ROLE OF ARTIFICIAL INTELLIGENCE IN ACHIEVING SUSTAINABLE DEVELOPMENT GOALS
Updated: Nov 30, 2021
Author: Vachsi Shah, V year of B.Com.,LL.B. from Institute of Law, Nirma University
Artificial Intelligence (AI) is the assistant that sustainable development needs to plan, execute, advise and plan the prospect of our earth and its sustainability more efficiently. Technology like AI will help us build more resourcefully, use resources sustainably and diminish and manage the waste we generate more effectively, among many other materials. The Sustainable Development Goals (SDGs) are a set of 17 interconnected global goals that serve as a "blueprint for a better and more sustainable future for all." The United Nations codified the SDGs in 2015, to achieve them by 2030.[i]
Uniting AI with sustainable development will facilitate all productions to design an improved planet, addressing current needs with no compromising future age groups due to climate change or other major challenges. Renewable energy efficiency can also be improved with AI. Companies are already employing this technology to determine the daily availability of energy-generating facilities (wind turbines, hydraulic plants, biomass plants, and so on) to forecast the amount of energy required in the coming days and, eventually, to avoid and diagnose problems.[ii]
Aside from the energy sector, AI may benefit a variety of industries and enterprises, all while benefiting the environment. It is used in agriculture, for example, to improve irrigation and fertilization efficiency. Artificial Intelligence can estimate crop needs using humidity, temperature, and fertilization sensors. Drones that assist farmers with surveillance, as well as hyperspectral image analysis for comprehensive pest control, are among the most advanced agricultural sustainability solutions.[iii]
AI'S IMPACT ON THE ENVIRONMENT
There is an increasing number of AI applications in the environmental sector, including those in energy (e.g., smart grids), agriculture, and water management. Recent advancements in IoT technology, as well as AI algorithms in vision and sensor fusion, have made this viable.[iv]
The introduction and possible adoption of Electronic Vehicles (EVs) and smart appliances, for example, could improve the efficiency and reliability of power generation in a smart city environment. Furthermore, AI could help incorporate renewable energy into smart networks by effectively controlling risk and bridging supply-demand gaps. Some of these AI technologies, on the other hand, can be computationally expensive. In countries where dirty coal is still used to generate electricity, efficiency benefits can be overshadowed by pollution.[v]
According to estimates, overall electricity demand from Information and Communications Technology (ICT) alone might consume up to 20% of global energy by 2030, up from 1% today. As a result, establishing a more efficient and renewable-energy-based data centre, as well as infusing human knowledge into current models through priors, is critical for green growth. This is because the human brain uses far less energy (and does so more efficiently) than current AI models, and improving on this integration (e.g., physics-informed deep learning) could be beneficial not only to the environment but also to communities at large, who are particularly vulnerable to AI-system-based pollution.[vi]
The final set of SDGs is concerned with climate action, life underwater, and life on land (SDGs 13, 14, and 15). For the Environment category, we identified 25 targets (93 percent) where AI could be useful. The ability to analyze large-scale interconnected databases and devise coordinated actions targeted at environmental preservation could be one of AI's benefits. In the case of SDG 13 on climate action, there is evidence that improvements in artificial intelligence will aid in the knowledge of climate change and the modelling of its potential consequences. Furthermore, AI will help low-carbon energy systems that incorporate a high level of renewable energy and energy efficiency, all of which are required to combat climate change.[vii]
AI AND SOCIETAL OUTCOMES
Unfortunately, complicated AI systems are still expensive and lower-income families and people with disadvantaged backgrounds may not be able to afford them. If the government fails to regulate how the benefits of AI are dispersed among the various stakeholders, the likelihood of increased inequality will increase. Big producers, for example, may gain, but smallholder farmers may be left behind because they cannot afford an expensive AI system that can increase their output and productivity. Furthermore, existing racism, gender stereotypes, xenophobic tendencies, and hate crimes could be amplified if there isn't enough transparency and diversity.[viii]
Systematic racism and bias can still be found in AI systems, particularly in the NLP and Computer Vision sectors. This is because the data used for training is already tainted with societal biases, and without an intentional de-biasing method during data engineering, the once objective machine may inherit our subjective and irrelevant opinions. Transparency and diversity are thus essential for keeping our models as objective as feasible. One option would be a decentralization procedure in which AI technologies are implemented by teams with diverse cultural, ethnic, racial, and gender backgrounds.[ix]
Crypto-currency applications like Bitcoin, for example, consume as much electricity as some countries' electrical demand, jeopardizing SDG 7 outcomes as well as SDG 13 on Climate Action. According to some projections, the total electricity demand for information and communications technologies (ICTs) might reach 20% of global electricity demand by 2030, up from 1% currently. As a result, ICT technology's green growth is critical. More effective cooling systems for data centres increased energy efficiency, and the use of renewable energy in ICTs will all help to limit the expansion of electricity consumption.[x]
Human expertise must be embedded in the construction of AI models, in addition to more efficient and renewable-energy-based data centres. Aside from the fact that the human brain consumes far less energy than what is used to train AI models, the available knowledge introduced in the model (for example, physics-informed deep learning) does not need to be learned through data-intensive training, potentially lowering the associated energy consumption. Although AI-enabled technology has the potential to accelerate the implementation of the 2030 Agenda, it may also exacerbate inequities, which might stymie progress on SDGs 1, 4, and 5. In the case of economic, an instance of using AI to achieve better sustainability is the technology that has been developed in tunnel boring machines, which are particularly multifaceted equipment. A crash can stop all or a momentous part of subversive work in its tracks.[xi]
Overall, evaluating risk is the first step toward ensuring that AI breakthroughs are egalitarian and long-term. We only have the power and understanding to design the next generation of AI technologies that are kinder and more sustainable if we acknowledge these flaws.
The SDGs are a powerful lens for looking at internationally agreed goals on sustainable development, and they represent a leap forward in the representation of all spheres of sustainable development, including human rights, social sustainability, environmental outcomes, and economic development, when compared to the Millennium Development Goals. However, because the SDGs are a political compromise, they may be restricted in their ability to capture some of the more complicated processes and cross-interactions across targets.
[i] Bisaga, I., Parikh, P., Tomei, J., &; To, L. S. (2020, November 19). Mapping synergies and trade-offs between energy and the Sustainable Development Goals: A case study of off-grid solar energy in Rwanda. Energy Policy. Retrieved September 29, 2021, from https://www.sciencedirect.com/science/article/pii/S0301421520307394.
[ii] Insights - Science. (n.d.). Retrieved September 28, 2021, from https://science.sciencemag.org/content/sci/358/6362/442.full.pdf.
[iii] LYNCH, Shana. Andrew Ng: Why AI Is the New Electricity. Stanford Business. 2017. https://www.gsb.stanford.edu/insights/andrew-ng-why-ai-new-electricity
[iv] Williams, A., & Dupuy, K. (2017, May 11). Deciding over nature: Corruption and environmental impact assessments. Environmental Impact Assessment Review. Retrieved September 28, 2021, from https://www.sciencedirect.com/science/article/pii/S0195925516303274
[v] Panch, T., Mattie, H., &; Celi, L. A. (2019, August 16). The "Inconvenient Truth" about AI in Healthcare. Nature News. Retrieved September 29, 2021, from https://www.nature.com/articles/s41746-019-0155-4.
[vi] (PDF) Sustainable Cities and communities in New Zealand. (n.d.). Retrieved September 29, 2021, from https://www.researchgate.net/publication/342360199_Sustainable_Cities_and_Communities_in_New_Zealand.
[vii] Arlidge, W. N. S., Bull, J. W., Addison, P. F. E., Burgass, M. J., Gianuca, D., Gorham, T. M., Jacob, C., Shumway, N., Sinclair, S. P., Watson, J. E. M., Wilcox, C., & Milner-Gulland, E. J. (2018, April 18). Global mitigation hierarchy for nature conservation. OUP Academic. Retrieved September 28, 2021, from https://academic.oup.com/bioscience/article-abstract/68/5/336/4966810.
[viii] Will democracy survive big data and artificial ... - springer. (n.d.). Retrieved September 29, 2021, from https://link.springer.com/chapter/10.1007/978-3-319-90869-4_7.
[ix] Wintle, B. A., Kujala, H., Whitehead, A., Cameron, A., Veloz, S., Kukkala, A., Moilanen, A., Gordon, A., Lentini, P. E., Cadenhead, N. C. R., & Bekessy, S. A. (2019, January 15). Global synthesis of conservation studies reveals the importance of small habitat patches for biodiversity. PNAS. Retrieved September 28, 2021, from https://www.pnas.org/content/116/3/909?sfns=mo
[x] authors, A., http://orcid.org/0000-0002-5892-245X, V. S., & Additional informationFundingThis work was supported by Vetenskapsrådet project grant 2013: Development Space. (n.d.). The Sustainable Development Oxymoron: Quantifying and modelling the incompatibility of sustainable development goals. Taylor &; Francis. Retrieved September 28, 2021, from https://www.tandfonline.com/doi/abs/10.1080/13504509.2016.1235624.
[xi] WWW-NATURE-COM-S.VPN.SDNU.EDU.CN. (n.d.). Retrieved September 29, from http://www-nature-com-s.vpn.sdnu.edu.cn/articles/s41893-021-00730-6.pdf.