The ability to sense the environment, learn environmental variables and their patterns, and process historical and contemporary information makes artificial intelligence a "game changer" also in the field of sustainability and environmental protection. In May 2019, a special UN conference on artificial intelligence and the environment was held, which was designed to promote this field and channel additional investments into it. One of the estimates was that the contribution of applications in the field of the environment to the global GDP will reach about 5.2 trillion dollars
By Ran Ben Michael, Angle - Science and Environment News Agency
The Jewish-Urban legend from the 15th century about the Golem from Prague is probably one of the first stories that dealt with introducing human abilities to inanimate objects and also about the dangers inherent in losing control over them. The breakthroughs in the field of artificial intelligence make the cocoon of the 21st century a valuable tool in dealing with the global environmental crisis.
Artificial intelligence is a branch of computer science that seeks to make machines act and react like humans. For this, the machines have to learn patterns and categories and absorb sensory information while experimenting - this training is also called deep learning. These components are common to applications such as Siri and Alexa, autonomous cars and robots: they process a lot of information to perform tasks according to fixed patterns, but also reacting to changes.
Recently, there has been a major breakthrough in the field: calculation capabilities, sophisticated algorithms, hardware and management of huge data (big data) - which allow Netflix to offer you a new movie, Facebook to tag your photos, and service centers to solve hardware problems for you. The leap in capabilities also translates into economic activity: by 2030, artificial intelligence is expected to contribute approximately $16 trillion to global GDP.
The ability to sense the environment, learn environmental variables and their patterns, and process historical and contemporary information makes artificial intelligence a "game changer" also in the field of sustainability and environmental protection. In May 2019, a special UN conference on artificial intelligence and the environment was held, which was designed to promote this field and channel additional investments into it. One of the estimates was that the contribution of applications in the field of the environment to the global GDP will reach about 5.2 trillion dollars.
A significant role in solving the climate crisis
A joint report by the World Economic Forum and the PWC company identified six environmental areas in which artificial intelligence has an important role: the climate crisis, biodiversity, oceans, water security, air pollution and resistance to extreme weather. The examples in each of these are many.
The climate crisis is the greatest contemporary human challenge and artificial intelligence has a significant role to play in it. Smart management of urban transport or the electricity network, for example, can save greenhouse gas emissions and air pollution. A report by PWC and Microsoft estimated that through artificial intelligence it is possible to reduce global greenhouse gas emissions by 4 percent by 2030 (equivalent to Gt2.4 carbon dioxide). On the side of resilience with the crisis, improving the weather forecast can ensure better dealing with extreme events (such as storms or fires) thanks to more accurate alerts. Artificial intelligence already helps in environmental decision-making in areas such as agriculture - to improve productivity by various means - or the management of the water network (as IBM does in China).
In 2017, Microsoft established "Artificial Intelligence for the Earth" which supports technological ventures to protect the planet
Another field to which artificial intelligence is mobilized is environmental enforcement, which for various reasons is a point of failure in environmental management. A study published in the scientific journal Nature in October 2018 estimated that the number of detected violations of anti-pollution regulations could increase up to seven times at lower costs for the authorities (from 6.7 currently detected to 13-50 percent with AI). Thanks to the immediate detection, it is possible to stop polluting processes (mainly industrial) at an earlier stage, which will help the industry meet the required standards and save indirect expenses of long legal processes. Similarly, patterns of illegal fishing or hunting of wild animals can be identified and thus prevented.
Windows to the environment
An important part of developing artificial intelligence projects is going out of the lab into the real world and learning in the field. One of the leaders in the field is the technology giant Microsoft, which launched in 2017 the "Artificial Intelligence for the Earth" (AI for Earth) initiative and is investing $50 million in grants over five years to expand the use of applications for the purposes of preserving biodiversity and nature.
One project is of an organization working to protect the snow leopard, a unique and mysterious Asian predator. The snow leopard population has dropped to about 6,000 individuals and is in danger due to the reduction of the living space, poisoning, hunting, climate change and the like. Because of the difficulty of tracking this species, the organization uses cameras with motion and heat sensors to monitor the population over vast areas. Microsoft's artificial intelligence solution helps the organization quickly analyze hundreds of thousands of images from dozens of cameras while reducing identification errors and omissions and thus obtain an accurate picture of the population and even of details within it.
Overcoming technical obstacles in nature conservation is also evident in the WildMe project, which is designed to help participants in citizen science projects correctly identify species and thus help experts analyze the vast visual information collected. In Germany, a team of experts developed "Nature's Shazam", a system that enables the sound identification of species and analysis of the size of their populations, and they are using Microsoft to expand the application, which overcomes the great effort involved in using other means, such as burying traps or visual tracking when it comes to insects. Thus, for example, the information collected with the help of the application in an agricultural field makes it possible to specify the use of pesticides and fertilizers and to better use insects for the needs of pollination or biological control.
"Artificial intelligence can also be harnessed for self-improvement," says Dr. Tomer Simon, VP of National Technologies at Microsoft Israel, who coordinates the AI for Earth initiative in Israel. "Thus, Microsoft reduced 24.7 percent of electricity consumption at its huge campus in Redmond, where there are 150 buildings and 60 thousand employees and saved millions of dollars in operating expenses. It is possible to raise the scale of such an application and reduce the environmental footprint at the municipal level, which is very suitable for the crowded Israeli reality."
What next?
It is understood that the large technology companies do not invest in environmental applications only for altruistic reasons. Artificial intelligence advances as the systems learn more, therefore cooperation with entities that can take advantage of it are necessary for continued development. In addition, such programs improve public relations - Microsoft, for example, used the snow leopard initiative as a television advertisement for its technology.
Artificial intelligence can take part in changing the course of the environmental crisis and alongside the effort of the technology companies, the academy also takes part in this. The University of Cambridge, for example, established a research center for artificial intelligence and sustainability, with a public investment of 200 million pounds (about one billion shekels). Recently, the leading scientists in the field, in a long article, called on the artificial intelligence community to harness their energy in the fight against the climate crisis and among decision makers in the business community there is optimism regarding the use of these technologies.
There are also risks in technology such as wrong decision-making (after all, the machines learn from imperfect people), ethical bias (who taught the machine?), lack of transparency (the decisions will not always be in line with human rationale), and dangerous use (for example, for terrorist purposes). But even if artificial intelligence is not a magic solution, the potential inherent in it for maintaining sustainability and dealing with environmental challenges is enormous.
"Israel," believes Simon, "can restore its days as a powerhouse for exporting agricultural technologies thanks to developments based on artificial intelligence, for example in the field of precision agriculture, which enables surgical treatment in almost all aspects of the agricultural unit. All you need is a smartphone and an app."
More of the topic in Hayadan:
