Chapter 5 Transforming Higher Education

This chapter discusses how the highly-skilled higher education industry may be on the verge of a huge disruption.

Analysis of algorithmic grading of papers shows machine marking is very consistent with human markers, and sometimes more reliable. It provides low-cost, fast, consistent, marking with instant feedback. However, some academics are against its usage. It could be argued that not implementing automation in education is partly responsible for the high cost of education fees, and student loans.

The chapter continues with a look at Massive Open Online Courses (MOOCs). Many universities have put their courses online, available for free or low-cost. This has both pros and cons. Students can attend a renowned university, learning from the best teachers, however problems with verifying student identity, and cheating exist. That said, it is hoped MOOCs might provide a way to help with the education of the poor. Potentially, MOOCs could disrupt education significantly.

Chapter 6 The Health Care Challenge

Like higher education, healthcare has been little impacted by automation. Medicine changes rapidly, and published research doubles every 5 years. There is too much information for individuals to keep up to date with, but all this data is ideal for processing with computer systems.

The chapter continues with a look at the impact of AI on medicine. IBM’s Watson team are working with AI and Machine Language to make huge volume of data queryable by doctors via the internet.

Automation could prevent human errors and reduce malpractice liability. It’s stated that 98,000 patients in the US die each year due to preventable medical errors. Systems exist that reduce human error, by eliminating any human involvement. Such systems ensure the correct dosage and the medicine is taken in a timely manner. AI imaging systems have proved particularly helpful.

With the aging population, there’s plenty of interest in elder-care robots. Much of research is occurring in Japan, which has a rapidly aging population.

Big Data processing is likely to identify new correlations between factors that impact health. Also, implanted sensors can monitor and report problems e.g. low glucose in blood of diabetics.

The chapter ends with a bit of a rant on healthcare being a dysfunctional market. AI and robots are making some progress in education and medicine, but much more could be done.

Chapter 7 Technologies and Industries of the Future

Historically, new technologies have caused ‘creative destruction’, where old jobs are lost, but other, often better jobs are created. The author suggests this pattern is no longer valid, since the newer technologies require fewer people. The chapter takes a look at 2 specific new technologies, 3D printing and autonomous (self-drive) cars.

3D printing allows the printing of objects using various materials (e.g. metal, plastic). The future hope is the ability to print human organs. It’s noted that 3D printing could impact manufacturing. There’s an interesting example of a house being printed.

Autonomous cars have improved dramatically since Google became involved in 2008. These cars consistently outperform human drivers in many aspects of driving. As a side-effect, many car makers are now creating semi-autonomous cars. It’s estimated that 1.25 million people are killed on the roads globally each year, autonomous cars should significantly reduce this total. However, since the car industry has political influence, these changes may be resisted.

Warren Buffet has recently said the autonomous car would be disaster for the insurance industry – what better recommendation could you want that these cars will significantly reduce accidents!

Chapter 8 Consumers, Limits to Growth . . . and Crisis?

This chapter looks at the dangers of automation. As jobs decline, the lack of disposable income may lead to a fall in goods and services, the closure of businesses, and a spiral of job losses.

Workers are also consumers, whereas robots aren’t, less consumers can a dramatic effect on the economy. There is no general agreement amongst economists, they can look at the same data but report on it differently, often along political lines. Although economics is increasingly mathematical, it uses past data, so forecasting has limits.

The economy is complex, with interdependences and feedback loops. Forecasting and interpretation is difficult. Remember how often these economists have failed to forecast recessions.

There’s a wonderful quote from Keynes about economists and their models: “Too large a proportion of recent ‘mathematical’ economics are merely concoctions, as imprecise as the initial assumptions they rest on, which allow the author to lose sight of the complexities and interdependencies of the real world in a maze of pretentious and unhelpful symbols.”

Another option, not examined, involves the small but real possibility of discontented people initiating unrest and revolution.

Chapter 9 Super-Intelligence and the Singularity

It is suggested that there may come a time (the Singularity) when AI systems are more intelligent that humans, and these systems then improve themselves exponentially. These systems could outperformance the finance markets , manipulate leaders, and invent new weapons etc. Stephen Hawking recently said ignoring this may potentially be our worst mistake in history.

The chapter discusses the pros and cons of getting to the Singularity, together with its many supporters and detractors.

This is quite a radical chapter, in some ways it doesn’t belong to this book, being set too far in the future and very speculative. For a deeper understanding, see Nick Bostrom’s Superintelligence (Oxford University Press)

Chapter 10 Toward a New Economic Paradigm 

This chapter discusses approaches that might ameliorate the impact of automation on job losses.

Historically, more education and training has been the answer, but the author has argued that these will be inadequate this time. There’s a short discussion about the diminishing returns from education and training. Next, the anti-automation view examines stopping automation, but typically someone somewhere will introduce such a system and undercut the competition, leading to job losses.

The chapter makes the case for a Basic Income Guarantee, a universal safety net to supplement low incomes. Factors discussed include: ensure there are sufficient incentives to encourage work and education, the public has paid for much automation research in the past so should be given something back, and the Peltzman effect (people take more business risks with a safety net in place). The various downsides are also discussed, together with some solutions to limit their worst effects.

The chapter ends with the suggestion it may not be feasible to implement the Basic Income Guarantee at the moment since US politics is currently very divisive.

It’s interesting to note that since this book was published (May 2015), Finland is examining the Basic Income Guarantee and a similar system is being tested in the Dutch city of Utrecht.

Conclusion

What will be the impact of automation on jobs? White-collar jobs? Your job? This book provides an interesting discussion on what has happened in the past, what is currently happening, and what may happen in the near future.

The book is generally well written, makes plausible suggestions, provides some evidence to support the assertions, is thought provoking, and has plenty of references and links for further information. The book is also selective, repeats itself, is disjointed, wanders into irrelevant areas (e.g. climate change, the Singularity), is overly gloomy, and polemic. Some sections are unnecessarily long.

To get some balance and perspective, I would suggest you also read Russell’s essay In Praise of Idleness www.zpub.com/notes/idle.html, and Keynes’s essay Economic Possibilities for our Grandchildren www.econ.yale.edu/smith/econ116a/keynes1.pdf, both written in the early 1930s depression era, both welcome a time when people could work 3 or 4 hours per day, spending their free time in pursuit of enjoyable pastimes.

This book misses a major underlying problem we have (but acknowledged by Keynes) – the need for more stuff. Do we really need to acquire so much stuff? Or need the latest product versions? Perhaps we need to educate ourselves on what’s important. Marketing has a lot to answer for...

Overall, an interesting, thought provoking read. Recommended.

To keep up with our coverage of books for programmers, follow @bookwatchiprog on Twitter or subscribe to I Programmer's Books RSS feed for each day's new addition to Book Watch and for new reviews.

<ASIN: 0465097537>
<ASIN: B00PWX7RPG >

<ASIN:0198739834>