of A in this game?”  Adding later…  “Could the interrogator distinguish between them?”

It is not clear if the judges in the University of Reading organized event focused on gender, maybe they will clarify this when the ‘peer reviewed’ papers are released which they have referenced.

I am not convinced the Turing Test has been passed although it is fair to say that in recent years we have seen considerable progress towards the goal of a machine that can ‘trick’ judges into thinking it is human or more precisely a woman.

Murray Shanahan, a professor of cognitive robotics at Imperial College London, who is quoted in the Daily Telegraph as saying: “I think the claim is completely misplaced, and it devalues real AI research. It makes it seem like science fiction AI is nearly here, when in fact it’s not and it’s incredibly difficult.” Or as Alan Turing himself concluded his famous paper: “We can only see a short distance ahead, but we can see plenty there that needs to be done.”