Articles on AI and ethics/society
From The Artificial Intelligence and Society discussion group
Media reports and opinions about future AI
- Nature editorial on AI (April 2016): Anticipating Artificial Intelligence
- Economist report on robots (March 2014): Immigrants from the future
Academic reports about future AI
- Mueller and Bostrom (2016): Future Progress in Artificial Intelligence: A Survey of Expert Opinion.
Business / economic reports about AI
- McKinsey Quarterly's report on AI and potential job losses: Four fundamentals of workplace automation (Nov 2015)
- The World Economic Forum's report on AI: The Fourth Industrial Revolution: what it means, how to respond (Jan 2016)
- Swiss bank UBS's report on implications of the fourth industrial revolution: Extreme Automation and Connectivity (Jan 2016). Here's a Guardian article about the report.
- Deloitte's State of the State report for 2016/17 predicts that 861,000 UK public sector jobs could be lost to automation by 2030 (see pp16-17). (Frey and Osborne contributed to the report.) Here's a Guardian article about the report.
Academic discussions about the likely impact of AI on jobs
- Frey and Osborne's 2013 paper estimating 47% of US jobs are 'highly automatable': The future of employment: how susceptible are jobs to computerisation?
- Arntz et al.'s 2016 OECD report estimating 9% of OECD jobs are 'automatable': The Risk of Automation for Jobs in OECD Countries: a comparative analysis
- Discussion of Frey and Osborne's paper in the Guardian: The knowledge economy is a myth
- Discussion of Arntz et al.'s paper: robotenomics.com: Robots and job fears: Destruction of large numbers of jobs unlikely, says new OECD Study
- David Autor's (2015) paper arguing against predictions about large-scale technology-driven job losses: Why Are There Still So Many Jobs? The History and Future of Workplace Automation
- John Danaher's (2015) response to Autor: Why haven't robots taken our jobs? The Complementarity Effect.
- Another piece by Danaher (2015), with a discussion of (and support for) Autor's arguments for a 'polarising' effect of technology on jobs: Automation and Income Inequality: Understanding the Polarisation Effect
- Frey and Osborne's 2015 paper on societal/economic impacts of new technologies: Technology at work: The future of automation and employment.
- The UN International Labour Organisation's 2016 report on the impact of automation on jobs in South-East Asia: ASEAN in transformation. This report estimates that 56% of the total workforce of ASEAN countries are at risk of displacement by robots.
- A study from the US National Bureau of Economic Research (Acemoglu and Restrepo, Working Paper No. 23285, 2017), Robots and Jobs: Evidence from US Labor markets. This reports a regression analysis that looks at the influence of robots on unemployment levels and wages. In the US areas studied, the authors estimate that 'one more robot per thousand workers reduces the employment to population ratio by about 0.18-0.34 percentage points and wages by 0.25-0.5 percent' - and that these influences are distinct from the impact of imports from China and Mexico, the decline of routine jobs, offshoring, and several other factors. (Here's an article on the study in the New York Times.)
Some links from Murat Ungor on AI and employment:
- National Academies of Sciences, Engineering, and Medicine. Information Technology and the U.S. Workforce: Where Are We and Where Do We Go from Here? (National Academies Press, 2017).
- Track how technology is transforming work by Tom Mitchell and Erik Brynjolfsson, published on April 13, 2017
- Why “How many jobs will be killed by AI?” is the wrong question by Andrew McAfee and Erik Brynjolfsson, published on June 24, 2017
- Society ‘flying blind’ over robots’ impact on jobs is the wrong question by Richard Waters, published on April 14, 2017
- The rise of the algorithm need not be bad news for humans by Luciano Floridi, published on May 5, 2017
- Imagine if Robo advisers could do emotions by Andrew W. Lo, published on June 6, 2016
- https://www.project-syndicate.org/commentary/mnuchin-automation-low-skill-workers-by-j--bradford-delong-2017-04 Artificial intelligence and artificial problems] by J. Bradford Delong, published on April 3, 2017
- http://voxeu.org/article/robots-and-jobs-evidence-us Robots and jobs: Evidence from the US] by Daron Acemoglu and Pascual Restrepo, published on April 10, 2017
- The Economist in your laptop by Casey Mulligan, posted on February 22, 2017
- Jobs in the age of AI by Simon Johnson and Jonathan Ruane, published on May 30, 2017
- The skills that matter in the race between education and technology by Harry A. Patrinos, published on 03/01/2017
- The race between machine and man: Implications of technology for growth, factor shares, and employment by Daron Acemoglu, published on June 21, 2017. (Here's a talk by Acemoglu on the same topic.)
- Financial advice by robots option could be fast-tracked by Sally Peart, Otago Daily Times, June 26, 2017
- Estimating the impacts of robots on productivity and employment by Guy Michaels and Georg Graetz, published on March 18, 2015
Articles relating to regulation of AI
- Wachter et al.'s 2016 report arguing that current EU legislation providing a 'right to explanation' of automated decision mechanisms is insufficient, and calling for a watchdog to assess whether such decisions are discriminatory: Why a Right to Explanation of Automated Decision-Making Does Not Exist in the General Data Protection Regulation. (Alan Turing Institute / University of Oxford)
Articles relating to bias / discrimination in AI
Some articles about a study from Princeton/Bath demonstrating many biases in language corpora taken from the web:
- The original study: Semantics derived automatically from language corpora contain human-like biases (Caliskan, Bryson and Narayanan, April 2017).
- A Guardian article summarising the study.
- Bias in Criminal Risk Scores Is Mathematically Inevitable, Researchers Say (Angwin and Larson, ProPublica 2016).
- There's more detail on the data analysis for this study here.
- A Washington Post blog article about the ProPublica study, by Sam Corbett-Davies, Emma Pierson, Avi Feller and Sharad Goel.
- A sample of COMPAS's risk assessment questionnaire (unconfirmed).
- A primer on Cox regression, the statistical paradigm used by COMPAS (with worked examples in R), by John Fox and Sanford Weisberg (Sage Publishing 2011)
Articles on uses of AI in criminal justice:
- Jason Tashea: Courts are using AI to sentence criminals. That must stop now. (Wired, April 2017)
- An interesting symposium organised by Jason Tashea: Justice Codes (October 2016)
- A useful review article on Algorithms in the Criminal Justice System (including a good bibliography of academic articles) by the Electronic Privacy Information Center - an appendix to their book 'An R Companion to Applied Regression' (2011)
- Goel, Rao and Shroff's paper about a proposed method for reducing police bias in stop-and-search events: Personalized Risk Assessments in the Criminal Justice System (American Economic Review, 2016)
Discussions of the impact on people of AI
- Anne Amnesia's blog post on the Unnecessariat (May 2016)
AI and legal issues
- An article about possible EU legislation to classify advanced robots as 'electronic persons' (June 2016)
- A white paper from Chapman Tripp, in association with the NZ Institute of Directors: Artificial Intelligence Opportunities and challenges for New Zealand: A call to action (October 2016)
Recent initiatives relating to the future of AI
- Wired magazine report on Elon Musk's new 'Open AI' company (May 2016): Inside OpenAI, Elon Musk's wild plan to set Artificial Intelligence free