2
BCS, The Chartered Institute for IT response to the
House of Lords Select Committee on Artificial Intelligence - Call for Evidence
1st September 2017
BCS, The Chartered Institute for IT
BCS is a charity with a Royal Charter. Its mission is to make IT better for society. It does this through leadership on societal and professional issues, working with communities and promoting excellence.
BCS brings together industry, academics, practitioners, educators and government to share knowledge, promote new thinking, educate, shape public policy and inform the public. This is achieved through and with a network of 75,000 members across the UK and internationally. BCS is funded through membership fees, through the delivery of a range of professional development tools for practitioners and employers, and as a leading IT qualification body, through a range of widely recognised professional and end-user qualifications.
www.bcs.org
BCS, The Chartered Institute for IT response to House of Lords Select Committee on Artificial Intelligence - Call for Evidence
7
The pace of technological change
Question 1. What is the current state of artificial intelligence and what factors have contributed to this? How is it likely to develop over the next 5, 10 and 20 years? What factors, technical or societal, will accelerate or hinder this development?
1. The current state of artificial intelligence is that one specific set of technologies, “machine learning”, which has seen some impressive practical advances recently, is dominating, to the point where the two are practically synonymous in the public eye, and also where it seems to be difficult to secure funding for other forms of artificial intelligence research. As regards the challenges, and potential dangers, of machine learning, we refer to the Institute of Mathematics and its Applications evidence to this inquiry, and to the joint BCS/IMA evidence base at: https://ima.org.uk/6910/the-debate-about-algorithms/ . We also note the IMA’s comments on how easy it seems to be to subvert the current machine learning systems, and feel that the risks are insufficiently appreciated, even by practitioners. We note the striking example they give, quoted from [Evtimov et al., 2017 1] of what clearly looks like a STOP sign, to human beings, but has been tampered with to appear like a 45 speed limit to a machine learning recogniser.
2. Stop sign to a human being, 45 speed limit to a machine [Evtimov et al., 2017 1].
3. The committee, and indeed the public, needs to differentiate between Specific and General AI. Specific AI is where the intelligence is used to deal with a specific task, such as a medical diagnosis. In this case it seeks answers to pre-programme
BCS, The Chartered Institute for IT response to House of Lords Select Committee on Artificial Intelligence - Call for Evidence
1. Evtimov et al., 2017] Evtimov,I., Eykholt,K., Fernandes,E., Kohno,T., Li,B., Prakash,A., Rahmati,A. & Song,D., Robust Physical-World Attacks on Machine Learning Models. https://arxiv.org/abs/1707.08945.
7
questions, but is incapable of forming new questions. General AI has the ability to ask new questions and is akin to human consciousness. Because we do not understand how human consciousness develops this is
4. Some way off (depending on what is meant by ‘akin’), but will eventually raise ethical issues such as is the AI ‘device’ alive in the human sense. BCS wishes to caution society about the dangerous nature of trying to deploy analogues without full understanding of the ethical and sociological implications. A good example of the weaknesses of current general AI is shown by the Chinese insistence that two chatbots be taken offline for political incorrectness: http://www.reuters.com/article/us-china-robots-idUSKBN1AK0G1 . A more humorous example of the difficulty of producing even vaguely plausible General AI is given in https://www.livescience.com/60275-ai-writes-next-game-of-thrones-novel.html where the AI system produces text such as “Varys poisoned Daenerys and another of the dead men”.
5. Specific AI relies on human input to develop the questions, generally to identify “training data” and “testing data” which characterise the question, analyse the answers and define the action to be taken. Specific AI has the potential advantages of consistency, speed and not getting tired. Like other computer programs (because that is what it is), Specific AI can operate in real-time 24 x 7. It is likely to replace, or change significantly, many human jobs in the near (10 years) future which raises many questions about what roles humans will have in the future.
Question 2. Is the current level of excitement which surrounds artificial intelligence warranted?
1. A great deal of what is said in the popular media is, of course, wrong, utopian or alarmist. But nevertheless society is right to be deeply interested in major technological changes which have already affected (for a small example, see http://metro.co.uk/2016/11/09/little-girl-uses-google-translate-to-invite-her-lonely-new-classmate-to-lunch-6246363/), and will undoubtedly affect far more, nearly everyone’s lives.
2. Some of the developments in artificial intelligence, notably the developments in automated reasoning which allow the production and fielding of fault-free software in critical applications such as air traffic control, jet engine operation and medical devices (see IMA submission) are wholly beneficial. Others, notably those clustered around “machine learning” (which is better described as “machine pattern recognition”), are more mixed. If we look back at the history of the automobile, we see that we should avoid both the equivalent of the Locomotive Act 1865 (the “Red Flag Act”) and the equivalent of the period before driving tests.
BCS, The Chartered Institute for IT response to House of Lords Select Committee on Artificial Intelligence - Call for Evidence
7
Impact on society
Question 3. How can the general public best be prepared for more widespread use of artificial intelligence?
BCS has no response to make to this question.
Question 4. Who in society is gaining the most from the development and use of artificial intelligence and data? Who is gaining the least? How can potential disparities be mitigated?
1. The impact of AI and Machine Learning is fuelling another generation of automation and is becoming an ever-stronger factor in the “appification” of routine tasks, formerly done by the less skilled, on smartphones and tablets. Map software that anticipates your destination, navigates you there and offers you places to visit or stay is replacing work previously performed by personal assistants, travel agencies, and even friends. Tasks such as these and others in offices and factory floors that are routine or repetitive will be the first to be taken over by machines. Tasks that are complex are more likely, at least for the foreseeable future, to be assisted, rather than replaced, by machine intelligence. Most of the one fifth of activities in the U.S. workplace that McKinsey [2016 2] has identified as highly susceptible to automation are performed by low skilled or unskilled workers. Already changes due, at least partly, to technology are visible. The share of U.S. workers employed in routine office jobs declined from 25.5% to 21% between 1996 and 2015 [Economist, 2017 3]. Far from being threatened by AI and Machine Learning highly skilled professionals such as doctors, lawyers and accountants will significantly benefit. Services in areas such as healthcare, law, and financial advice, are usually in high demand and can be made more affordable through AI. Professionals will be able to do more with less [e.g. Cohen, 2016 4] But in areas like retail and services, jobs are already being replaced by machines [e.g. Pierce, 2017 5].
2. The implications of these developments for policymakers are stark. First, policymakers need to understand that “unemployment caused by technology” is simply another way of saying “unemployment caused by a lack of skills”; second, when workers’ skills fall behind then inequality follows [Economist, 2017] In a highly educated and highly skilled society there is no reason for anyone to suffer unemployment due to displacement by technology for more than a short period. Other than managing worker displacement, for which Denmark’s flexicurity system
BCS, The Chartered Institute for IT response to House of Lords Select Committee on Artificial Intelligence - Call for Evidence
2. [McKinsey, 2016] McKinsey, ‘Where machines could replace humans—and where they can’t (yet)’, July, 2016, See: http://www.mckinsey.com/business-functions/digital-mckinsey/our-insights/where-machines-could-replace-humans-and-where-they-cant-yet
3. [Economist, 2017] Lifelong learning: special report, Economist, January 14, 2017
4. [Cohen, 2016] Mark A. Cohen, 'How Artificial Intelligence Will Transform The Delivery Of Legal Services', Forbes, September 6, 2016, http://www.forbes.com/sites/markcohen1/2016/09/06/artificial-intelligence-and-legal-delivery/20ebaa842647
5. [Pierce, 2017] David Pierce, 'This Robot Makes a Dang Good Latte', Wired, January 30, 2017, https://www.wired.com/2017/01/cafe-x-robot-barista/
7
3. may offer a model, the best policy responses are likely to include measures for teaching people how to learn new skills. Workers who have learned how to learn are likely to need this skill throughout their careers
Public perception
Question 5. Should efforts be made to improve the public’s understanding of, and engagement with, artificial intelligence? If so, how?
BCS has no response to make to this question.
Industry
Question 6. What are the key sectors that stand to benefit from the development and use of artificial intelligence? Which sectors do not?
1. It is very hard to say that a given sector cannot benefit from Artificial Intelligence. Take the thousands of people building and maintaining the Rolls Royce Trent 10000 engines, the only engines for the Boeing 787, Airbus A350-1000 and A330 neo. Their jobs are there only because of the small (less than 100) team who applied Automated Reasoning tools to verify the avionics and therefore give it that market edge. They themselves are only effective because of the team of about 10 who developed that methodology at Altran.
2. The barriers to adopting this sort of technology are largely demand-side ignorance, and supply-side lack of trained staff.
Question 7. How can the data-based monopolies of some large corporations, and the ‘winner-takes-all’ economies associated with them, be addressed? How can data be managed and safeguarded to ensure it contributes to the public good and a well-functioning economy?
BCS has no response to make to this question.
Ethics
Question 8. What are the ethical implications of the development and use of artificial intelligence? How can any negative implications be resolved?
BCS has no response to make to this question.
BCS, The Chartered Institute for IT response to House of Lords Select Committee on Artificial Intelligence - Call for Evidence
7
Question 9. In what situations is a relative lack of transparency in artificial intelligence systems (so-called ‘black boxing’) acceptable? When should it not be permissible?
BCS has no response to make to this question.
The role of the Government
Question 10. What role should the Government take in the development and use of artificial intelligence in the United Kingdom? Should artificial intelligence be regulated? If so, how?
1. The EU’s General Data Protection Regulation, due to enter in force in May 2018, talks, in Article 22(1), about “a decision based solely on automated processing, including profiling, which produces legal effects concerning him or her or similarly significantly affects him or her.” At this point, the Regulation says (Article 14(2)(g), see also 15(1)(h)) that the subject must be “provided with meaningful information about the logic involved, as well as the significance and the envisaged consequences of such processing for the data subject.”.
2. While this seems plausible as a statement of general principles, it is extremely vague. Leaving it for clarification by case law seems unhelpful to the many practitioners. The meaning of “similarly significantly affects” is unclear. One significant question is whether this includes short-listing for a job.
Learning from others
Question 11. What lessons can be learnt from other countries or international organisations (e.g. the European Union, the World Economic Forum) in their policy approach to artificial intelligence?
1. In Answer 10 we noted that the EU’s General Data Protection Regulation talking about “a decision based solely on automated processing, including profiling, which produces legal effects concerning him or her or similarly significantly affects him or her.” As we said, this is plausibly a correct statement of general principles, it is extremely vague. The U.K. should learn from this vagueness, and seek, whether by legislation or by much firmer guidance from the Information Commissioners Office to give greater clarity to
2. “significantly affects” – this almost certainly includes the length of prison sentences, probably affects the granting of major loans (it does in the US: Fair Credit Act), but it is currently unclear whether this includes short-listing for jobs, university entrance etc.;
3. “meaningful information about the logic involved” – presumably this is more than “our deep neural net says you shouldn’t be shortlisted”, but how much more?
4. “based solely on” – there has been much concern in the US recently about “robo-signers”, humans who sign papers prepared by computers, and do not effectively review them, and while one would hope that robo-signers were not sufficient to bypass “based solely on”, the situation is not clear.
END
BCS, The Chartered Institute for IT response to House of Lords Select Committee on Artificial Intelligence - Call for Evidence