人工智能伦理：政府的下一件大事

The World Government Summit is a global platform dedicated toshaping the future of governments worldwide. Each year, the Summitsets the agenda for the next generation of governments with a focuson how they can harness innovation and technology to solve universalchallenges facing humanity.The World Government Summit is a knowledge exchange center at theintersection of government, futurism, technology, and innovation. Itfunctions as a thought leadership platform and networking hub forpolicymakers, experts and pioneers in human development.The Summit is a gateway to the future as it functions as the stage foranalysis of future trends, concerns, and opportunities facing humanity.It is also an arena to showcase innovations, best practice, and smartsolutions to inspire creativity to tackle these future challenges.AnsweringTomorrow’sQuestionsToday2 Table of ContentSetting the Context - Ethics at the heart of AIHow should governments respond?Shaping the governance guidelines for AIIs a global AI Ethics framework the solution?Topics 07141622 Despite remarkableachievements, the rapiddevelopment of AI hasraised a host of ethicalconcerns. Governmentsface challenges andchoices pertaining to howto apply AI technologiesin the public sector and ingovernance strategies. According to Deloitte’s 2018 Global Human CapitalTrends report, 42 percent of surveyed executivesexpect Artificial Intelligence (AI) - intelligentmachines that imitate human behavior - may bewidely deployed in their organizations in the nextthree to five years. The public sector follows the samepath seeking and adopting applications to improvepublic services and manage the growing difficulty ofanalysis and decision making by effectively exploitingincreasingly available amounts of information.Through cognitive applications, AI already helpsgovernments reduce backlogs and cut costs, predictfraudulent transactions and identify criminal suspectsvia facial recognition. By adopting AI for automation,governments can focus on more creative and complexaspects of service delivery to citizens.Despite remarkable achievements, the rapiddevelopment of AI has raised some concerns being asubject of fear and skepticism in the media. Could AIlong-term development lead to the end of humankindas Elon Musk, Bill Gates and numerous technologistshave speculated? What is the role of ethics in thedesign, development and application of AI? How willethics help maximize the benefits of AI to increasecitizen well-being and common good?Setting the Context –Ethics at the heart of AIArtificial intelligence looks to make all of governmentmore efficient by automating and improving routinetasks allowing public sector employees to spend fewerhours on noncore tasks and more on innovation. Forinstance, finance ministries in the Gulf region are usingmachine learning for detecting complicated fraudscenarios; public welfare organizations are employingmachine learning to disburse welfare payments ina more efficient and equitable manner; numerouspublic-facing entities have deployed Chatbots tointeract with citizens; health organizations areemploying AI for triaging health care cases; and policeagencies are using facial recognition AI for improvingits surveillance capabilities. In the 2017 Deloitte reportAI-augmented government, our analysis of the USPublic Sector found cognitive technologies could saveup to 1.2 billion hours and potential annual savings of$41.1 billion1as indicated in figure 1 below.AI- augmented governmentFIGURE . HOW MUCH SAVINGS CAN AI IN GOVERNMENT GENERATE?High in investment(Tasks speed up by 200%)(Tasks speed up by 20%)Hours freedPotential savingsLow investmentSource : Deloitte analysis1.2 billion hours$41.1 billion$3.3 billion96.7 million hours Person-hours peryear for taskTime 1Time 27Deloitte University Press | dupress.deloitte.comO*NET program has been surveyingworkers on how much time isdevoted to each taskObserving the same tasks at twodifferent points in time showschanges in labor allocated tothat task 5 While there is an increasing interaction between AItechnologies and our socio-political and economicinstitutions, consequences are not well defined. Theadvent of AI raises a host of ethical issues, relatedto moral, legal, economic and social aspects of oursocieties and government officials face challenges andchoices pertaining to how to apply AI technologies inthe public sector and in governance strategies. FromUber’s self-driving car fatality to Amazon’s genderbiased recruitment tool, examples of AI ethicalconcerns abound and reinforce the idea that theyshould be taken into account before an AI system isdeployed. In this perspective, “ethics” can be definedby the pursuit of “good” actions based on “good”decision-making —decisions and actions that leadto the least possible amount of unnecessary harmor suffering.2It implies that our government andbusiness leaders understand and define what “good”means for AI systems. Gaining societal consensuson the ethics of AI i