Artificial intelligence for sustainability: Challenges, opportunities, and a research agenda

Probably not a substantial claim: may not pass S3 (load-bearing).

Computer vision and natural language processing are emerging themes related to AI's capability to mimic human cognition.

“Computer vision (CV) and natural language processing (NLP) are emerging themes related to AI's capability to mimic human cognition.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Genetic algorithms are the most popular instance of evolutionary computation.

“Genetic algorithms (GA) are the most popular instance of evolutionary computation (EC) (Drugan, 2019).”

• cites Drugan, 2019Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Studies on AI increased significantly from 2015 to 2019.

“when studies on AI increased significantly (Lu, 2019).”

• cites Lu, 2019Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based). May combine two claims.

Biodiversity, water, energy, and transportation are rich areas of investigation for AI.

“Biodiversity, water, energy, and transportation are rich areas of investigation for AI.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The literature on AI for sustainability extends across many disciplines and domains and addresses environmental challenges.

“The literature on AI for sustainability extends across many disciplines and domains and addresses environmental challenges.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

Water systems research has predominantly developed and tested ML models using techniques from linear regression, GA, ANN, SVM, ARMA, and ANFIS.

“Water systems research has predominantly developed and tested ML models using techniques from linear regression, genetic algorithm (GA), artificial neural network (ANN), support vector machine (SVM), autoregressive moving average model (ARMA), and the adaptive neuro-fuzzy inference system or network-based fuzzy inference system (ANFIS).”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

ML models are used to predict stream flow and examine water quality parameters.

“ML models are used to predict stream flow and examine water quality parameters (Demirci et al., 2019).”

• cites Demirci et al., 2019 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S4 (based).

Theory development regarding the use of AI for sustainability is scant.

“Theory development regarding the use of AI for sustainability is scant”

Probably not a substantial claim: may not pass S3 (load-bearing).

Evolutionary computation is popularly applied to optimize ML models.

“EC is also popularly applied to optimize ML models (Al-Sahaf et al., 2019).”

• cites Al-Sahaf et al., 2019Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Energy use of buildings, energy production, energy distribution, operations and maintenance, and public opinion for renewable energy have been core areas of interest in energy research.

“Energy use of buildings, energy production, energy distribution, operations and maintenance, and public opinion for renewable energy have been core areas of interest in this domain (González Ordiano et al., 2018; Merizalde et al., 2019; Wang and Srinivasan, 2017).”

• cites González Ordiano et al., 2018Not checked yet

  Not checked in the free run.

• cites Merizalde et al., 2019Not checked yet

  Not checked in the free run.

• cites Wang and Srinivasan, 2017Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S4 (based).

Research to date has been primarily restricted to technical and design-oriented engineering disciplines that emphasize problem solving and technical solutions.

“Research to date has been primarily restricted to technical and design-oriented engineering disciplines that emphasize problem solving and technical solutions.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Major CV applications include feature extraction in images and videos, object identification and classification, and perception of the surrounding environment as used in autonomous vehicles.

“Major CV applications include feature extraction in images and videos (Buch et al., 2011), object identification and classification, such as in face recognition (Kumar et al., 2019) and perception of the surrounding environment as used in autonomous vehicles (Elliott et al., 2019).”

• cites Buch et al., 2011Not checked yet

  Not checked in the free run.

• cites Kumar et al., 2019Not checked yet

  Not checked in the free run.

• cites Elliott et al., 2019Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

AI for sustainability research strongly focuses on ML models and algorithms to show how machines can analyze and learn from data.

“AI for sustainability research strongly focuses on ML models and algorithms to show how machines can analyze and learn from data.”

Probably not a substantial claim: may not pass S3 (load-bearing).

In energy research, ANN, expert systems, pattern recognition, and fuzzy logic models are the main focus areas.

“In energy research, ANN, expert systems, pattern recognition, and fuzzy logic models are the main focus areas (Demirci et al., 2019; Tyralis et al., 2019).”

• cites Demirci et al., 2019 sourceNot checked yet

  Not checked in the free run.

• cites Tyralis et al., 2019Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Much biodiversity research has developed ML or NLP solutions to predict ecosystem services, model social media data to predict public opinion, and model habitat suitability and availability for biodiversity conservation.

“Similarly, much biodiversity research has developed ML or natural language processing (NLP) solutions to predict ecosystem services, has modeled social media data to predict public opinion, and has modeled habitat suitability and availability for biodiversity conservation (Bagstad et al., 2016; Tuuli et al., 2019).”

• cites Bagstad et al., 2016Not checked yet

  Not checked in the free run.

• cites Tuuli et al., 2019Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Deep learning is gaining traction as a type of ANN with more layers of units.

“Deep learning is gaining traction as a type of ANN with more layers of units (Shrestha and Mahmood, 2019).”

• cites Shrestha and Mahmood, 2019 sourceNot checked yet

  Not checked in the free run.

Classified as probably not a substantial claim.

There has been relative silence on how to explore the specific context of AI for environmental sustainability in IS reviews and opinions.

“there has been relative silence on how to explore the specific context of AI for environmental sustainability.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The final theme identified in the literature was AI robotics.

“The final theme we identified in the literature was AI robotics.”

May combine two claims.

NC algorithms are used to solve multiobjective problems or generate optimal model parameters in energy.

“NC algorithms are used to solve multiobjective problems or generate optimal model parameters (Li et al., 2018).”

• cites Li et al., 2018Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Solar is the most popularly studied renewable energy source.

“Solar is the most popularly studied renewable energy source.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Although AI has not yet achieved human-level general intelligence, it has expanded broadly to include strategic planning, sequential actions, self-learning, and self-improvement.

“Although AI has not yet achieved human-level general intelligence, it has expanded broadly to include strategic planning, sequential actions (Ghallab et al., 2016), self-learning, and selfimprovement (Omohundro, 2007).”

• cites Ghallab et al., 2016Not checked yet

  Not checked in the free run.

• cites Omohundro, 2007Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S4 (based).

The literature commonly takes a reductionist approach to deal with each element of sustainability independently.

“The literature commonly takes a reductionist approach to deal with each element independently.”

May combine two claims.

Only one article discussed public opinion about energy production and none examined financial feasibility of market deployment.

“Only one article discussed public opinion about energy production (Nuortimo and Härkönen, 2018) and none examined financial feasibility of market deployment.”

• cites Nuortimo and Härkönen, 2018Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Dimensions of human cognition, emotions, and thinking are combined to allow AI robots to perform human motions and manipulations.

“Dimensions of human cognition, emotions, and thinking are combined to allow AI robots to perform human motions and manipulations.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

More than 250 publications from 2015 to 2019 applied AI to energy conservation and renewable energy.

“More than 250 publications from 2015 to 2019 applied AI to energy conservation and renewable energy.”

May combine two claims.

CV techniques were used for road marking, incident detection, and autonomous vehicles.

“CV techniques were used for road marking, incident detection, and autonomous vehicles (Aymen and Mahmoudi 2019).”

• cites Aymen and Mahmoudi 2019Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Popular ML models combined ANN (including ANFIS) with GA.

“Popular ML models combined ANN (including ANFIS) with GA (Rodriguez-Soto et al., 2017).”

• cites Rodriguez-Soto et al., 2017Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Popular combinations include ANN with other models, optimizing such models with NC algorithms.

“Popular combinations include ANN with other models (statistical or time series), optimizing such models with NC algorithms (Wang and Srinivasan 2017).”

• cites Wang and Srinivasan 2017Not checked yet

  Not checked in the free run.

The quote only partly states this claim.

Cognition and robotics are largely overlooked components of AI in energy research.

“First, cognition and robotics are largely overlooked components of AI.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

Attempts to emulate human emotions and thinking through AI include affective computing, fuzzy logic, and expert systems.

“Attempts to emulate human emotions and thinking through AI include affective computing, fuzzy logic, and expert systems.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

Studies of AI seem to overemphasize technical issues.

“Studies of AI seem to overemphasize technical issues.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

GA is the most popular EC optimization technique.

“GA is the most popular EC optimization technique.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The ARIES software incorporates several ML models for data analysis to help researchers understand complex, nonlinear relationships.

“The software incorporates several ML models for data analysis to help researchers understand complex, nonlinear relationships (Death, 2015).”

• cites Death, 2015Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

Rule-based systems such as ARIES are popular for modeling ecosystem services.

“Rule-based systems such as ARIES 1 are popular for modeling ecosystem services.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Since 2015, ML models have been increasingly used for predicting and optimizing water resource conservation.

“Since 2015, ML models have been increasingly used for predicting and optimizing water resource conservation.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

New methods that leverage AI provide better probabilistic estimates of biodiversity components in time and space.

“New methods that leverage AI provide better probabilistic estimates of biodiversity components in time and space.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Some researchers use decision trees (especially random forest), multiple regression, and spline regression in water resource conservation.

“Some researchers use decision trees (especially random forest), multiple regression, and spline regression (Ochoa and Urbina-Cardona 2017).”

• cites Ochoa and Urbina-Cardona 2017Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

AI applications in water resource conservation have attracted significant research attention.

“AI applications in water resource conservation have attracted significant research attention (Table 2-3).”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

With GA frequently used as an optimization technique, popular ML models for biodiversity include ANN and BN.

“With GA frequently used as an optimization technique, popular ML models for biodiversity include ANN and BN (popular with modeling networks in ecosystems).”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

Biodiversity research identified a range of techniques to provide estimates of land-related service provisions, in contrast with more traditional static spatial modelling technique.

“Biodiversity research identified a range of techniques to provide estimates of land-related service provisions, in contrast with more traditional static spatial modelling technique.”

May combine two claims.

Research on AI for sustainability has explored many environmental sustainability topics and has focused on specific AI applications for enhancing biodiversity.

“Research on AI for sustainability has explored many of those topics, and has focused on specific AI applications for enhancing biodiversity (Table 2-2).”

Probably not a substantial claim: may not pass S3 (load-bearing).

Recent studies of expert systems have utilized ML models and fuzzy logic to support decision making.

“Recent studies of ES have utilized ML models and FL to support decision making (Leo Kumar, 2019).”

• cites Leo Kumar, 2019Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S2 (checkable).

If we adhere to the current self-interested reductionist paradigm, we will devise only irrational 'approximate solutions', often under the guise of rationality.

“Consequently, if we adhere to the current self-interested reductionist paradigm, we will devise only irrational 'approximate solutions' (Cox and Raja 2007:1), often under the guise of rationality.”

• cites Cox and Raja 2007:1Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing). May combine two claims.

The challenges of transforming human activities are exceedingly complex and require all our technological expertise to develop immediate and long-term solutions.

“although the challenges are exceedingly complex and require all our technological expertise to develop immediate and long-term solutions (for a detailed review of techniques, see Duan et al., 2019).”

• cites Duan et al., 2019 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

AI can support the derivation of culturally appropriate organizational processes and individual practices to reduce the natural resource and energy intensity of human activities.

“we argue that artificial intelligence (AI) can support the derivation of culturally appropriate organizational processes and individual practices to reduce the natural resource and energy intensity of human activities.”

• Also stated in: § 1. Introduction · p. 2 · plain

  “In this opinion, we argue that AI can support the derivation of culturally appropriate organizational processes and individual practices to reduce the natural resource and energy demands of human activities.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The literature review was drawn from Web of Science and Google Scholar databases, prioritizing journal articles from 2015 to 2019.

“Next we present our literature review, drawn from the Web of Science and Google Scholar databases. We prioritized journal articles from 2015 to 2019”

Probably not a substantial claim: may not pass S3 (load-bearing).

There is an increasing call to explore the social and ethical issues that surround AI.

“there is an increasing call to explore the social and ethical issues that surround this technology (Chia, 2019).”

• cites Chia, 2019Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S2 (checkable).

The true value of AI is not in reducing energy, water, and land use intensities, but in facilitating and fostering environmental governance.

“The true value of AI will not be in how it enables society to reduce its energy, water, and land use intensities, but rather, at a higher level, how it facilitates and fosters environmental governance.”

• Also stated in: § 1. Introduction · p. 2 · plain

  “The true value of AI will not be in how it enables individuals and society to reduce its energy, water and land use intensities. Rather, the true value of AI will be realized at a higher level, in how it facilitates and fosters effective environmental governance.”

Probably not a substantial claim: may not pass S2 (checkable).

The success of AI solutions will be determined by how well they navigate and influence psychological, sociological, and organizational factors that currently impede human progress in this area.

“Although AI solutions are inherently technical, their success will be determined by how well they navigate and influence psychological, sociological, and organizational factors that currently impede human progress in this area.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

Going beyond current thinking patterns to devise science-based solutions and policies requires a paradigm shift.

“But going beyond current thinking patterns to devise science-based solutions and policies requires a paradigm shift.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

Leveraging AI to conduct large-scale pattern-recognition with vast amounts of data gives society the opportunity to leapfrog, expedite, or morph its way of thinking and devising science-based solutions and policies for environmental problems.

“Leveraging AI to conduct large-scale pattern-recognition with vast amounts of data gives society the opportunity to leapfrog, expedite, or morph its way of thinking and devising science-based solutions and policies for environmental problems.”

Probably not a substantial claim: may not pass S4 (based).

As AI consumes and learns from vast amounts of data, the resulting decisions will be different from those taken by even expert humans.

“as the machine consumes and learns from vast amounts of data, the resulting decisions (informed by objective data and free of cognition biases and emotions) will be different from those taken by even expert humans.”

• AssumptionRests on an assumption: AI decisions are informed by objective data and free of cognition biases and emotions.

Probably not a substantial claim: may not pass S3 (load-bearing).

The world has entered a critical state of climate crisis.

“Unfortunately, the world has entered a critical state of climate crisis (Leahy, 2019).”

• cites Leahy, 2019Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

AI offers humans the opportunity to think, plan, and execute holistic solutions to environmental degradation and the climate crisis, free from reductionism and the self-interest of individuals and small collectives.

“Alternatively, AI offers humans the opportunity to think, plan, and execute holistic solutions to environmental degradation and the climate crisis, free from reductionism and the self-interest of individuals and small collectives.”

May combine two claims.

AI has the potential to overcome information asymmetries and the bias of human emotions, two factors that create barriers to developing solutions for environmental sustainability.

“AI has the potential to overcome information asymmetries and the bias of human emotions, two factors that create barriers to developing solutions for environmental sustainability (Cullen-Knox et al., 2017).”

• cites Cullen-Knox et al., 2017Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

The world continues to struggle with the need to transform human activities to ensure long-term existential survival.

“More than three decades after the notable Brundtland Commission Report, the world continues to struggle with the need to transform human activities to ensure long-term existential survival”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based). May combine two claims.

Focusing on AI for sustainability is the next logical step for the subdomain of green IS and will extend knowledge related to Green IS solutions.

“Focusing on AI for sustainability is the next logical step for the subdomain of green IS (IS for sustainability research) and will extend knowledge related to Green IS solutions.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Predictions from climate models can appear to be alarmist or to undermine threats due to strong socio-political elements in the climate crisis debate.

“Considering that the climate crisis debate has such strong socio-political elements, predictions from such models can appear to be alarmist or to undermine threats.”

• AssumptionRests on an assumption: The climate crisis debate has strong socio-political elements.

Probably not a substantial claim: may not pass S3 (load-bearing).

Stochastic approaches that permit inherent randomness with probabilities for different scenarios make the resulting climate models more speculative.

“Stochastic approaches that permit inherent randomness with probabilities for different scenarios make the resulting models more speculative.”

Probably not a substantial claim: may not pass S2 (checkable).

Understandings of behavioral responses to climate crises are critical.

“Thus, understandings of behavioral responses to climate crises are critical.”

• AssumptionRests on an assumption: Coordinated behavioral and policy responses are essential for developing climate crisis mitigation strategies.

Probably not a substantial claim: may not pass S3 (load-bearing).

Heightened cybersecurity and data management risks require abilities to navigate data standards and integration protocols.

“Heightened cybersecurity and data management risks require abilities to navigate data standards and integration protocols.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

If ML models are simplified and highly biased, they will always perform poorly for both training and test data.

“In contrast, if ML models are simplified and highly biased, they will always perform poorly for both training and test data.”

Classified as probably not a substantial claim. May combine two claims.

DL and NN models are often static and isolated, whereas climate crisis and environmental degradation are complex issues entailing responses and interactions among human, ecological, political, economic, social stakeholders, and subsystems.

“Thus, DL and NN models are often static and isolated, whereas the climate crisis and environmental degradation are complex issues entailing responses and interactions among human, ecological, political, economic, social stakeholders, and subsystems.”

May combine two claims.

AI applications try to mimic human decision-making but do not focus on human responses to decisions.

“Although AI applications try to mimic the human decision-making process, they do not focus on human responses to decisions.”

Classified as probably not a substantial claim.

Rebound effects occur when savings paradoxically increase consumption and offset savings.

“A compelling example of interconnections between social and technical dimensions are rebound effects, identified initially by Jevons in 1865, occurring when savings paradoxically increase consumption and offset savings.”

• cites Jevons in 1865Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

The technical limitations of climate models are attributable to other factors such as political ideologies or special interest lobbying that increase science-policy gaps and interfere with environmental governance.

“In either case, the models' technical limitations are attributable to other factors such as political ideologies or special interest lobbying that increase science-policy gaps and interfere with environmental governance.”

Probably not a substantial claim: may not pass S3 (load-bearing).

How human actors respond to AI is predicated on various psychological and sociological aspects.

“How human actors respond to AI is predicated on various psychological and sociological aspects.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

AI generally, and DL in particular, present opportunities for organizational decision makers to leverage computation models such as game theory and agent-based models that can extend resource capabilities.

“AI generally, and DL in particular, present exciting opportunities for organizational decision makers to leverage computation models such as game theory and agent-based models that can extend resource capabilities.”

Classified as probably not a substantial claim. May combine two claims.

The debate on the environmental value of IS is still not settled, as research in this area is in the nascent stage.

“The debate on the environmental value (Soomro et al., 2019) of IS is still not settled, as research in this area is in the nascent stage.”

• cites Soomro et al., 2019Not checked yet

  Not checked in the free run.

May combine two claims.

AI application creates a huge carbon footprint, which represents a direct form of rebound effects.

“Although AI is a promising and powerful technology for sustainability, its application creates a huge carbon footprint, which represents a direct form of rebound effects.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

If AI is meant to enable or extend an employee's opportunities, then overall morale and acceptance of AI may be high.

“If the AI is meant to enable or extend an employee's opportunities, as we may see in an expert based AI application, then the overall morale and acceptance of AI may be high.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The convergence of AI with other emerging technologies, such as blockchain, is expected to generate industry-wide benefits.

“Furthermore, the convergence of AI with other emerging technologies, such as blockchain, is expected to generate industry-wide benefits (Hughes et al., 2019).”

• cites Hughes et al., 2019Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

Current AI applications for sustainability have failed to adequately resolve cybersecurity and data governance aspects.

“Nevertheless, current AI applications for sustainability have failed to adequately resolve these aspects, which typically fall under general data governance concerns.”

Classified as probably not a substantial claim.

The mindful use of AI systems can allow society to expand beyond traditional levers for environmental governance.

“However, the mindful use of AI systems can allow society to expand beyond these traditional levers for environmental governance.”

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

AI can identify strategies to reduce emissions, water footprints, and bio-diversity risks, and to handle increasingly frequent and severe natural disasters.

“AI can also identify strategies to reduce emissions, water footprints, and bio-diversity risks, and to handle increasingly frequent and severe natural disasters.”

May combine two claims.

Quantum computing will aid ML models and increase processing speed, but it will also increase hacking speeds.

“Although quantum computing will aid ML models and increase processing speed, it will also increase hacking speeds (Denning, 2019).”

• cites Denning, 2019Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S4 (based).

Multilevel research models can distinguish the explanatory power of different levels of analysis, which can help avoid variance-bias tradeoffs.

“Multilevel RM can distinguish the explanatory power of different levels of analysis, which can help in avoiding variance-bias tradeoffs.”

Probably not a substantial claim: may not pass S2 (checkable).

With multilevel research models, researchers will be able to examine environmental sustainability in an integrated fashion with social and economic sustainability and overcome limitations related to uncertainty of human behavioral responses, unintended adverse consequences, and performance measurement.

“Finally, with multilevel RM, researchers will be able to examine environmental sustainability in an integrated fashion with social and economic sustainability and, in this way, overcome limitations related to the uncertainty of human behavioral responses, unintended adverse consequences, and performance measurement.”

Probably not a substantial claim: may not pass S2 (checkable).

AI for sustainability has to be approached from a design thinking perspective to develop effective, net zero emission AI applications.

“Hence, AI for sustainability has to be approached from a design thinking perspective to develop effective, net zero emission AI applications.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

Prospect theory suggests that responses to negative stimuli are stronger than responses to positive stimuli, but when AI based models predict doomsday climate change scenarios, we do not often observe stronger negative reactions.

“Prospect theory suggests that responses to negative stimuli are stronger than responses to positive stimuli, but when AI based models predict doomsday climate change scenarios, we do not often observe stronger negative reactions.”

• cites Prospect theoryNot checked yet

  Not checked in the free run.

May combine two claims.

The IS value literature recognizes that IS alone is ineffective in generating value and complementary assets are key to realizing value from IS.

“The IS value literature recognizes that IS alone as ineffective in generating value and complementary assets as key to realizing value from IS (Shea et al., 2019).”

• cites Shea et al., 2019Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based). May combine two claims.

AI can become autonomous, with checks and balances for managing risks and exceptions.

“Thus, AI can become autonomous, with checks and balances for managing risks and exceptions.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based). May combine two claims.

AI will help to move society forward in terms of monitoring the potential unintended consequences of AI and supporting performance measurement of AI solutions for sustainability.

“Moreover, it will help to move society forward in terms of monitoring the potential unintended consequences of AI (e.g., rebound effects) and supporting performance measurement of AI solutions for sustainability.”

Probably not a substantial claim: may not pass S2 (checkable).

Such measurement is a highly challenging undertaking.

“Such measurement is a highly challenging undertaking.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

A design thinking approach will help minimize potential unintended consequences and improve the effectiveness of AI solutions.

“Proposition 3: AI for sustainability should be approached from a design thinking approach, as a design thinking approach will help to minimize potential unintended consequences and improve the effectiveness of AI solutions.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

Understanding the psychological and sociological underpinnings of human response is necessary for effective long-term solutions.

“Proposition 4: AI for sustainability should incorporate environmental psychology and sociology perspectives, as understanding the psychological and sociological underpinnings of human response is necessary for effective long-term solutions.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Siloed approaches minimize cybersecurity risks but are inefficient and arguably less effective.

“Although siloed approaches minimize cybersecurity risks, they are inefficient and arguably, less effective.”

Probably not a substantial claim: may not pass S4 (based).

Unless conscious efforts are made, the environmental footprint of AI will increase with the increasing sophistication of techniques and complexity of models.

“Unless conscious efforts are made, the environmental footprint of AI will increase with the increasing sophistication of techniques and complexity of models.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

New forms of environmental governance can minimize self-interest by removing the bias of human emotions and information asymmetry.

“New forms of environmental governance can minimize self-interest by removing the bias of human emotions and information asymmetry.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

AI can identify factors that adversely or positively affect environmental performance.

“In other sectors, AI can identify factors that adversely or positively affect environmental performance.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Government plays a key role in safeguarding AI's impacts on society.

“Thus government plays a key role in safeguarding AI's impacts on society (research proposition 12; Duan et al., 2019).”

• cites Duan et al., 2019 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

AI is justified as an investment priority for multiple organizational performance outcomes.

“As a new opportunity to reduce operational costs of small and large industrial processes, AI is justified as an investment priority for multiple organizational performance outcomes.”

Probably not a substantial claim: may not pass S4 (based).

AI-related context diverges from conventional IS behavior contexts.

“AI-related context diverges from conventional IS behavior contexts.”

Probably not a substantial claim: may not pass S4 (based).

AI presents an opportunity to leverage hardware and software to learn new ways of organizing industrial processes and global supply chains to reduce water, energy, and land inputs and net carbon emissions.

“AI presents an opportunity to leverage hardware and software to learn new ways of organizing industrial processes and global supply chains to reduce the water, energy, and land inputs associated with upstream commodity production, such as mining, oil, gas, and utilities, while reducing net contributions to carbon emissions.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Interactions among stakeholders can result in feedback loops in the system.

“These interactions can result in feedback loops in the system.”

Probably not a substantial claim: may not pass S2 (checkable).

AI cannot be a panacea for all our complex environmental sustainability problems.

“AI cannot be a panacea for all our complex environmental sustainability problems.”

• AssumptionRests on an assumption: Historically new technologies have promised to solve immediate problems, but in time proved unsustainable.

Probably not a substantial claim: may not pass S2 (checkable).

AI, being a disruptive IS, brings new questions to the forefront.

“AI, being a disruptive IS in terms of its unique characteristics to learn from data and new computational capabilities, brings new questions to the forefront.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

Psychology and sociology investigation is essential for understanding human responses to AI.

“Consequently, psychology and sociology investigation is essential for understanding human responses to AI.”

May combine two claims.

In the building sector, expert decision-support tools are often widely inaccurate, while theoretical models fail to calibrate to the real world.

“For instance, in the building sector, expert decision-support tools are often widely inaccurate, while theoretical models fail to calibrate to the real world (Kolokotsa et al., 2016).”

• cites Kolokotsa et al., 2016Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based). May combine two claims.

Money is central to the way capitalist societies have evolved and function, therefore, the economic rationale will strongly influence how outcomes of AI-based models are perceived, accepted, and used.

“Money is central to the way capitalist societies have evolved and function, therefore, the economic rationale will strongly influence how outcomes of AI-based models are perceived, accepted, and used.”

• AssumptionRests on an assumption: Money is central to the way capitalist societies have evolved and function.

Probably not a substantial claim: may not pass S3 (load-bearing).

If AI has transcended the limitations associated with traditional or legacy IS, it will overcome concerns related to productivity paradox.

“If AI has transcended the limitations associated with traditional or legacy IS, it will overcome concerns related to productivity paradox (Brynjolfsson, 1993).”

• cites Brynjolfsson, 1993Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

AI presents opportunities for operating in high-risk areas and environments, for advanced extraction techniques, and for managing highly volatile materials without risking human health.

“AI presents opportunities for operating in high-risk areas and environments, for advanced extraction techniques (e.g., space-based mineral extraction), and for managing highly volatile materials (e.g., nuclear waste) without risking human health.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

The significance of economic rationales that will influence different responses to AI applications for sustainability cannot be ignored.

“Besides psychology and sociology, the significance of economic rationales that will influence different responses to AI applications for sustainability cannot be ignored.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

A system dynamics perspective will capture interactions and feedback loops among technology, users, and other stakeholders.

“Proposition 2: AI for sustainability should be viewed from a system dynamics perspective, as a system dynamics perspective will capture interactions and feedback loops among the technology, users, and other stakeholders.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

Timely, precise, and accurate measurement and monitoring of interventions is required to advance toward environmental sustainability.

“To advance in any meaningful way toward environmental sustainability, timely, precise, and accurate measurement and monitoring of interventions is required.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

IoT device integration provides real-time data that can be analyzed in settings most appropriate for learning and organizational sensemaking.

“Internet of Things (IoT) device integration provides real-time data that can be analyzed in settings that are most appropriate for learning and organizational sensemaking.”