A Look Into The Future – AI & Automation by Shamus Mahmood, MD at theimcs.com

I was waiting at Waterloo train station.  I went straight towards Foyles, the bookstore.  I did my usual browsing of the latest titles.  Among them, biographies, of business titans discussing how they made their fortunes.  Rather more nauseatingly, TV celebrities writing autobiographies.  We all know it’s massively important to know how the mind of a reality-tv star/model/footballer works. Then, I spotted a title – The Rise of the Robots – by Martin Ford

Rise-of-the-Robots-Book-Cover-198x300

Immediately, the mind conjures images of The Terminator & Cyberdine.  I picked it up.  It said “Technology & the Threat of Mass Unemployment”.  Ok, here is a genuinely serious issue that warrants further investigation.

The book took my breath away.  Brilliantly researched & written by a tech industry insider.  The article below is a summary of some of the points made in the book & a strong argument for people (and governments) to start thinking about the implications of automation & AI.

Do people know how much is already being done by robots? I wondered.  I suspect they don’t.

I’ve spent the last 10 years in digital media & I’ve worked in a tech start-up which achieved 2 things:

1 – Grew from zero -$100 million in 6 years – All via a small but significant technology that changed how the industry operates today.

2 – It taught me how big players like Google & Facebook employ slick algorithms to automate processes that were once the remit of teams of analysts.

A paradigm shift has been taking place for many years now and people are struggling to keep up.

  • In 1979- General Motors earned $11 billion with 840,000 Staff.
  • In 2014- Google (for whom I served as a Premier Partner) earned $14 billion with only 38,000 staff.

This is an incredible statistic.  38,000 staff out-performed 840,000 staff

For those who remember long queues in a supermarket, how much better is it that you can take your shopping and just pay via your card?  In the near future, this too will disappear.

The real impact, however, will be felt over the next 10-30 years.  By that point, many jobs will have become automated & the workforce no longer required

Martin Ford states, “The soaring Fortunes of those at the very top may ultimately represent a threat to democratic governance.  However, the most immediate problem for most middle & working class people is that job opportunities are broadly deteriorating”.

This is going to get very real with the election of billionaire businessman Donald Trump. The USA lead the way, with regards automation technology.

Less Jobs & More Automation:

US President Trump was voted in on the promise of jobs (but hasn’t even mentioned the automation challenges).  It doesn’t make for pretty or smart politics to discuss this so early into his presidency. In time, automation will do serious harm to his promise of increased jobs & make it near impossible to compete with the progress & advancements heading our way.

Moore’s Law famously states that the size of technology is halved every 2 years.  This brings us to an interesting place.  If technology is advancing at this speed, where will we end up?

Man v Machines

Professor Stephen Hawking & Elon Musk, among other great minds talk about their worry of the “singularity”.  This is a point at which computer & AI become so advanced that they are able to out-think human minds & there is no going back.  It sounds like sci-fi but Gary Kasparov was beaten by IBMs Deep Blue back in 97.  Chess, of course, is a rigid series of moves that can be pre-learned.  It’s different to knowing everything there is to know & out-smarting a human mind.  Enter IBMs Watson.  This machine beat humans on Jeopardy which is not like chess. The subject matter could be anything, anytime, anywhere.  It’s a staggeringly impressive achievement both in terms of computing power & no less importantly for IBMs, PR.

IBM-Watson-Jeopardy

In 2009, researchers at the Mayo Clinic in Minnesota, built an AI network to diagnose endocarditis.  This normally involves a probe inserted into the oesophagus to determine whether or not the inflammation is caused by a deadly infection. A study involving 189 patients found the system to be 99% accurate & saved over half having to undergo an expensive & unnecessary diagnostic procedure.

Between 2009 – 2013/14 – the UK set aside up to £15.6 Billion for clinical negligence.  With sophisticated AI machines, such as Watson, a second opinion could reduce errors & claims.  This money can go back to the NHS.  Eventually, machines will prove reliable enough to free up the time of the physician & save both time & money.  Robotic arms will dispense medicines & drugs instead of traditional pharmacists.

  • Retail

Amazon employs over 37,000 Robots.  37,000. They are on the verge of employing an army of drones to deliver.

In the USA, in one example, a famous chain of chemists was able to predict women’s pregnancy, based on analysing their spending habits (big data).  They were then targeted with pregnancy related ads, even before some of them had disclosed their pregnancy to partners.

We briefly mentioned supermarkets & the ability to pay for your shopping by yourself.  Amazon has already advanced this concept with their brilliant “Go” concept.  You just walk into the supermarket, pick up what you need & walk out. Technology does the rest. See below

  • Driverless Cars / Trucks & Automotive

Google, Uber & Tesla are all at the forefront of the driverless vehicle concept.  This has already been proven & it’s just a matter of time before they supersede the present model of having to drive your own vehicle.

Tesla’s tweet in April 2016 speaks for itself:  Over 47M miles driven on Autopilot, the more you drive, the more we’ll learn

In Tesla’s plant in Fremont, California, they employ 180 robots.  These robots assemble 400 cars a week. The robots autonomously swap tools so that they can do multiple jobs.

In Google’s vision, robotic cars are distributed in fleets. This would allow for the centralisation of car maintenance, fuelling, repair & insurance.  To put this into context, in the UK alone, over 200,000 people work in car washes.  Small businesses will be destroyed & will quickly need to adapt or find something else to do.  Big businesses will also be at risk, unless, like Uber, they are big enough to get in early.  Time will tell how this scenario plays out.

  • Media & AI – Journalism & Blogging?

Narrative Science is a company that employs powerful technology, which allows for the creation of articles via automation, rather than the traditional journalism approach.  Forbes is among the companies to employ this technology.

When the co-founder of this technology was asked to predict the percentage of news articles that would be algorithmically written within 15 years, his answer was over 90%.

As we speak, this technology is being enhanced to cover other business verticals. Feel free to research Quill.

Narrative-Science

  • Middle Management will slowly vanish

Big Data will eliminate the need for middle managers to make decisions.  The statistical & computational power of big data systems will provide real-time & useful information before a middle-level manager would even know he needs it.

Clerical & skilled analysts will be replaced by algorithmically ascertained information which will be accurate and cost effective

Companies will emerge – with the ability to take a project, determine which components can be outsourced & then automatically post these roles to Elance & Freelancer, thus removing the need for someone in-house to perform these roles. WorkFusion, a New-York based startup does exactly this, but there will be many others, all competing to be the king of the industry

  • Marketing/Advertising

Google is king here, along with Facebook.  However, IBMs Watson also offers its own analytics tool.  Imagine what can be achieved with that kind of computational power?

Google AdWords already employs algorithms to automatically run ad campaigns while making the optimisation process automatic. This used to be the job of teams of marketing analysts.

Big Data companies are emerging all the time & offer analytics & consumer insights.  Remarketing technology can now target consumers with a level of accuracy that is both impressive & scary, as per the “pregnancy example” above.

The Cloud – Why does it matter?

For those in digital, this is already common practice but just to give an example to those who do not work in IT or digital.  Here is the best way to demonstrate the benefits (I shall resist going into long explanations of what it is & how it works – in short – it’s saving everything in large supercomputers, rather than local house or home-based computers.  This information is readily accessible anywhere and anytime, with the right tools.

cloud-computing-768x432

A small company called Cycle Computing was able to solve a complex problem that would have taken 260 years, in just 18 hours.  This was achieved by utilising the tens of thousands of computers that power Amazons Cloud service.

One of my own hobbies is a movie-based website called www.schmoovies.com.  The images are housed in Amazons cloud servers because the website has over 10 million data points in it.  This costs approx £150 a month. As we move on, it becomes even less expensive.

Facebook & the cloud

One of the best examples is this – Facebook employs a cloud-based system called Cyborg.  It continuously monitors tens of thousands of servers, detects problems & performs repairs autonomously.  This used to require hiring hundreds of skilled IT workers.

However, even in this example, huge data storage centres in urban towns in the USA employer few full-time staffers because the real work is done by robots & AI programmes, such as Cyborg.

Good Data – San Francisco – 30,000 Jobs down to 180

Another example is Good Data in San Francisco.  The CEO Roman Stanek states that the company uses Amazon cloud services to perform data analysis for over 6,000 clients.  Each company used to require 5 people to do this work – that’s 30,000 jobs.  He now performs this task with 180.

Back to automation:

The UK Banking Industry

By the end of 2015 – Lloyds, Barclays, HSBC & RBS had cut 189,000 jobs between them.

It is accepted that after the financial crisis, there were a lot more regulations introduced but you can be sure that this is the result if increasing automation.

Impact of Offshoring

In 2007, the former vice-chairman of the US Federal Reserve Board of Governors (Alan Blinder) said that 30-40 million jobs (at that time) were “offshorable”.

In the UK, we are affected much more than the rest of Europe. The given figure (by researcher group Forrest Group) is that by 2015, of the million European jobs offshored, 75% were from the UK.  That’s a possible 750,000 jobs.

The UK Ministry of Defence said that 30% of the RAF fleet, by 2030, will be unmanned aerial vehicles.

UK Sec Def Hammond holding a drone

In 2015, the House of Lords estimated that 35% UK jobs will fall victim to automation in the next 20 yrs.  This time is upon us now and we must plan ahead.  I don’t just mean the government.  I’m including the small business & tomorrow’s students, everyone.  The consequences of not being prepared will be dire.

Legal Industry

Large software companies will appear.  They will have powerful algorithms that have the ability to analyse tens of thousands of documents & shoot out key responses in a short space of time, often seconds.  This removes the need for junior level paralegals and admin backup staff who would usually do this kind of work.

UK Students – An Important Observation on the situation today

  • 2007 – average starting salary for graduates (bachelor’s degree) – £24,293
  • 2012 – average starting salary for graduates (bachelor’s degree) – £21,701

This is a drop of 11% – for people with high-quality degrees.

Over the same period of time, student debts doubled from nearly £22 Billion to £46 billion.

Higher Education

In 2016, the University of Akron’s College of Education did a study comparing machine marking to human instructor marking of exam results.  The results?  Virtually identical & in some cases, the machines proved more reliable. (16,000 pre-marked students essays across six US States).

These results proved similar to an investigation carried out by the Cambridge Assessment looking at GCSE Biology results.  They set out to test how well a machine can mark these exams.  The early results proved similar to the US example with virtually identical levels of reliability, matching the human markers 88% of the time.

Algorithmic marking systems are on the way.  They are not yet perfected, but they are coming.

3D Printing

GE plans to use 3D printing to produce up to 100,000 parts by 2020.  This can potentially reduce the aeroplane weight by 1,000 pounds.  To put this in context – in 2013, American Airline removed all paper flight manuals and put them onto IPads in digital format.  That saved 35 pounds per plane.  That result saved $12 million in annual fuel costs.

To give a more extreme example of the disruptive potential of 3D printing, we need to take it to construction size.  An engineer (Dr. Behrokh Khoshnevis) at the University of South California is building a 3D printer (Contour Crafting), capable of fabricating an entire house in just 24 hours. This can be used for building habitats on other planets & has many applications.

3D-Printing-a-house-300x192

In the UK – over 2.1 million people are employed in the construction industry.  Imagine the impact this kind of technology will have on their jobs.

The future

This isn’t all doom & gloom.  There’s plenty of good news too.  Nanotechnology represents incredible opportunities.  In 2013, The Indian Institute of Tech announced a nano-particle-based filtering technology that can provide clean water for a family at a cost of just $16 per year.  This is truly incredible. There is the desalination of the ocean to consider.

Carbon Nanotubes have proven to be extremely strong & also better at conducting electricity.   This offers opportunities for cars & aircraft construction.  We also have Graphene, for which the Physicist inventors won the Nobel Prize.  This also offers incredible opportunities for electronics & biotechnologies.

The purpose of this article was to just get one thinking about the challenges that lay ahead.

Martin Ford provides a full chapter on a potential solution with a completely new economic paradigm.  It’s not the intention of this article to discuss the merits of the approach or the chapter itself.  Rather, to draw attention to these incredible challenges this at once seemed like the future.  Today, they represent the present.

There are hundreds of references to back up the examples provided in this book.  All examples & references are to be found in the work on which this article is based.  For those who wish to know more, you can obtain a copy on Amazon by clicking the link below.

The 2015 McKinsey Business Book of the Year.  The Rise of the Robots, by Martin Ford.

I do hope you enjoyed this glimpse into the next 20 years. It’s an exciting time to work in digital & I look forward to hearing your thoughts on some of the items discussed here.

Written by Shamus Mahmood

Headshot Guy

MD – The IMCs.

Shamus@theimcs.com

@TheIMCs

© Sumner & Scott. Website by Outlines Design.