Predicting future AI failures from historic examples

foresight ◽  
2019 ◽  
Vol 21 (1) ◽  
pp. 138-152 ◽  
Author(s):  
Roman V. Yampolskiy
Keyword(s):  
Artificial Intelligence ◽  
Intelligent Systems ◽  
Design Methodology ◽  
Security System ◽  
Moderate Level ◽  
Catastrophic Event ◽  
Data Set ◽  
Content Type ◽  
Public Data ◽  
Ai Safety

Purpose The purpose of this paper is to explain to readers how intelligent systems can fail and how artificial intelligence (AI) safety is different from cybersecurity. The goal of cybersecurity is to reduce the number of successful attacks on the system; the goal of AI Safety is to make sure zero attacks succeed in bypassing the safety mechanisms. Unfortunately, such a level of performance is unachievable. Every security system will eventually fail; there is no such thing as a 100 per cent secure system. Design/methodology/approach AI Safety can be improved based on ideas developed by cybersecurity experts. For narrow AI Safety, failures are at the same, moderate level of criticality as in cybersecurity; however, for general AI, failures have a fundamentally different impact. A single failure of a superintelligent system may cause a catastrophic event without a chance for recovery. Findings In this paper, the authors present and analyze reported failures of artificially intelligent systems and extrapolate our analysis to future AIs. The authors suggest that both the frequency and the seriousness of future AI failures will steadily increase. Originality/value This is a first attempt to assemble a public data set of AI failures and is extremely valuable to AI Safety researchers.

Autonomous vehicles: a review of the ethical, social and economic implications of the AI revolution

2020 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Shweta Banerjee
Keyword(s):  
Artificial Intelligence ◽  
Autonomous Vehicles ◽  
Design Methodology ◽  
Technological Advancement ◽  
Economic Implications ◽  
Content Type ◽  
The Public ◽  
Autonomous Cars ◽  
The Subject ◽  
Practical Implications

PurposeThere are ethical, legal, social and economic arguments surrounding the subject of autonomous vehicles. This paper aims to discuss some of the arguments to communicate one of the current issues in the rising field of artificial intelligence.Design/methodology/approachMaking use of widely available literature that the author has read and summarised showcasing her viewpoints, the author shows that technology is progressing every day. Artificial intelligence and machine learning are at the forefront of technological advancement today. The manufacture and innovation of new machines have revolutionised our lives and resulted in a world where we are becoming increasingly dependent on artificial intelligence.FindingsTechnology might appear to be getting out of hand, but it can be effectively used to transform lives and convenience.Research limitations/implicationsFrom robotics to autonomous vehicles, countless technologies have and will continue to make the lives of individuals much easier. But, with these advancements also comes something called “future shock”.Practical implicationsFuture shock is the state of being unable to keep up with rapid social or technological change. As a result, the topic of artificial intelligence, and thus autonomous cars, is highly debated.Social implicationsThe study will be of interest to researchers, academics and the public in general. It will encourage further thinking.Originality/valueThis is an original piece of writing informed by reading several current pieces. The study has not been submitted elsewhere.

scholarly journals A surrogate-based generic classifier for Chinese TV series reviews

2017 ◽  
Vol 45 (2) ◽  
pp. 66-74
Author(s):  
Yufeng Ma ◽  
Long Xia ◽  
Wenqi Shen ◽  
Mi Zhou ◽  
Weiguo Fan
Keyword(s):  
Design Methodology ◽  
Experimental Result ◽  
Digital Information ◽  
Data Set ◽  
Content Type ◽  
Category Labels ◽  
Information Supply Chain ◽  
Feature Selection Techniques ◽  
Practical Implications

Purpose The purpose of this paper is automatic classification of TV series reviews based on generic categories. Design/methodology/approach What the authors mainly applied is using surrogate instead of specific roles or actors’ name in reviews to make reviews more generic. Besides, feature selection techniques and different kinds of classifiers are incorporated. Findings With roles’ and actors’ names replaced by generic tags, the experimental result showed that it can generalize well to agnostic TV series as compared with reviews keeping the original names. Research limitations/implications The model presented in this paper must be built on top of an already existed knowledge base like Baidu Encyclopedia. Such database takes lots of work. Practical implications Like in digital information supply chain, if reviews are part of the information to be transported or exchanged, then the model presented in this paper can help automatically identify individual review according to different requirements and help the information sharing. Originality/value One originality is that the authors proposed the surrogate-based approach to make reviews more generic. Besides, they also built a review data set of hot Chinese TV series, which includes eight generic category labels for each review.

Risk-based capital regulation revisited: evidence from the early 2000s

2015 ◽  
Vol 23 (2) ◽  
pp. 115-134 ◽  
Author(s):  
Thomas L. Hogan ◽  
Neil R. Meredith ◽  
Xuhao (Harry) Pan
Keyword(s):  
Design Methodology ◽  
Banking Regulation ◽  
Bank Performance ◽  
Capital Regulation ◽  
Data Set ◽  
Content Type ◽  
Future Performance ◽  
Capital Ratios ◽  
Using Data ◽  
Capital Regulations

Purpose – The purpose of this study is to replicate Avery and Berger’s (1991) analysis using data from 2001 through 2011. Although risk-based capital (RBC) regulation is a key component of US banking regulation, empirical evidence of the effectiveness of these regulations has been mixed. Among the first studies of RBC regulation, Avery and Berger (1991) provide evidence from data on US banks that new RBC regulations outperformed old capital regulations from 1982 through 1989. Design/methodology/approach – Using data from the Federal Reserve’s Call Reports, the authors compare banks’ capital ratios and RBC ratios to five measures of bank performance: income, standard deviation of income, non-performing loans, loan charge-offs and probability of failure. Findings – Consistent with Avery and Berger (1991), the authors find banks’ risk-weighted assets to be significant predictors of their future performance and that RBC ratios outperform regular capital ratios as predictors of risk. Originality/value – The study improves on Avery and Berger (1991) by using an updated data set from 2001 through 2011. The authors also discuss some potential limitations of this method of analysis.

Artificial intelligence-based automatic visual inspection system for built heritage

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Lukman E. Mansuri ◽  
D.A. Patel
Keyword(s):  
Artificial Intelligence ◽  
Visual Inspection ◽  
Image Data ◽  
Inspection System ◽  
Data Set ◽  
Web Based ◽  
Content Type ◽  
Built Heritage ◽  
Automatic Visual Inspection ◽  
Inspection Systems

PurposeHeritage is the latent part of a sustainable built environment. Conservation and preservation of heritage is one of the United Nations' (UN) sustainable development goals. Many social and natural factors seriously threaten heritage structures by deteriorating and damaging the original. Therefore, regular visual inspection of heritage structures is necessary for their conservation and preservation. Conventional inspection practice relies on manual inspection, which takes more time and human resources. The inspection system seeks an innovative approach that should be cheaper, faster, safer and less prone to human error than manual inspection. Therefore, this study aims to develop an automatic system of visual inspection for the built heritage.Design/methodology/approachThe artificial intelligence-based automatic defect detection system is developed using the faster R-CNN (faster region-based convolutional neural network) model of object detection to build an automatic visual inspection system. From the English and Dutch cemeteries of Surat (India), images of heritage structures were captured by digital camera to prepare the image data set. This image data set was used for training, validation and testing to develop the automatic defect detection model. While validating this model, its optimum detection accuracy is recorded as 91.58% to detect three types of defects: “spalling,” “exposed bricks” and “cracks.”FindingsThis study develops the model of automatic web-based visual inspection systems for the heritage structures using the faster R-CNN. Then it demonstrates detection of defects of spalling, exposed bricks and cracks existing in the heritage structures. Comparison of conventional (manual) and developed automatic inspection systems reveals that the developed automatic system requires less time and staff. Therefore, the routine inspection can be faster, cheaper, safer and more accurate than the conventional inspection method.Practical implicationsThe study presented here can improve inspecting the built heritages by reducing inspection time and cost, eliminating chances of human errors and accidents and having accurate and consistent information. This study attempts to ensure the sustainability of the built heritage.Originality/valueFor ensuring the sustainability of built heritage, this study presents the artificial intelligence-based methodology for the development of an automatic visual inspection system. The automatic web-based visual inspection system for the built heritage has not been reported in previous studies so far.

Fast and accurate detection of surface defect based on improved YOLOv4

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Jiawei Lian ◽  
Junhong He ◽  
Yun Niu ◽  
Tianze Wang
Keyword(s):  
Feature Extraction ◽  
Real Time ◽  
Surface Defect ◽  
Steel Ingot ◽  
Industrial Applications ◽  
Data Sets ◽  
Data Set ◽  
Processing Technologies ◽  
Content Type ◽  
Public Data

Purpose The current popular image processing technologies based on convolutional neural network have the characteristics of large computation, high storage cost and low accuracy for tiny defect detection, which is contrary to the high real-time and accuracy, limited computing resources and storage required by industrial applications. Therefore, an improved YOLOv4 named as YOLOv4-Defect is proposed aim to solve the above problems. Design/methodology/approach On the one hand, this study performs multi-dimensional compression processing on the feature extraction network of YOLOv4 to simplify the model and improve the feature extraction ability of the model through knowledge distillation. On the other hand, a prediction scale with more detailed receptive field is added to optimize the model structure, which can improve the detection performance for tiny defects. Findings The effectiveness of the method is verified by public data sets NEU-CLS and DAGM 2007, and the steel ingot data set collected in the actual industrial field. The experimental results demonstrated that the proposed YOLOv4-Defect method can greatly improve the recognition efficiency and accuracy and reduce the size and computation consumption of the model. Originality/value This paper proposed an improved YOLOv4 named as YOLOv4-Defect for the detection of surface defect, which is conducive to application in various industrial scenarios with limited storage and computing resources, and meets the requirements of high real-time and precision.

Intention to use analytical artificial intelligence (AI) in services – the effect of technology readiness and awareness

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Carlos Flavián ◽  
Alfredo Pérez-Rueda ◽  
Daniel Belanche ◽  
Luis V. Casaló
Keyword(s):  
Artificial Intelligence ◽  
Technology Adoption ◽  
Design Methodology ◽  
Technology Readiness ◽  
Intention To Use ◽  
Content Type ◽  
Financial Investment ◽  
Managerial Implications ◽  
Barriers To Technology

PurposeThe automation of services is rapidly growing, led by sectors such as banking and financial investment. The growing number of investments managed by artificial intelligence (AI) suggests that this technology-based service will become increasingly popular. This study examines how customers' technology readiness and service awareness affect their intention to use analytical AI investment services.Design/methodology/approachThe automation of services is rapidly growing, led by sectors such as banking and financial investment. The growing number of investments managed by AI suggests that this technology-based service will become increasingly popular. This study examines how customers' technology readiness and service awareness affect their intention to use analytical AI investment services.FindingsThe results indicated that customers' technological optimism increases, and insecurity decreases, their intention to use robo-advisors. Surprisingly, feelings of technological discomfort positively influenced robo-advisor adoption. This interesting finding challenges previous insights into technology adoption and value co-creation as analytical AI puts customers into a very passive role and reduces barriers to technology adoption. The research also analyzes how consumers become aware of robo-advisors, and how this influences their acceptance.Originality/valueThis is the first study to analyze the role of customers' technology readiness in the adoption of analytical AI. The authors link the findings to previous technology adoption and automated services' literature and provide specific managerial implications and avenues for further research.

scholarly journals Subnational institutions and open innovation: evidence from China

2017 ◽  
Vol 55 (9) ◽  
pp. 1942-1955 ◽  
Author(s):  
Fei Sun ◽  
Junjie Hong ◽  
Xiuying Ma ◽  
Chengqi Wang
Keyword(s):  
Open Innovation ◽  
Design Methodology ◽  
Emerging Market ◽  
Market Development ◽  
Data Set ◽  
Content Type ◽  
Complementary Effect ◽  
Performance Consequences ◽  
Prior Literature ◽  
External Actors

Purpose The purpose of this paper is to examine how subnational institutions within a country explain the performance consequences of open innovation (OI) in emerging market enterprises (EMEs). Design/methodology/approach The paper conducts a regression analysis by using a novel panel data set comprising of 438 innovative Chinese firms over the period of 2008-2011. Findings The authors show that although on average openness to external actors improves innovation performance this effect is pronounced for EMEs that operate in subnational regions with a higher level of intellectual property rights (IPR) enforcement and of factor market development. The findings point to the context-dependent nature of OI strategy and the complementary effect of institutional parameters in emerging markets and help to reconcile the contrasting findings regarding the effect of OI in the prior literature. Originality/value This paper extends the literature on OI by suggesting that the analysis of the performance consequences of OI strategy should go beyond the nexus between OI and firm performance, and instead, focus on subnational-specific institutions, such as region-specific IPR enforcement, factor market development and intermediation market development, that may facilitate or constrain the effect of OI model.

How firms borrow in international bond markets

2018 ◽  
Vol 26 (1) ◽  
pp. 135-169
Author(s):  
Alberto Fuertes ◽  
Jose María Serena
Keyword(s):  
Emerging Economies ◽  
Design Methodology ◽  
Choice Behavior ◽  
Global Financial Crisis ◽  
Securities And Exchange Commission ◽  
Bond Markets ◽  
Data Set ◽  
Content Type ◽  
Parametric Analyses ◽  
The Global Financial Crisis

Purpose This paper aims to investigate how firms from emerging economies choose among different international bond markets: global, US144A and Eurobond markets. The authors explore if the ranking in regulatory stringency –global bonds have the most stringent regulations and Eurobonds have the most lenient regulations – leads to a segmentation of borrowers. Design/methodology/approach The authors use a novel data set from emerging economy firms, treating them as consolidated entities. The authors also obtain descriptive evidence and perform univariate non-parametric analyses, conditional and multinomial logit analyses to study firms’ marginal debt choice decisions. Findings The authors show that firms with poorer credit quality, less ability to absorb flotation costs and more informational asymmetries issue debt in US144A and Eurobond markets. On the contrary, firms issuing global bonds – subject to full Securities and Exchange Commission requirements – are financially sounder and larger. This exercise also shows that following the global crisis, firms from emerging economies are more likely to tap less regulated debt markets. Originality/value This is, to the authors’ knowledge, the first study that examines if the ranking in stringency of regulation – global bonds have the most stringent regulations and Eurobonds have the most lenient regulations – is consistent with an ordinal choice by firms. The authors also explore if this ranking is monotonic in all determinants or there are firm-specific features which make firms unlikely to borrow in a given market. Finally, the authors analyze if there are any changes in the debt-choice behavior of firms after the global financial crisis.

Students' perceptions of education for sustainable development

2014 ◽  
Vol 15 (2) ◽  
pp. 242-253 ◽  
Author(s):  
Yvonne Zeegers ◽  
Ian Francis Clark
Keyword(s):  
Design Methodology ◽  
Pedagogical Approach ◽  
Data Set ◽  
Content Type ◽  
Tertiary Institutions ◽  
Reflective Journal ◽  
Reflective Journals ◽  
The Social ◽  
Perceptions Of Education ◽  
Practical Implications

Purpose – This study investigated whether a course which focused on raising students' awareness of sustainability, from a balanced perspective, that is, one which gives equal consideration to the social and economic aspects as well as the environmental would produce graduates with the knowledge and commitment required to drive the sustainability agenda forward. The paper aims to discuss these issues. Design/methodology/approach – An analysis of students' final entries in their reflective journal was used to explore whether their views on sustainability reflected a balanced view. Findings – The findings of this research confirmed previous studies showed that initially students do have an enviro-centric bias. It also showed that despite experiencing a pedagogical approach which challenged views by encouraging discussion, debate, and reflection and which provided what was considered to be a balanced view of sustainability, many of the students still leaned towards an environmentally focused perspective of sustainability. Research limitations/implications – The conclusions are based on one data set but are supported by other data described in the paper. Practical implications – The finding led the authors to conclude that a concerted holistic effort within and across courses is needed within tertiary institutions if students' views about sustainability are to be challenged. Originality/value – The outcomes demonstrate that students' reflective journals can be used to gather information about the change in students' perceptions about sustainability.

Software documentation is not enough! Requirements for the documentation of AI

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Florian Königstorfer ◽  
Stefan Thalmann
Keyword(s):  
Artificial Intelligence ◽  
Business Process ◽  
Design Methodology ◽  
Governance Mechanism ◽  
It Industry ◽  
Specific Nature ◽  
Content Type ◽  
Software Documentation ◽  
Disruptive Technologies ◽  
Present Information

Purpose Artificial intelligence (AI) is currently one of the most disruptive technologies and can be applied in many different use cases. However, applying AI in regulated environments is challenging, as it is currently not clear how to achieve and assess the fairness, accountability and transparency (FAT) of AI. Documentation is one promising governance mechanism to ensure that AI is FAT when it is applied in practice. However, due to the nature of AI, documentation standards from software engineering are not suitable to collect the required evidence. Even though FAT AI is called for by lawmakers, academics and practitioners, suitable guidelines on how to document AI are not available. This interview study aims to investigate the requirements for AI documentations. Design/methodology/approach A total of 16 interviews were conducted with senior employees from companies in the banking and IT industry as well as with consultants. The interviews were then analyzed using an informed-inductive coding approach. Findings The authors found five requirements for AI documentation, taking the specific nature of AI into account. The interviews show that documenting AI is not a purely technical task, but also requires engineers to present information on how the AI is understandably integrated into the business process. Originality/value This paper benefits from the unique insights of senior employees into the documentation of AI.

Sign in / Sign up

Export Citation Format

Share Document