The upcoming holidays offer the opportunity to highlight some books that we've so far missed out and that -- depending on your idea of fun -- might provide entertainment over the break. Before we start on 2019, we should note that in 2018 we overlooked two important books: Virginia Eubanks' Automating Inequality (2018) and Safiyah Umoja Noble's Algorithms of Oppression (2017). Both provide important perspectives on the unfairness and discrimination endemic to the data being fed into today's decision-making algorithms. Eubanks focuses on the scoring systems being put in place in benefits offices, the criminal justice system and financial services, while Noble writes about the negative biases against women of colour embedded in search engines and other algorithms. Biased data -- as news headlines have gone on to make increasingly clear since they were writing -- is everywhere.
Data will be generated in huge quantities by the Internet of Things (IoT), which is Alexandra Deschamps-Sonsino's territory. She runs IoT-LDN, a monthly London-based discussion group that attracts entrepreneurs, techies and others who are interested in this fast-growing sector. In Smarter Homes, she traces the changes in home design over the past century as new technologies such as electricity, telephones, gramophones, televisions, central heating and air conditioning, and computers entered our living spaces.
In Deschamps-Sonsino's historical sweep, the IoT is just the latest in a long series -- but one that may change entirely how we relate to our homes. As she writes, today's smart home designs represent a marked disruption: where in past decades the idea was labour-saving, today, we're being asked to spend much more of our time and attention monitoring and updating the apps and devices we install, all of which require ongoing after-care from both the companies that supply them and ourselves as purchasers.
SEE: Sensor'd enterprise: IoT, ML, and big data (ZDNet special report) | Download the report as a PDF (TechRepublic)
Every generation, however, needs its new 'modernity', and so, Deschamps-Sonsino concludes, today's IoT is bound to become mundane and uninteresting. While she can't guess what will be next, she does surmise that in the shorter term we will become much more discriminating about balancing the value of a service against its privacy-intrusiveness. Expect, she concludes, greater variation in individual homes than the sameness depicted in traditional images.
A key element is the business models that will govern those apps and services and the data they collect. In Don't Be Evil, Rana Foroohar, an associate editor at the Financial Times, mercilessly documents the metamorphosis of today's big technology companies from the do-gooder images they began with into the threats to democracy that are being called in front of government committees in multiple countries.
'Don't Be Evil' is, of course, the motto that Google famously included in its IPO prospectus and then quietly dropped from its code of conduct in 2015. Foroohar, however, reminds us that in giving Sheryl Sandberg free rein to build the company's advertising behemoth they knew that what they were doing could compromise the results served up to users. As Google founders Sergey Brin and Larry Page wrote in the original 1998 paper that outlined their new search engine, "We expect that advertising funded search engines will be inherently biased towards the advertisers and away from the needs of the consumers." They have spent the past two decades testing their own belief, followed by fellow Silicon Valley big names such as Amazon, Facebook, Apple and Uber.
Two things sparked Foroohar's decision to write this book. The first was reading that 80% of corporate wealth was now held by just 10% of companies -- those that were leveraging information and networks. The second was more personal: a $900 bill run up by her usually responsible ten-year-old son inside an online soccer game that showed Foroohar the hold the tech companies have on their customers.
While Foroohar supports the idea of reverting to the pro-competition antitrust law of earlier in the 20th century, she doesn't believe that doing so will dent the profitability and power of these huge companies. Instead, she believes that rebalancing their power will need to include extracting a "digital dividend" from data collectors to benefit all of us who feed their machines. Privacy law should also be improved to ensure that users unambiguously opt in for collection and that the tech companies are required to be more transparent about how their algorithms work, including keeping audit logs of the data fed into them. She also favours creating a legal framework for digital rights and consumer protection to ensure we can access our personal data and understand how it's being used. Finally, she favours greater transparency about assets and algorithms for the benefit of investors and a global flat tax to prevent multinationals from leveraging varying tax regimes.
In part, Foroohar blames the tech companies' hero-to-villain conversion on the lack of diversity among Silicon Valley's "engineer-kings". Far more detail on the consequences of this appears in Caroline Criado-Perez's award-winning Invisible Women, which looks at the many ways women are left out of the design of our world. Some of these aspects -- particularly the failure to include women in tests of new pharmaceuticals -- have been known for years, but others, such as using only crash test dummies sized to match the 'average male', were new to us.
Criado-Perez, who successfully led campaigns to get Jane Austen onto the £10 bank note and a statue of suffragist Millicent Fawcett into Parliament Square alongside those of 11 men, writes with a fine, angry passion about the gaps in data caused by presuming that the "average white male" is the default human. As a result, designers create spaces -- everything from cars to kitchens -- that are uncomfortable for many women. Smaller hands struggle to grip today's larger smartphones, and a failure to appreciate anatomical and sartorial differences results in long lines in front of the women's toilets while men zip up and zip out. (Since the book appeared, Criado-Perez has developed a Twitter persona as the "toilet lady" because she's fielded so many of these complaints about railway stations and theatres.)
Criado-Perez does not limit her canvas to 'women's issues': workplace safety suffers whether the plastic fumes are emanating in an automobile factory or a nail salon. Similarly, the "myth of meritocracy" applies whether you're a woman struggling to be taken seriously in Silicon Valley, where "its God is a white male Harvard dropout", or a historian trying to teach "a people's history" where 'people' has a broader meaning than 'white men'. Overall, there's much here that's new even if you think you're already well-informed on the subject of gender bias -- and Criado-Perez has done her research to present a solid evidence base.
At least men and women share human language, even if the most-translated written work, the Bible, has only been translated into about 10% of the world's 7,000 languages. In Extraterrestrial Languages, Daniel Oberhaus surveys the effort to first contact and then communicate with aliens. My personal model for this sort of problem is aquarium fish, because everything about them adds to the difficulty of bridging the human-piscine gap: the scale is wrong, the environments are incompatible, and there's no common culture. The one thing we manage is that the human appearing next to the tank suggests food is imminent, and the fish rushing to the side of the tank shows they're hungry. (Hint: they're always hungry.)
For this reason, it's never been clear to me whether any of the things that have been sent into space -- from mathematical symbols and radio bursts to music, art and ideograms, to artificial intelligence pioneer Marvin Minksy's suggestion of sending a cat -- could actually become effective communication. Granted, fish are not typically lauded for their intelligence. But, as Oberhaus writes, while humans have managed to teach dolphins a small array of human words, we have had no success in learning anything about what all those clicks mean (perhaps partly because dolphins have displayed no interest in teaching us lesser intelligences their language).
Oberhaus makes good companion reading to either viewing the 2016 movie Arrival or the Ted Chiang story it's based on, Story of Your Life, both of which describe the process by which a linguist decodes an alien language and finds that doing so alters the way her mind works. Chiang's 2019 book, Exhalation, contains nine evocative stories whose themes, even when they're familiar, seem completely new. In 'Exhalation', a being whose nature is not specified (but which reminds one of Philip K. Dick's self-observing robots), dissects its own brain and develops a new understanding of the universe it lives in. In the novella 'The Life Cycle of Software Objects', a small group of developers guide a group of "digients", virtual beings that are sold as pets, but that these developers seek to help reach their maximum potential through a long series of threats, including unscrupulous predators and changing business models. In 'The Merchant and the Alchemist's Gate', Chiang explores time travel through a nested Arabian Nights-style series of stories. In 'The Truth of Fact, the Truth of Feeling', Chiang draws a parallel between a native learning to read and write under the guidance of a missionary and the advent of completely searchable life-logging. While the modern parts of the story are reminiscent of Black Mirror (specifically, season 1, episode 3, 'The Entire History of You'), the effect is entirely different: what is honesty in a world where everything can be checked?
Read more book reviews