Robots Are Coming...

Robots, replicants, androids, cyborgs, automata… call them what you will; we humans have been fascinated by the idea of creating beings to do our bidding since the beginnings of civilization. Homer tells in The Illiad how Hephaestus, the blacksmith of the gods, created two golden maidens in his forge to keep him company, endowing them each with the powers of speech and reason. Ovid’s Metamorphosis has Pygmalion carving a statue that he falls in love with, and which eventually comes to life. The tale was reprised in the early 20th century by George Bernard Shaw, in which diction expert Henry Higgins believes he teach Cockney flower seller Eliza Doolittle to pass herself off as a high-born lady, and of course he too ends up enamored of his creation.

And a hundred years later, motion pictures and television continue to reflect our fascination with robotics, highlighting the dangers they present to humans, which are as often sentimental as existential. Ridley Scott’s seminal Blade Runner portrays a dystopian future where the dirty work is done by replicants—genetically engineered beings with many human qualities programmed to live for four years. The job of the film’s hero is to liquidate rogue replicants, but of course, echoing Pygmalion, he falls in love, eventually fleeing with her, unsure if he can program her to live longer, and so coming to terms with our other obsession, mortality.

In fairness, perhaps love really is the acid test of our relationship with automatons: if we can accept the idea of affection flowering between us and them, then surely it is just a short step to establishing other kinds of emotional ties such as friendship or the unique bond that can develop between a teacher and a pupil.

By 1968, when Arthur C. Clarke wrote 2001: A Space Odyssey, which was adapted that same year for the movies by Stanley Kubrick, computers was already being used in some areas of business and were at the heart of our project to conquers space. In the film, a computer called Hal (Heuristically Algorithmic Computer) is partly responsible for running a spaceship on an interplanetary mission. In many ways, he is just another member of the crew, interacting with his human colleagues and is, for example, able to express humor. But Hal soon starts taking his own decisions, dispatching his human colleagues one by one until the captain manages to disable him. In his final, pathos-filled moments, as he winds down, Hal intones an old music hall tune he was taught by his human creator. The story highlights our lurking fear about robots: that they cannot really be trusted, and that one day, they will try to take over the world.

According to Martin Ford, author of the sinisterly titled The Rise of the Robots: Technology and the Threat of a Jobless Future (1), computers are not so much taking over the world, as our jobs, which, thinking about it, may amount to the same thing. In part, the blame for this shift lies with globalisation, says Ford, which has seen companies relocate to areas of the planet where wages are cheaper, a process aided by the decline of the unions. The result is growing inequality and an ever-wider wage gap. All this has created the perfect storm, with diminishing job creation, economic recoveries that take longer to generate employment, and soaring long-term unemployment that means lower wages and under-employment among recent graduates. The picture not only applies to the United States, but also to Europe.

Technology initially impacts most on sector susceptible to scaling up such as agriculture or industry, but over time it ends up affecting many other business areas. As we have seen over the last century, repetitive and routine jobs are the most vulnerable to the impact of technology. Automation is set to transform the logistics and transport sector in the coming years with the arrival of driverless vehicles, which are already being road tested in the United States and some European countries. At the same time, more creative activities, or those with greater added value, will also be hit. For example, the development of sophisticated software and the evolution of algorithms make it easier than ever for companies to manage big data.

High value added professions such as financial analysis are increasingly susceptible to automation, with complex applications that can follow the markets in real time, relieving analysts of part of their traditional workload. That said, a final risk evaluation of a financial operation is still an activity that, due in large part to a range of subjective factors, cannot easily be determined by algorithms.

The Rise of the Robots cites a recent study predicting that over the next two decades half of the jobs in the United States currently carried out by humans will be automated. Ford even goes so far as to suggest that this prediction is overly conservative: “A great many college-educated, white collar workers are going to discover that their jobs, too, are squarely in the sights as software automation and predictive algorithms advance rapidly in capability” (2)

As he points out, machines have already been developed that are able to beat humans at many intellectual tasks or abstract thinking, among them IBM’s Watson, which has scored better than many contestants on the popular US quiz show Jeopardy. Google Translate, while still far from perfect, is now widely used around the planet for basic communication, while a company called Narrative Sciences is developing ever-more sophisticated software able to write syntactically and grammatically correct reports drawing on a wide range of sources and at speeds flesh and blood consultants cannot imagine competing with.

Narrative Science’s software can write articles and highly detailed analyses synthesizing the vast amount of information available on the cloud or that is held by businesses and other organizations, formulating reports that can be used to make decisions. What’s more, the software can be used to produce news items, literary reviews, and even opinion articles, and may well put many journalists out of a job or enable enterprising editors to set up new online publications with just a few key staff (3). In fact, Ford notes that several respected major publications already use such software. With an eye on the future, the CIA has invested in Narrative Sciences, presumably hoping to make better use of the huge amounts of data it produces.

But other scholars have described Ford’s vision of a robot-run world in our lifetime as exaggerated. Thomas Davenport, for example, Distinguished Professor in Information Technology and Management at Babson College, argues that even in the most automated sectors, traditional jobs will remain. (4) Automation will take time to implement, he says, and in the meantime, new types of jobs will appear. What’s more: “there are cultural, legal, and financial barriers to widespread automation in a variety of industries (and) these obstacles won’t ultimately prevent automation-based job loss, but they will certainly slow it down”.

Even Martin Ford has to admit that some sectors are relatively protected from the rise of the robots, and along with other academics, identifies healthcare, management, engineering services, and notably education as relatively resistant to automation (5).

There are perhaps two main reasons why education is holding its own: greater government control and regulation, and the way most academic institutions are run, which tends to slow down decision-making processes.

Ford discusses the disruptive phenomenon of Massive Open Online Courses (MOOCs), but concludes that for the moment, they have not had the devastating impact on formal education that so many people predicted. I have explored the ways MOOCs have affected education, along with their possible medium-term benefits, mainly as ancillaries to traditional teaching programs. (6)

Our ivory towers will not protect us forever against the rise of the robots, and so we in academia are obliged to think about the educational tasks, specifically teaching, that could be substituted or complemented by technology. In short, will we really see classrooms led by a robot teacher?

Perhaps not, but there are any number of routine activities related to teaching that could easily be assisted or replaced by software.

For example, marking papers, assessment, and performance. At present, software applications are largely limited to marking multiple choice tests or simple exams, but we can easily imagine how the programs Narrative Science is working on could be used to mark essays. That said, surely a student’s final assessment, even taking into account what technology can provide us with in terms of performance, would require a more complex analysis that only a teacher could satisfactorily provide.

But as Ford notes, even when it comes to using computers to lighten their workload, relieving them of tedious tasks such as marking exam papers, many academics suspect the worst. In 2013 academics throughout the United States signed a petition against the introduction of automated essay scoring (AES) called Professionals Against Machine Scoring Student Essays in High Stakes Assessment. The organisers, from some of the United States’ most prestigious academic institutions, dismissed AES as, “simplistic, inaccurate, arbitrary and discriminatory,” citing research to back up their opposition, but without providing specific examples of where students’ work had been mismarked. At the same time, we should remember that similar programs are used when processing applications to the same universities where many of these academics work.

When it comes to assisting with one-on-one activities such as tutorials, coaching, or mentorship, the software currently available is still not sufficiently developed, and progress is unlikely in the medium term, with the exception of activities focused on transmitting information, testing knowledge, or self-help activities to strengthen the learning process. A motivation session with a coach cannot be reduced to the limited replies that recent artificial intelligence projects such as EMOSpark, which can analyse 90 different facial expressions, thus supposedly enabling it to work out its owner’s feelings and emit a corresponding message of greeting, encouragement, or support, for example. (7)

It is easy to see how lectures, where it is not necessary to attend in person any longer, will soon be replaced by MOOCs, videoconferencing, recordings in a range of formats, and other solutions. This is already evident in the phenomenon of the “flipped classroom” a blended approach where teaching is often delivered online, moving activities like homework, into the classroom, creating a more engaging and deeper learning experience. For this approach to work, the teacher must be able to innovate and take a more creative role, something that is still outside the remit of algorithms.

The flipped classroom extends the learning momentum, creating a continuum between classroom teaching and individual work, aiding the understanding and application of ideas and concepts. To a large degree, learning is about repetition, as well as about finding ways for students to assimilate knowledge, and this is one approach that can help.

Similarly, interactive sessions such as case studies require a teacher to coordinate debate; again a role unlikely to be given to a robot. But teachers can use computer programs to help them in the classroom, setting time limits, monitoring who is taking part, or aiding them in assessing students’ contributions to the discussion.

Once again, artificial intelligence machines are more like to be able to manage classrooms when it comes activities related to facts, data, and information, which are easier for them to recognize. But at present, most such machines would not be able to deal with contributions that include abstract thinking, concepts, or more original thinking, all things that identify the brightest and best students.

Applications also exist that can measure the extent to which students are following and understanding topics by using rapid confirmation tests, and that over time will doubtless be perfected. We can also imagine more sophisticated solutions that adapt teaching to the habits and tastes of individual students, as well as to their distinctive forms of intelligence. For example, a student with artistic inclinations, math could be explained in the context of painting or architecture, while more action-oriented pupils could be taught through the use of sporting examples.

The benefits of using specialist technologies in teaching certain groups with difficulty integrating, such as autistic children, have already been shown. (8)

Technology also makes it possible to convert a classroom session into an event where content, information, and opinions can all be accessed in real time, while permitting the virtual participation of any guest, regardless of their location. Similarly, students scattered all over the world can interact, making for greater diversity, with the corresponding learning benefits. Many universities give classes or seminars in which their own students connect via streaming with those from other places of learning. SNOCs (Small Network Online Courses) such as those developed by schools that are part of the Global Network for Advanced Management use this type of methodology, allowing students from member business schools around the world to take part in classes.

Taking everything into account, my feeling is that technology isn’t a zero-sum game. There is no point in thinking about the relationship between teachers and technology, or robots and software in terms of one or the other; instead they should be seen as complementary, the goal of which is to improve students’ teaching and learning experience, and ultimately, education overall. In reality, teaching technology should be an extension of the teacher, which may mean at some point in the future that we see biotechnological integration. Bill Gates is one of the biggest investors in introducing technology into education, and asked the question in his The Road Ahead whether teachers should or could be replaced by technology; answering with an emphatic no.

The creativity, originality, and sincerity of the best educators are not qualities that will ever be replicated in robots or programmed into artificial intelligence machines. Anybody who doubts this need only remember their school or university days and those unique moments when suddenly something makes sense, thanks to the patience and empathy of a skilled teacher. Do we really believe that a machine or an algorithm could ever help shape our character or inspire us to pursue a particular profession?

In Woody Allen’s 1973 comedy Sleeper, the inhabitants of a future world have a machine, the Orgasmatron, a capsule big enough for two people that produces a near-instantaneous orgasm seemingly without the need for any physical contact. As a result, we learn that the members of this civilization have become impotent, and that traditional forms of sex have been abandoned.

Let’s imagine for a moment that we could invent a similar machine, but this time with didactic properties, called the Educatron, a capsule in which students could, in few minutes, learn an entire university degree course, or acquire the contents of an encyclopedia, or learn any number of new languages. The problem with such an approach to education, as with the Orgasmatron’s toward sex, is that we would miss out on the learning experience, on the effort and demands it puts on us, as well as the benefits of making choices and thinking things through. Learning from a teacher is not just about the end result, whether that applies to learning a theory or a language, but the road travelled and the means used to take us to our destination, which should be enjoyable, satisfying, and as valuable in terms of our personal development as in the knowledge accrued. And it is for this reason that those who see technology as a shortcut, as a replacement for the learning process, are wrong: there is as much value in what we learn as how we learn. Education can be seen as a voyage that lasts a lifetime, where knowledge of the destinations we reach is just as important as the experience lived along the way.

 Notes

Photo: https://wallpaperswide.com/robots_of_the_future-wallpapers.html

(1)Martin Ford, The Rise of the Robots: Technology and the Threat of a Jobless Future, (Basic Books: New York, 2015). References to this book here are made to the Kindle version.

(2) ibid, location 147.

(3) An article in Wired mentions this software: Steven Levy, Can an Algorithm Write a Better News Story Than a Human Reporter?, Wired 24/4/2012

https://www.wired.com/2012/04/can-an-algorithm-write-a-better-news-story-than-a-human-reporter/

(4) Thomas H. Davenport, Just How Serious Is The Automation Problem? The Wall Street Journal, 20 May 2015

https://www.dhirubhai.net/pulse/article/just-how-serious-automation-problem-tom-davenport

(5) Carl Benedikt Frey and Michael A. Osborne, The Future of Employment: How Susceptible are Jobs to Computerization? Paper, 17 September 2013, Department of Engineering Science, University of Oxford, Oxford, UK.

https://www.oxfordmartin.ox.ac.uk/downloads/academic/The_Future_of_Employment.pdf

(6) Santiago I?iguez, Taking MOOCs Seriously,

https://www.dhirubhai.net/pulse/article/20140709182023-557690-taking-moocs-seriously/edit

(7) Sally Adee, Say hello to machines that read your emotions to make you happy, New Scientist, 14 May 2015

https://www.newscientist.com/article/mg22630212.900-say-hello-to-machines-that-read-your-emotions-to-make-you-happy.html

(8) Tina R. Goldsmith, Linda A. LeBlanc: Use of Technology in Interventions for Children with Autism, JEIBI Vol. 1, Issue 2, 2014

https://files.eric.ed.gov/fulltext/EJ848688.pdf

(9) Bill Gates, The Road Ahead