Photo by @franckinjapan

National AI Strategies and the Climate Crisis

Are Strategies for Artificial Intelligence Around the World Adequately Addressing the Defining Issue of Our Time?

We could look into the strategies of the five permanent members of the UN Security Council — China, France, Russia, United Kingdom and United States. However I chose to bring in India’s strategy as well, with its 1,366,417,754 people it is the second most populous countries in the world; it has a large military power; and is investing heavily in AI. Looking at AI strategies from China, France, India, Russia, the United Kingdom and the United States there is a clear difference in how the climate crisis is being addressed. China, France and India has a clear priority to do so whereas Russia, the United Kingdom and the United States do not.

The strategies are part of an overall policy within each country as such the described strategies is used to represent AI Policy within each country and may not do so fully. There is more to the AI Policy for each country than is written in these documents whether they are plans or strategies, however these various strategy documents or plans will be the focus of this article.

According to Jessica Cussins Newman in her report Toward AI Security there has been a great increase in the number of AI National Strategies & Policy Initiatives in the years between 2016–2018. I am very indebted to the work of Newman in writing this article, as she has been mapping AI Security and by doing so gathering an overview of different strategies that in turn gave me a better overview.

Retrieved on the 10th of November 2019

I will run through the six countries in alphabetical order.

1. China
2. France
3. India
4. Russia
5. United Kingdom
6. United States

1. China, AI Strategy and the Climate Crisis

1.1 Is the Climate Crisis Addressed in China’s Plan for Artificial Intelligence?

One of the most extensive analysis of AI Policy around the world which I have found so far has been undertaken by Jessica Cussins Newman and is focused on AI Security. To begin I would say clearly that climate is not mentioned specifically in China’s A New Generation of Artificial Intelligence Development Plan (新一代人工智能发展规划), AI Policy from 2017. On the other hand ‘sustainable’ is mentioned four times. There was two mentions of environmental protection. I will discuss these four mentions of sustainable; two of environmental protection; and the context within which they are mentioned. I have found six focal points that can be linked to addressing the climate crisis within China’s AI strategy from 2017.

1.2 Sustainable Development as the Center of the Intelligent

“We shall develop intelligent economy, construct intelligent society, safeguard national security, build the ecosystem with cluster integration of knowledge, technology, industry and mutual support of talent, system, culture, foresee the risks and challenges, propel sustainable development as the center of the intelligent, and as a result comprehensively enhance the social productive forces, national strength and national competitiveness for the purpose to speed up the construction of innovative countries and the science and technology power of the world, to achieve “two hundred years” goals and to provide strong support for the great rejuvenation of the nation.”

The first mention of sustainable is mentioned in the second section called ‘Overall Requirements’ within the first point ‘(A) Guiding Thought’. Towards the end the 200 year goals are mentioned. The Two Centenaries (Chinese: 两个一百年) is a set of goals advanced by General Secretary Xi Jinping following the 18th National Congress of the Communist Party of China held in 2012. It is said to be the basic foundation for achieving the “Chinese Dream”, another ideology advanced by Xi. They were formed back in the 80’s. The goal for 2020 is the elimination of extreme poverty in China, linking it to GDP.

Sustainability
可持续性
Kě chíxù xìng

‘Can hold continued nature’, as such this relates perhaps to ecology as well as the other notion of ability to sustain.

Sustainability has been a great part of China’s current strategies, however linking it partly to GDP may be unfortunate as this has been shown increasingly to be an indicator which can have adverse effects. Exchanging this measurement tool could be fortunate, however then it may not be an easy way to retrospectively track.

1.3 Healthy and Sustainable Developments of Artificial Intelligence

“The development of artificial intelligence is a complex systemic project related to the overall situation. In accordance with the layout of “build a system, grasp the dual attributes, adhere to the Trinity, strengthen the four support” , we shall form the strategic path for an healthy and sustainable developments of artificial intelligence.”

The Trinity mentioned here is dubious, and I may be wrong. However Carl von Clausewitz, who sought to illuminate the “trinity” of strategy by connecting the relationship between the people, the military, and the government, Sun Tzu looked at the interdependencies between these actors while accounting for the role of nature, terrain, and laws in shaping strategy. Granted I may be wrong.

On the other hand it is more likely that it refers to the goal of (1) a stable yuan, (2) open borders and (3) control of its own interest rates. This has been dubbed the impossible trinity.

This section talks not only of the goal, and that sustainability is at the center, it also insists that the path to the goal must be healthy and sustainable.

1.4 Transdisciplinary Sustainable Development

“We shall focus on the major scientific frontier issues of artificial intelligence, take into account the current needs and long-term development, break through the basic theory of artificial intelligence bottlenecks as the focus, and ensure that advanced layout may lead to artificial intelligence paradigm transformation of basic research, promote crossintegration of disciplines for sustainable development and depth applications of artificial intelligence and provide strong scientific reserves.”

In the translated document crossintegration is mentioned, and I have chosen to translate it to trandisciplinary.

It is unsure in what way this will be ensured, and which ‘type’ of disciplines that should be involved. The focus on integrating disciplines is an interesting thought as this has proven challenging in many setting. It is certainly necessary in an academic, business or government environment that there are strong scientific reserves to help facilitate this collaboration as well.

1.5 Sustainable Ecosystem for Consistent Development of Infrastructure

“High performance computing infrastructure. Continue to strengthen the supercomputing infrastructure, distributed computing infrastructure and cloud computing center construction. Build a sustainable ecosystem for the consistent development of highperformance computing applications. Promote the next generation of supercomputer research and development applications.”

Since the pathway to AI must be ‘sustainable’ it makes sense that there needs to be a sustainable ecosystem for development. Sustainable and consistent may of course be open for interpretation, and the speed at which the development is outlined may be a challenge to the sustainability of the projects that occur.

1.6 Application of AI for Environmental Protection

“Artificial intelligence brings new opportunities for social construction. China is in a comprehensive stage of building a well-off society, population aging, resource environment constraints and other challenges are still grim and the application of artificial intelligence in education, medical care, pension, environmental protection, urban operation, judicial services and other fields will greatly improve the precision level of public service, comprehensively enhancing the quality of life of the people.”

This goal is too general to say much in specific, and yet specifically mentioning AI to be opportunities in a set of different areas. The main point to drive home here is that there is a wish to increase precision in public service. There has been a want from governments to be more accurate and specific operating in an ambiguous world with large population. China is the most populous country in the world with 1,379,302,771 people estimated in 2017. It is hard to equate quantitative understanding to quality of life, however with this large number it is clear that making sense of this data or making decisions is likely to involve a large degree of automation.

1.7 Intelligent Environmental Protection

“Intelligent environmental protection. Establish big data platform for smart monitoring in the atmosphere, water, soil and other environmental areas. Build land and sea interaction, land and earth interaction, and information-sharing intelligent environment monitoring network and service platform. Research and develop intelligent forecasting models and early warning programs for resource energy consumption and environmental pollutant discharge. Strengthen the environmental protection and intelligent prevention and control systems for sudden environmental events in BeijingTianjin-Hebei, Yangtze River Economic Zone and other major national strategic regions.”

This relates both to ensuring the protection of environmental sites and predicting sudden changes in the environment. Sudden changes are expected to occur more often going forward, and therefore it is important to be able to use more accurate models to predict or understand what is happening across a large area. Forecasting would necessarily have to be a collaboration which has been the case across the world for weather data, however the will to measure and forecast is nothing new — adding artificial intelligence to the mix is not a strange course of events.

Summarising China’s AI Strategy in Relation to Climate

We can see six focal points that can be linked to addressing the climate crisis within China’s AI strategy from 2017. I would argue it is addressed within the strategy and there can be initiatives developed on the basis of these points, or arguments made to bring projects to life within these focal points. I have added one question to each point.

1. Sustainable development as the center of the intelligent. How do we make sustainability the center in the development of AI?
2. A strategic path to healthy and sustainable developments of artificial intelligence. How can the pathway to the development of AI itself be sustainable, so that the goal and the way there is?
3. Transdisciplinary sustainable development or/and integration of disciplines. Which disciplines needs to be integrated and which ways can this be done inside and beyond China?
4. Sustainable ecosystem for consistent development of infrastructure. How can the Chinese government audit its cloud providers or resource providers to ensure development is responsible? Which type of energy use can be seen in the different solutions?
5. Application of AI for environmental protection and increasing the precision level of public service. Which type of GovTech relating to AI can be explored by the Chinese government to improve public service?
6. Intelligent environmental protection and monitoring through big data. How can China collaborate with other countries on environmental protection and monitoring, and in which way can 5G networks be utilised for this purpose?

As a final remark I realise there are controversies about China that do not make these points unproblematic. I am not blind to these, however it is important to understand how we can work together to address the climate crisis — because it is a dire crisis and China has the largest population on the planet, we are all dependant on how China handles this pervasive issue.

2. The French AI Strategy and the Climate Crisis

2.1 How Does the French AI Strategy Address the Climate Crisis?

My topic for this week is looking at the AI Strategy in seeing the context in which they mention the climate, sustainability, environment and ecology. Largely taking a look at how they address the climate crisis in their AI Strategy. I have taken a surface look at: “For a Meaningful Artificial Intelligence: Towards a French and European Strategy.” written from the 8th September 2017 to 8th March 2018.

In the most ultimate French way it starts with a picture of the author who is a Mathematician and Member of the French Parliament. It was a mission assigned by the Prime Minister Édouard Philippe. Part of the report consisted of travelling around to different AI hubs in the world and learning more about the topic. It was a team consisting of a varied collection of people.

Straight off it is clearly visible that the French AI strategy has space dedicated to considering the climate and ecology when thinking about the development of artificial intelligence. The fourth section in their document is dedicated to using artificial intelligence to create a more ecological economy.

In the executive summary of the report it is said that:

“Carving out a meaningful role for artificial intelligence also means addressing its sustainability, especially from an ecological standpoint. This does not just mean considering the application of AI in our ecological transition, but rather designing natively ecological AI and using it to tackle the impact of human action on the environment. This is an urgent matter as world data storage requirements, inherently correlated to the development of digital technology and AI, could exceed available worldwide silicon production out to 2040.”

Natively ecological AI sounds a bit like fluff, but there has been talk of different AI solutions that are not trained on massive datasets. As well as storing less data, so it has a practical discussion point elsewhere, and is talked of in the industry particularly in France. The question of resources is a good one too, which I have not seen mentioned enough elsewhere (particularly in the US strategy).

Then again it was the Paris agreement which the world talks of, and it would be very sad indeed if the epicenter of a massive change with great agreement did not take these points seriously. “AI must be included in initiatives emerging as part of the Paris Climate agreement and the Global Pact for the Environment.” It is said as well to be a strategic imperative:

• France and Europe can spearhead this smart ecological transition by raising awareness on the international arena.
• The primary task is to consider both the impact of AI on achievement of the UN’s sustainable development goals, how it puts pressure on certain goals and how it can accelerate others.

2.2 How Can This be Done?

To do this it is argued that there has to be a devoted space for AI research and energy resource optimisation research to meet. These thoughts include too a consumer perspective creating a platform to make consumers more aware of their choices in technology. “This platform should also include a simple calculator to enable all citizens to gain greater awareness of these impacts and compare the environmental footprint of the various products, services, software and hardware.”

The cloud industry is mentioned in particular, that: “Public authorities must also act to make the value chain greener and support the European cloud industry to promote its ecological transition.” Uniquely, because I have not seen it elsewhere, this report argues for not only open software, but open hardware: “Lastly, making the AI value chain greener will clearly require open hardware and open software…” The last point, that was mentioned by China as well, was gathering large amounts of climate data.

On page 103 it is argued that it has to be part of the international agenda.

“France could propose setting up a major event along the lines of the COP 21, to showcase exemplary and highimpact initiatives. It could also be more closely involved in the convergence of the two transitions, ecological and digital, within international forums, particularly the G7, where discussions concerning AI were initiated and where France is shortly to take over the presidency”

2.3 A Threat and a Solution

One specific existing project is mentioned that I would like to bring forward as well. The Tara Oceans Project: freeing-up massive amounts of data concerning the oceans for the purposes of understanding and modelling a planetary biome. It is argued too that France needs to make more public data available. There is a table presenting the possible use of public data in regards to the environment.

Table from “For a Meaningful Artificial Intelligence: Towards a French and European Strategy.” p.29 retrieved the 6th of November.

It is stated in the report:

“By 2040 the energy required for computation will equally have exceeded world energy production”

It is said in the report that digital energy consumption is increasing by 8.5% per year and its contribution to world electricity consumption could reach 20% in a moderate scenario or even 50% in a pessimistic scenario by 2030. We can therefore see consumption multiplied 10-fold in 20 years’ time.

Although AI is a potential threat to the environment, it is also a potential solution. They argue that the ESEC, the French Economic, Social and Environmental Council, needs to play a major role in the strictly political debate on artificial intelligence and its consequences.

The French/European AI strategy has a clear focus on the environment and does not only talk of abstract priorities, but specific ways to address the issues.

3. India’s Artificial Intelligence Strategy and the Climate Crisis

3.1 How India Planning to Tackle the Climate Crisis in their AI Strategy?

I found India’s AI Strategy to be thorough dealing with a wide variety of different topics. The discussion paper National Strategy for Artificial Intelligence was released in June 2018. Jessica Newman in her analysis claims that the AI Strategy for India is quite extensive compared to others (such as the UK), covering a wide variety of topics. I am not saying this is bad or good, however I was interested to see how much climate concerns are present in the strategy.

In the Indian strategy ‘climate’ is mentioned six times. Environment is mentioned in a different context relating more to people. However sustainability is mentioned quite a few times too.

3.2 Agriculture

“As global climate becomes more vulnerable and unpredictable, dependence on unsustainable and resource intensive agriculture will only heighten the risks of food scarcity and agricultural distress […] The Indian agriculture sector is vulnerable to climate change due to being rain dependent. Varying weather patterns such as increase in temperature, changes in precipitation levels, and ground water density, can affect farmers especially in the rainfed areas of the country. AI can be used to predict advisories for sowing, pest control, input control can help in ensuring increased income and providing stability for the agricultural community.”

Not too long ago in 2017 more than 55% of Indians made a living from farming. Therefore it is not strange to see the focus on agriculture being front and center of the report.

3.3 Research Focus on Climate Change

“The research, focused on IT services and social good, will aim to provide powerful AI insights and recommendations for improved productivity. It also includes software analytics — building, testing, managing and modernisation of applications, solving real-life social issues such as malnutrition, human trafficking and climate change through prediction and recommendation models using AI.”

In the section describing research climate change is mentioned early on, although hidden in a list of other priorities.

3.4 Solar Power

“India is already playing a leading role in climate leadership, with Hon’ble Prime Minister Narendra Modi vowing to go “above and beyond” India’s commitment on Paris Agreement on climate change. Similarly, India has been a pioneer in a sustained push for clean energy revolution by leading the International Solar Alliance, and setting an ambitious target of 100GW of installed solar energy capacity by 2022. With 20GW of installed solar capacity, India is well and truly on its way to achieving this target.”

India has made a goal to install large amounts of solar power in the country to become more renewable, and this is a goal worth striving for.

3.5 Government Data Sharing

“Government data sharing: Government of India has large amounts of data lying in silos across ministries. The government can launch a mission of making all these data available for public good after undertaking proper privacy checks. For example — climate data, non-strategic remote sensing data, regional language speech (from All India Radio), soil health data etc.”

There are government sharing initiatives popping up in different locations, and it could be useful to follow this process. If the government shares data responsible there can be huge benefits to society.

3.6 India as a Test-Bed for Further Scaling

“Solving for India, given the complexity and multi-dimensional aspects of most of our economic and societal challenges, can easily be extended to the rest of the emerging and developing economies. An integral part of India’s strategy for AI involves tackling common and complex global challenges that can be solved through technology intervention, and India’s scale and opportunity landscape provides the ideal test-bed to ensure sustainable and scalable solutions.”

Piloting solutions in India and taking them to an international market seems a clear goal in this context.

3.7 Transportation

“Need for sustainable transportation: The recent initiative of the Government of India for announcing development of 100 Smart Cities is aimed at addressing this anomaly and catalyse smart strategies for urban planning which promote sustainable land use design and multimodal integration […] While new initiatives could take time to show realisable impact, the existing issues in urban mobility related to congestion, efficient traffic flow, movement of goods etc. can indeed be solved using AI technology.”

When we used AI-technology to solve problems it is likely to be causing new problems, however integrating technologies in society more seamlessly is a goal mentioned in different contexts.

3.8 Education

“The education sector needs to be re-aligned in order to effectively harness the potential of AI in a sustainable manner. In primary and secondary schools, there is a need for transition to skill based education in subjects relevant to AI. Often criticised for being overly knowledge intensive, Indian education is in urgent need of transition particularly in subjects relevant to STEM, or computer based education. As jobs based on technology become prominent, so will the need to develop applied skills in a continuously changing environment.”

AI in education plattforms is an interesting proposition.

3.9 Sustainable Business Models Aggregating Data

“Today, incumbents continue to enjoy an oligopoly in building sustainable business models in AI for two main reasons: (a) they can successfully buy data in the informal market setting due to availability of resources and reach to negotiate one-time contracts continually, and (b) they have specialised departments to work on different facets of the development value chain.”

This is a conversation of informal versus formal data. It seems informal data is relatively easy to get hold of in India, however as otherwise good data quality is needed so I can see this being an issue.

3.10 Responsible AI Development

Retrieved on the 9th of November 2019

Lastly this slide makes some good points regarding which areas we could invest effort into addressing the climate crisis.

4. Russian AI Strategy

According to the Russian news agency TASS: “Russian President Vladimir Putin approved the National Strategy for the Development of Artificial Intelligence for the period until 2030. The presidential executive order of October 10 was published on Friday on the official website of legal information.”

The Center for Security and Emerging Technology (CSET) at Georgetown University published a translation of the Russian report. The word climate or sustainability is nowhere to be found. The word environment is only mentioned in relation to the word economy referring to ‘economic environment’.

The Russians do not have a focus on climate and sustainability in their current AI Strategy scheduled to roll towards 2030.

5. The UK Artificial Intelligence Strategy and the Climate Crisis

5.1 How Focused is the UK AI Strategy on the Climate Crisis?

Talking to I realise strategy or policy is not the most exciting word in casual discussions. Even at a governance or ethics tracks it does not seem like this word is the one that gets a crowd excited. On the other hand perhaps that depends on what ‘types’ of policy that you suggest to implement — discussing issues or areas for possible policy-making may not seem appropriately action oriented. Then again policy is a course or principle of action adopted or proposed, as such a framework for action. The British framework for action is the AI Sector Deal and it is in short what I will discuss in this article: how does it related to the climate crisis?

First off we can start by saying that despite being updated on the 21st of May 2019 the AI Sector Deal does not mention the word ‘climate’, it does not mention the word ‘crisis’ either for that sake. In terms of ‘environment’ it mentions this 19 times, but usually in the context of business environment. The word ‘sustainable’ is not mentioned a single time.

“Similarly, open environmental data has been used to create flood risk and water quality apps. To test the autonomous vehicles of the future we will need good quality 3D topographic data on road conditions and roadside obstacles.”

Still they have been marked as falling within the security domain in terms of sustainability and still they received a marking in Toward AI Security, Jessica Cussins Newman’s report providing an overview of the different AI strategies. This grid was based on that strategy.

I would therefore say after having look at the AI Strategy from China and France in comparison that there is a great difference in the degree of focus on the climate change. None of the key actions towards the end of the British AI strategy indicates anything regarding climate change.

I would like to write more, however there is not much more to write about regarding the current strategy.

There was a select committee appointed by the House of Lords.

It is very strange to see Lords, Viscounts and Baronesses examine artificial intelligence. Seeing the prolonged Brexit negotiation is not an enlightening or bright process at all. Reading the current AI strategy in the United Kingdom was equally drab, it seems the United Kingdom is out of touch with its population or the global community when it does not mention at all climate or sustainability.

In academia however one initiative that I find of interest is the Oxford Initiative on AI×SDGs. One possibility is for this initiative led by professor Luciano Floridi which lecture AI in Society — Opportunities and Risks you have to see. He is as well the Director of the Digital Ethics Lab of the Oxford Internet Institute. To consider itself successful I would argue the initiative has to influence at least the local strategy in the United Kingdom in a different direction.

6. The United States Needs a National Vision for Artificial Intelligence

6.1 Will a New Strategy Will Consider the Climate Crisis?

Looking at the strategies of China, France and India there is a general focus on the environment and sustainability. It is easy to spot early on in the United States statement on the Whitehouse website in the Artificial Intelligence for the American People that there is no focus on climate change. The National Artificial Intelligence Research and Development Strategic Plan: 2019 did not contain many thoughts regarding the climate crisis either nor the ecological environment. Then again this may not come as a surprise since the President Trump formally pulled out of the Paris agreement, and this is being formalised to come into effect one day after the election of a new president.

Thus since the current strategies and plans can be said to have no vision in regards to climate change, environment nor sustainability we have to look elsewhere for vision.

6.2 A National Vision for AI

The 23rd of October 2019 a post was made to the Stanford Institute for Human-Centered AI (HAI). This post was called We Need a National Vision for AI, it was written by By Fei-Fei Li and John Etchemendy who are both co-founders of HAI. They start the text with a clear message:

“Establishing global leadership through a bold AI policy and plan is critical for the economic growth and stability of our society”

6.3 Climate Change

“AI has the ability to be a force multiplier of our very best — and very worst — intentions. It can help us address our most vexing challenges: managing natural resources; mitigating climate change…”

The issue of climate change is not very visible in the post, in fact only mentioned a single time, so it does not feature prominently. Yet it is mentioned, and that is far better than any current US strategy as done. As such this vision is already an improvement compared to the current vision in the United States.

6.4 Sustainability

“We should provide early-stage support for emerging technologies through grants, investment and technical resources, with an emphasis on agriculture, manufacturing, healthcare, sustainability and clean energy.”

Sustainability was also mentioned just one time in the vision and it could feature more prominently, however again it is already better than the current AI Strategy.

6.5 Make AI a Clear Strategic Priority

The text suggests a new AI ecosystem across education, research and entrepreneurship, with an investment of at least $120 billion over ten years.

1. Support public research to pursue the next generation of AI breakthroughs, with an emphasis on interdisciplinary research.
   Budget: $7 billion/ year. Establish national and regional research hubs in partnership with leading universities and emphasise cross-disciplinary research and diverse teams. Launching a National Research Cloud would provide high value data and high-performance computing for public-interest research.
2. Invest in education, with an emphasis on inclusion.
   Budget: $3 billion (double the current annual federal K-12 STEM spend). The US needs to educate a more diverse future workforce in science, technology, engineering and math (STEM), including artificial intelligence and computer science, as well as support research and programs to address job displacement and reskilling.
3. Spur innovation and support entrepreneurs. Budget: $2B. Entrepreneurship is the heart of the US economy. The Small Business & Entrepreneurship Council estimates firms with fewer than 100 employees comprise 98% of US businesses.
4. Implement clear, actionable international standards and guidelines for the ethical use of AI. Partner with foreign governments, companies, and civil society organizations to concretely implement global AI principles, such as those developed by the OECD.

It must be mentioned that while Obama was in office there was a report called Preparing for the Future of Artificial Intelligence that was released October 2016. Within this report both climate and sustainability is mentioned several times. This is as you by now know quite contrary compared with later reports and strategic communication during the Trump administration.

Overall Conclusion

Are AI policies around the world adequately addressing the defining issue of our time? I would say the answer is no. Having only half of the six countries outlined even addressing climate or sustainability is too weak. None of the countries mentioned ‘climate crisis’ in their strategies, so the sense of urgency is not present in any strategy. It must on the other hand be said that China, France and India have began addressing the climate crisis in their AI strategies or plans — this is a great step in the right direction.

If there are updates on the AI Policies of these different countries or strategy documents that I have missed please be sure to send these to me through a response here or elsewhere, so that I can amend my article.

This is #500daysofAI and you are reading article 160. I write one new article about or related to artificial intelligence every day for 500 days.