No Match Found
Freddie Martin: Hello and welcome to all new episode of the economics and business podcast. I am your host Freddie Martin.
Climate change poses one of the greatest threats to our world at the moment and fears over our ability to tackle this issue are growing. Whilst change in behaviours is going to be fundamental to addressing the issue, increasingly people are turning to developing new technology to help improve efficiency.
Today, I am joined by two economists from our team, Saloni Goel and Edmond Lee, who are part of a team commissioned by Microsoft to look at how one such technology, AI, can be used in the fight against climate change.
Thank you both for joining me today.
Saloni Goel: Thank you Freddie.
Edmond Lee: Thank you.
Freddie: I was wondering first, Edmond, could you give me some background as to why Microsoft commissioned this work?
Edmond: Microsoft already have been running this AI program, which awards grants to projects that use artificial intelligence to address issues in climate change and biodiversity. They would like to know more about the impact of the program and as well as the relative importance of different areas that they chose to invest in. They asked us to look into different AI applications in four key sectors, which are our agriculture, energy, transport and water. That’s how we started this program and that culminated in this public report, now available in the public domain.
Freddie : Right, so this is clearly quite a large and complicated question’s answer then. I was wondering how did you decide to approach the issue, how did you break down the problem?
Edmond: In short, there are three steps. In the first step, we looked at 16 different applications or pathways that AI could be applied to four sectors, how they are going to improve productivity in different ways. In second step, we then put these quantified numbers into our big economic and environmental model, to map out how industries are linked to each other, how households are likely to react, and how the feedback loop between mitigated climate change and better economic outcome would look like. After doing this model exercise, in the third step, we then mapped out what kind of recommendations there could be for Microsoft to prioritise their work, and what everyone in the policy and technology community can do to make our future more sustainable.
Freddie : So that was again businesses, and governments, and individuals.
Edmond : Yes.
Freddie: What was the impact of this report for clients or in the general media?
Saloni : I must say it was truly ground-breaking research. As Edmond highlighted, this is one of the few studies that sits at the intersection of artificial intelligence and climate change. For Microsoft particularly, our steady is helping them inform their investment strategy for the AI for earth fund. It’s a 50 million dollars fund. As Edmond mentioned they are looking to use this fund for AI applications that are targeted towards sectors like agriculture, water, energy, transport and other sectors with the sustainable angle. Our study has provided them the evidence to understand, which applications have the greatest environmental and economic benefit? Moving beyond Microsoft, the study has attracted a lot of media attention from the likes of Forbes.
We’ve been invited to speak at conferences like the AI for Good Global Summit and the CogX summit, which is UK’s largest artificial intelligence summit. In this world where most of the AI debate is around labour automation and robots stealing our jobs, I think our study really widens the debate and adds a positive angle on how we can leverage this technology for tackling some of the deepest problems our earth faces today.
Freddie: That’s incredibly impressive and just adds to the value of the report, directing tens of millions of pounds of investment.
Saloni : Yes exactly.
Freddie: You are talking about AI levers, could you talk about some specific examples because it’s quite hard to imagine how AI technology can be used in these different sectors?
Saloni : Absolutely, one of the examples I would pick would be from agriculture, because as you said it’s quite difficult to imagine how AI would play a role in these sectors. Something we studied was around precision monitoring of environmental conditions. This includes the application of field sensors to monitor local weather conditions and crop health to exactly understand the precise level of inputs required. For example, based on the crop health and the temperature and moisture conditions, it would recommend the amount of water that the crop needs, the amount of fertilisers and pesticides that should be used.
Now, taking a step back, the advantage of all of this is the cost savings that are enjoyed by the sector, because of the precise measurement, you are able to use less number of inputs, less amount of inputs, but to get the same yield.
Freddie: Because there is no waste at all.
Saloni : Yeah, there is no wastage, and you are precisely monitoring the need of the crop. As a result, we modelled that the agriculture sector, for example, would enjoy all these productivity benefits and cost reductions from applications such as precision monitoring.
Freddie: Were there any other levers that had particularly large impact that you studied?
Edmond : One of the levers or pathways with larger impact would be monitoring and management of energy consumption. To start with, we are trying to move to a world where energy is more renewable. One problem we would often face is that renewable energy like solar and wind, some of them are actually quite variable, sometimes you get wind and sometimes you don’t. One thing that could be done is to use artificial intelligence and internet of things technology to more actively monitor, and to optimise how energy is used and how some nonessential machines and appliances can be switched on and off from time to time.
We can also adjust prices of energy, either wholesale or retail, to allow people to respond to these price signals and that could allow energy to be used when it’s cheapest, and to reduce wastage, and as a result there will be a two-fold impact. One, it could mean a lower cost base for certain industries, which means they can produce more, become more efficient; and on the other hand that could also mean lower energy consumption and a more renewable energy mix, which means less carbon emission.
Freddie: It seems there are many different applications of AI across all these different sectors.
Edmond : We are just looking at the subset of what’s possible.
Freddie: Overall, what were the results of your report?
Saloni: I am going to break this question into two parts, because as we said, we are looking at how AI can enable a sustainable future. So, this involves environmental benefits, but at the same time we are also looking at economic benefits of these applications. Focusing on the second one, which is the economic benefits, we found that our AI environment applications have the potential to boost global GDP by up to 4.4% by 2030, this is relative to business as usual. This amounts to an economic uplift of 5.2 trillion US dollars. This is mainly driven by automation of manual and routine tasks, savings of specific inputs, for example, in the case of precision monitoring, saving of specific pesticides, water, etc., for the same amount of yield, and in general higher productivity enjoyed by different sectors in the economy.
Looking at the second pillar, which is environmental benefits, we found that AI for environment applications can reduce global greenhouse gas emissions by up to 4% by 2030. This would accelerate the move to a low carbon world, because this amounts to a reduction of 2.4 gigaton of carbon equivalence. Just to put this number in context, this 2.4 gigaton is equivalent to the 2030 annual emissions of Australia, Canada, and Japan combined. We are looking at quite a significant reduction.
Freddie: Also, it goes against the idea that environmental policies have to be expensive or cost us, we can achieve growth and reduce emissions at the same time.
Saloni : Yes, because you are boosting economic growth while also benefitting the environment.
Freddie: Are there any regional differences across the world in these impacts?
Saloni : That’s a very good question. We found that the economic benefits, and actually the environmental benefits as well are predominantly captured by Europe, North America, and East Asia, and you can imagine that this is primarily because of the higher propensity to adopt new technology, the current digital readiness level, and current proportion of large proportion of skilled workforce. These are the regions that would gain the most as per our study, but we also believe that with targeted public and private sector investments, and with proper upskilling of labour force, regions like Latin America and Sub-Saharan Africa can also leapfrog and go beyond these model to projections.
Freddie: There is a real scope for policy there.
Saloni: There is some part to be played by public and private sector players.
Edmond : That’s not the only story there. Another story is that multiple studies are saying that tropical regions are going to be more adverse and be effected by climate change. Put it this way, mitigated economic loss is an economic gain compared to the base line of business as usual, which means by mitigating climate change, we are bringing positive economic outcomes to Africa, South America and South-East Asia, and that would become more and more pronounced in the further future beyond 2030.
Freddie: Right, so the region, Edmond, is actually very key here. One concern that many people have of AI technology is this idea of labour automations, robots coming in and taking people’s jobs, etc. Did you find that there was increased unemployment?
Edmond : I guess there are two layers to this. One of course, technology would mean that certain jobs, humans can do it more efficiently and fewer people will be needed, but it would also create more demand for new goods and services as a result of the society become more productive and richer. Indeed the good news is, in our report, what we found is that AI applications in the four key sectors would create 38.2 million more jobs in that terms across the world that’s 1.0% on the current global workforce.
The region most benefitted is East Asia followed by Europe, partly because they both have a very educated workforce, but also a very balanced industry mix. On the other hand, you can see that Indo-Pacific, which is South and South-East Asia might have a slightly negative impact on their employment, because despite they benefit more in specific terms in white collar working level jobs, they are very culture intensive and as a result they would probably require more investment into reskilling and probably improving their industry mix.
Freddie: Interesting, so this research has led to some really fascinating results, but it does speak to the question of where do we go from here. I was wondering if either of you had thoughts on what you think governments and businesses can take from this research into using AI to tackle climate change?
Saloni : I think, that’s a very important question. All of us, businesses, governments, and societies, need to move forward on different dimension in order to make the positive scenario. We have just highlighted a realistic scenario. If you look at businesses first, then one of the key elements is the adoption of responsible AI. All these large tech companies that are deep into the digital transformation need to champion responsible technology practices that take not just the economic benefit into account, but also the social and environmental benefits, and basically focus on long-term value creation for everyone in the society. If you look at governments, there is an urgent need to address gaps in digital infrastructure, and as we highlighted earlier, this need is more significant in some sectors and in some regions like Sub-Saharan Africa and Latin America.
There is also an urgent need to reskill and upskill certain sections of the labour force. As Edmond was saying, that depending on the sector and depending on the region, certain sections stand to lose out as of now. There is an urgent need to make sure they are involved in this technological transformation and they are trained so that they can get absorbed in the new jobs that are created as a result of AI.
Edmond : Actually, learning from our previous experience in technological change and how do we actually help people to move careers without loss of course, otherwise we’ve seen some of the results already and it’s not good for everyone at the end.
Freddie: Overall, it seems like there are lot of different policy implications taken from the report, but my overall impression is that of a very positive outlook.
Saloni : Definitely, as we said there is a significant potential for gains, both economic and environmental, and with proper policy action in this area and with proper action from the business, we can achieve this positive scenario that we’ve outlined.
Freddie: Thank you both for joining me for a really interesting discussion today.
Saloni : Thank you so much for hosting us.
Edmond : Thank you very much.
Freddie: If you would like to find out more about the research that our PwC team did, please take a look in the description for our podcast for a link to the full report as well as some really great interactive data tools powered by Microsoft.
Thank you for joining us and please subscribe.
UK Chief Economist, PwC United Kingdom
Tel: +44 (0)7711 562331