AI

20 May 2018

Posted on May 20, 2018 Leave a Comment

Late Hans Rosling’s “Factfulness: Ten reasons we’re wrong about the world – and why things are better than you think” should be mandatory reading for anybody working in development. It is a fun and educating book. It uses anecdotes from Rosling’s eventful life to illustrate the absurdity of widespread development beliefs shaped by the media and activists. Because it is good storytelling, it sticks with you. It also infects the reader with Rosling’s passion for exploring data and taking the long view. So even if you think you have seen his bubble graphs or heard his monkey survey story one too many times, go and get yourself a copy.

McKinsey Global Institute’s “Notes from the AI frontiers” analyzes 400 applications of AI in 19 industries. It has useful maps of analytical techniques and AI potential per industry and functional areas. It shows that AI is mostly used in combination with traditional analytics: AI augments more often than it replaces. And it estimates the potential annual value creation of AI around $3.5-5.8 trillion with greatest promises in marketing, sales, supply chains and logistics.

I liked Iman Ghosh’s “A world of languages” and wished I had the time to merge its data with demographics to see what it would look like in 2030 and 2050. With Africa doubling its population by mid-century, the size of French will grow significantly, “making it the most-spoken language by 2050”. Wait. What? I need to share this with my kids as they start packing to come join me in France soon.

My quote this week is from Chobani CEO Hamdi Ulukaya: “We are entering a new era, when the center of gravity for social change has moved to the private sector. It’s business, not government, that is in the best position to lead today. It’s not government hiring refugees, it’s business. It’s not government cutting emissions, it’s business. It’s not government standing up to gun violence, it’s business. It’s not government that’s going to end inequality, or create opportunity. It’s business.”

13 October 2017

Posted on October 13, 2017 Leave a Comment

The IMF’s Fiscal Monitor 2017 is launched during the Annual Meetings of the IMF and the World Bank where the broader economic mood is positive with messages welcoming a global economic recovery. It is labelled “Tackling inequalities”. It shows that since the late 1980ies global inequality has declined while inequalities within countries have increased for half countries in the world. It unpacks 3 policy debates — progressive taxation, universal basic income (UBI) and public spending in health and education — without making explicit recommendations to policy makers. As UBI remains a hotly debated topic (that Horizons has summarized for you) and as the conversation around our house is growing on universal child grant, the report is worth browsing through.

In “The Red Cross presses Silicon Valley to fight cyberwarfare”, FastCompany interviews ICRC President Peter Maurer who met big techs to discuss responsibility and possible game changing role in humanitarian contexts. Arguing that governments with cyberwarfare capacities are not interested in regulation, he points to areas of convergence between the private sector and humanitarian organizations and suggests that collaborating with big techs could be a way to shape necessary norms and standards. Watch that space.

My graph this week is from Alvaredo and al “Measuring Inequality in the Middle East, 1990-2016: The World’s Most Unequal Region?” It combines household surveys, income tax data, wealth rankings and national accounts to estimate income concentration for 1990-2016 in Middle East countries. It then compares results with other parts of the world to show that the top income share in the Middle East is higher than in Western Europe, USA, Brazil and South Africa.

One week after unveiling the “Human capital project” aiming to create a human capital index for each country with the help of the Institute for Health Metrics and Evaluation (yes, same crowd who worked on the Gates Foundation SDG report), World Bank Group Jim Kim announces: “Soon, we’ll publish The Changing Wealth of Nations, and for the first time, we’re looking at human capital as part of the overall wealth of nations. It turns out that more than 65 percent of the wealth of all nations in the world is in human capital.”

19 May 2017

Posted on May 19, 2017 Leave a Comment

Delivered earlier this week during the Belt and Road Summit, Zheng Bijian’s “China’s ‘One Belt, One Road’ (OBOR) plan marks the next phase of globalization” presents OBOR as a global game changer. Indeed, this is the biggest infrastructure plan ever: 65 countries, 62% of the world population and 30% of global GDP. The total investment need is $5 trillion. China already pledged $113 billion, 9 of which in the form of aid to poor countries. More funding will come from the BRICS’s New Development Bank and the Asia Infrastructure Investment Bank. Unless you are in the Americas, chances are that what you do is somehow connected to this initiative. I guess, even if you are in the Americas, there will be externalities. Looking at power dynamics in the region, much of the talk focused on India’s lack of support. Next summit scheduled for 2019.

Danny Hernandez’ “How our company learned to do better predictions about everything” is what I shared when asked about best practices to improve organizational data literacy. Let me be clear here: I don’t know anything about best practices in that area. But I read this interesting article the day before I was asked the question. The company is Twitch, an Amazon offshoot. It invested in growing the predictive skills of every single staff member irrespective of functions and ranks. An initial phase warmed up employees by asking them to play with past company metrics, like “how much do you think our sales grew last quarter?” Once they got comfortable, they were asked to make predictions that would impact their work, like “how confident are you that you can complete this project in 3 weeks?” At first, Twitch employees were not excited: they did not believe in predictions, were afraid of making bad predictions, and thought that there was not enough evidence to make good predictions. But after the training, 96% of them said they would recommend it to a colleague. This initiative improved Twitch’s collective forecasting capacity. And it boosted efficiency as employees got better at evaluating ideas, defending funding proposals, and setting expectations. Other good ideas I heard this week: have data days in your teams or spend 10 minutes everyday on data issues. I am adding the latter to my daily routine.

CBinsights’ “From virtual nurses to drug discovery: 106 artificial intelligence startups in healthcare” maps the growing AI health market which is revolutionizing the whole industry from direct assistance to patients, to diagnosis, to drug discovery and mental health treatment.

My quote this week is from Elon Musk’s “The future we are building – and boring”: “I am not trying to be anyone’s savior. I just want to think about the future and not be sad.“

25 November 2015

Posted on November 25, 2015 Leave a Comment

Martin Ford’s “Rise of the Robots: Technology and the threat of a jobless future” just won the Financial Times and McKinsey Business Book of the Year award (last’s year winner was Piketty). I read this book while preparing the last issue of Horizons where we looked at the impact of artificial intelligence (AI) on the future of work (and education needs). This is a well written and compelling book with a lot of killer facts: “global shipments of robots increased by more than 60% between 2000 and 2012” (p. 3); “The percentage of news articles that will be written algorithmically within 15 years: over 90%” (p. 85); “42 % of researchers in human-level AI believe a thinking machine will arrive by 2030” (p. 231). Unlike other publications on the topic, it is not too technology-optimistic. It shows that the new phase of automation brought by AI threatens all job types including white-collar ones. Ford argues that new machine-augmented jobs will not compensate for job losses: “While human-machine collaboration jobs will certainly exist, they seem likely to be relatively few in number and often short-lived. They may also be un-rewarding or even dehumanizing” (p. 126). He also claims that investments in education and training, which helped absorb previous technological revolutions, will not be sufficient to adjust to the AI disruption. His core policy recommendation is the adoption of a basic income guarantee that he argues could be supported by liberals seeking to promote equity as well as pragmatic conservatives seeking an “insurance against adversity”. In short: a good read with a broadening fan-club.

Yang and Chou’s “Impacts of being downwind of a coal-fired power plant on infant health at birth” shows that pregnant women living downwind from a coal-fired power plant increase their likelihood of having low birth rate babies by 42%, and this is true when they live as far as 40 miles away from the plant. This is one of the latest scientific reminder of the highly concerning effects of air pollution on human health, and that of children in particular. Two months ago, a study estimated that outdoor air pollution coming from different sources (e.g., transportation, heating, dust, fires) was responsible for 3.3 million deaths every year globally, a number predicted to double by mid-century. This is to be added to the 3.5 million people dying every year from indoor air pollution. These alarming facts and prospects are for a large part affecting Asia where some of our colleagues, e.g., Mongolia CO, working with partners and leading experts are identifying high impact interventions to reduce the impact of air pollution on children’s health.

My interactive tool this week is Fengler’s “What is my place in the world population?” which gives you a vivid picture of where you (or anybody else you choose, like a child you are working for) fits in the global population dynamics: How many people are older/younger than you? When will you be the 5/6/7^th billion person in the world? When are you expected to die? If you lived in another country, how long will you be expected to live? I would lose 10 years if I were in Paraguay, 20 in Cote d’Ivoire, and gain 2 in Japan.

An important quote which has been attributed to many, including Benjamin Franklin, Cicero, Marie Curie, and Mark Twain, is in fact originally from Blaise Pascal’s Lettres Provinciales [1657]: “I have made this [letter] longer than usual because I have not had time to make it shorter.”