National AI Policies and The Climate Crisis
Does National Planning for the Field of Artificial Intelligence Consider the Climate Crisis?
Greta Thunberg says we need a war effort to be able to turn things around towards a planet that can avert the absolute worst consequences of the climate crisis. Technology and development; racing towards the future; however does not have a great reputation when it comes to taking these concerns seriously. We can choose to follow the linear conception of ‘revolutions’ from (1) mechanisation and steam; (2) mass production and electricity; (3) electronic IT systems automation; (4) cyber physical systems.
It is highly possible to feel uncomfortable with such a linear and simple definition — perhaps for good reason.
We are not ‘all’ moving forward. It is not a ‘revolution’.
As much as we examine what we stand to gain we must examine what we stand to loose or what we need take care we maintain.
The perception of winning or loosing is too apparent. Even in the conception of a ‘digital divide’.
Therefore going into the 4th industrial ‘revolution’ should perhaps not be seen all too much as a grandiose charge into the future with sparkly digital confetti flying across the room. It is easy to talk about AI, blockchain, 3D printing, IoT, robotics etc. I am of course guilty of this myself, and build on that narrative too despite attempting at times to address inequality.
Sustainability and Ecology in AI Security Domains
Stanford Institute for Human-Centered Artificial Intelligence held a conference on Ethics, Policy, and Governance. At this conference Jessica Cussins Newman was there showing a collection of how different countries’ prioritised different in their policies. within different AI Security Domains you can see the bottom right corner as sustainability and ecology.
Jessica Cussins Newman is a Research Fellow at the UC Berkeley Center for Long-Term Cybersecurity.
There is of course a lot to consider, and she mention that only EU took all into consideration.
She wrote a report called Toward AI Security examining different security domains in a white-paper.
“This report uses the lens of global AI security to investigate the robustness and resiliency of AI systems, as well as the social, political, and economic systems with which AI interacts. The report introduces a framework for navigating the complex landscape of AI security, visualized in the AI Security Map.”
Her report provides real-world examples across four security domains is social. Sustainability and ecology is placed within this social category.
“Sustainability and Ecology AI applications rely upon data centers for massive quantities of computational power, which could be cause for concern.131 Data centers in the United States currently use more than 90 terawatt-hours of electricity per year, relying upon the equivalent of more than 34 large coalfired power plants to provide this energy. Within these area she mentioned a few points which is summed up well in its negative ramifications:
- Data centers use roughly 416 terawatt-hours per year of electricity, and some analyses project this number will triple in the next decade.
- The production of digital devices is also unsustainable, as less than 16% of global e-waste is formally recycled.
- The volume of e-waste is predicted to rise to 50 million metric tons or more every year.
- In Asia, e-waste has increased in volume by approximately 63 percent since 2012.
- The use of AI chips could exacerbate the need for electronic manufacturing, which is the most carbon-intensive phase of manufacturing.
The positive possibilities in helping environmental goals are summed up as well:
- DeepMind was able to use AI tools to identify hidden inefficiencies in Google’s data centers and reduce their energy consumption by 40 percent.
- Apple has made the transition to running its operations on 100 percent renewable energy, powering its data centers with numerous solar farms.
This initiative prompted two-dozen other companies in Apple’s supply chain to pledge to do the same. These successes help expand local clean energy markets while making a strong business case for sustainable growth, and may inspire others to follow the same path.
China, France, UK and India Has Policies Addressing Sustainability and Ecology
All following figures are from the Toward AI Security report and were retrieved the 4th of November 2019.
China has a focus on sustainability and ecology.
She mentions France and the UK as possible policy leaders which we may see more of in terms of tackling environmental threats.
In her whitepaper India is already mentioned as a country that has a focus on sustainability and ecology.
Seeing as these large countries have been identified as important in regards to sustainability and ecology within their policies it makes perfect sense to examine them closer.
As such I will start with these four.
I will attempt to do examine the policies that has been referred to in this document in the following days.
This is #500daysofAI and you are reading article 154. I write one new article about or related to artificial intelligence every day for 500 days.