Gen AI for Good with James Hodson: Using AI to create a better tomorrow, today
Do you know where to go to get medicine when your child gets sick? Do you know where to go to recharge your devices when the power has been out for 20 hours? Are you able to communicate with other people in your society? How the hell do you find out what benefits you're entitled to from the international community or from your government? Because of the situation that you find yourself in when you've been in a blackout for three days? In a world where AI is a terrifying, culture changing behemoth, is there another side to the story? A side where AI is actually out there making positive change and impacting humanity for the better? My guest today, James Hodson, would like to believe so. He's a researcher, leader and entrepreneur, as well as the visionary behind AI for Good. James, thank you so much for being here. Welcome to AI in action.
Fantastic pleasure to be here, to be with you today. Thank you so much for inviting me. For sure. We're going to really enjoy picking your brain, but first we're going to start with the elephant in the room. We know that AI is scary to a lot of people. So when did you start to think that instead it could also be a force for positive change and good? Fantastic question. It's been a really long journey.
The foundation was started around ten years ago, but the idea that we need to use technology in society in more sustainable ways and to effectively attack kind of challenges that we face, is something that's age old, right? It's not something that we came up with. For us, it was in the late 90s. Over time, right during the 2000s, and you have to remember with AI, we've had multiple waves of the expectation of commercialization and the anticipation that AI is going to transform the universe. And it's been like that since the 50s, and now we're in a wave where we are actually seeing structural, transformative change with the technologies, and it's extremely exciting. It's also, for many people, seeing it in person, right, interacting with technologies that their parents even would not have been able to imagine or seeing a self-driving car stop for them at a crossroads in a city, it's a little bit, you know, takes them aback.
And I can see why people would be anxious about what could happen next. For us, that's what, that vacuum in terms of the understanding of how technology interacts with society, was the reason for starting our organization. We wanted to be the providers of expertise, kind of a backbone for society to understand how to best utilize emerging technologies in a way that's going to strengthen our resilience, it's going to strengthen our communities. It's going to make us better problem solvers overall. It's going to make us better able to use the resources that we have on this earth, which are not infinite, in order to give our children a better future right, and live more harmoniously together so that we're actually building a positive, forward looking force in the world.
I love that overall commitment. Give me some examples though. I want to hear what are some ways that you've really found that you can use AI for forces of positivity? Our mission as an organization can be best described as economic and community resilience through technology. And it's the through technology lens which changes kind of how we think about what we're doing. In the back, we are, yes, technologists, right.
And I've been a researcher in AI for two decades, and I've worked across a whole variety of problems, and I've seen a lot of what can be done with the technologies. But we are also economists and economists are people who study human behavior. They study the structure of society and try to understand what kinds of changes will have the impact that we desire, and what's the most effective way of achieving that end result.
You have to bring those two parts together in order to have an impact on the world. You can't just design a technology and expect that it's going to go out there and have good. You need to do it in context. You can't design a strategy for a whole nation state and their approach to technology and society, and how they can attack certain challenges that they're facing without actually being within that community. We prefer to strengthen the communities with the skills and the resources and the ideas that are required. So that they can build things for themselves that are going to have a long lasting impact on adoption.
Now, let me dive into the examples that you asked for. A few years ago, we were approached by the Tony Blair Institute, based in London, to work with them and the government of Ethiopia in order to develop a forward thinking strategy for the next 10, 20 years of economic development and societal transformation in Ethiopia. We’ll detach this conversation from how Ethiopia has developed politically since that stage. So, for example, in Ethiopia, they have the same amount of internet bandwidth for the whole country. As a few blocks around my house in California. Imagine how that constrains your ability in the modern world to build an innovative society that's solving its own challenges.
It's a massive bottleneck, right? People actually don't need much to be entrepreneurial problem solvers in the modern world, right? Laptop, good internet connection, and... good coffee, right? Right. When you are lacking one of these kind of foundational columns. Right. You need to start thinking about
how can you quickly get infrastructure in place. It turns out the strong internet backbone and that kind of infrastructure is actually really important when it comes to health care. In a country that's as massive as Ethiopia, with you know, more than 100 million people, most of them living in rural areas, most of them farming, right? Most of them not really having any background in technology and not thinking about opportunity through a technological lens. You need to actually reach them as a first stage. So a lot of what we then ended up thinking about was, what's the basic infrastructure that you need to put in place first, and how do you get people on board with these ideas? So that's the first example that a lot of the work we do actually is kind of advocacy oriented in communities to make them understand that the transformative or the foundational layer that they need in order to be able to build with technology, be able to transform their society, it's not that far away from where they are now, but you do need some resources in order to enable that to take place. Another example on the other side, we're very involved in Ukraine.
Now, you know, obviously some of the viewers will have seen that. I'm wearing a, rather bright outfit today. Right. Now, this is a traditional Ukrainian Vyshyvanka. Right. It's an ornate set of patterns related to particular regions of Ukraine.
And for us, as an organization, we started to enable better solutions to these challenges. Right. 2014, they were ratified by the United Nations. And that's when we also started. And that's the taxonomy that we use for thinking about the problems that we want to solve in the world. And if there's one thing in the world right now which is causing us to fall even further behind on achieving these goals, it's the fact that we have massive conflicts that are liable to spill over and are already having deep economic effects around the world. So we use the fact that, we have the technology expertise, the scalability expertise to solve problems, to also help in Ukraine.
We have 50 staff on the ground. We have about eight different locations that we work across, and we're active with the entire Ukrainian government and every municipal authority in the country, which means that we're developing technologies that are used by Ukrainians on a daily basis. We're developing strategies and policies based on data analysis and machine learning and and AI to enable the Ukrainian economy to withstand the invasion.
In all of known history, whenever there's been an all out invasion of a country, there has been a collapse of the banking system, for instance. And you need banks in Ukraine because of the way that the economy was being managed. There was no run on the banks, there was no crisis in the banking system. It's one of the few examples of where good economic thinking actually was able to prevent something that could have then led to a much, much, much deeper set of problems. And as a result, you actually have a functioning society, right? Still. And that means that we're able to go in and use technology to advance that.
Of course, if you didn't have that layer of well-operating, well-oiled society, then you would have challenges putting AI and other technologies into place. And all of this starts, of course, as you mentioned, with the strong foundation, you've got to have that infrastructure that’s set on up and then you can build. Can you give me some more ideas in terms of how can we think about AI outside of the the traditional scope of productivity and actually enhance that? I love this because, you know, we're, we're getting into conversations that impinge both upon kind of modern economic thinking and how we understand society and our relationship with productivity and growth. And so we use measures like GDP, right, and GDP is an economic measure. It shows you basically how money cycles through the economy, right.
And how quickly it moves around right from the available pool of capital. It's not actually a very good measure of how productive human beings are. So we have these measures that are kind of divorced from, you know, how productive are we being sat here? Right. Maybe it's a net negative. We don't have good ways of knowing what accelerates our ability to solve the problems that are important, or reduce the amount of energy that we need to expend to solve the problem.
Right? Or to make a sale or to develop a new product or to build infrastructure. We don't know actually what the relationship is between how many hours we work and how much we're paid, and what actually we're building in society, and how much closer we are to achieving what we want. So I try to step back from that and not worry too much about the productivity angle, because it's not useful for actually putting technology into the context where it can be used for societal transformation. So I'm going to take an example.
Imagine that you have a war. There's a lot of uncertainty. How do you find the resources that you need in order to stay calm, stay productive, stay alive, and thrive as much as possible and in such a situation. So you might imagine, okay, do you know where to go to get medicine when your child gets sick? Do you know where to go to recharge your devices when the power has been out for 20 hours? Are you able to communicate with other people in your society? How the hell do you find out what benefits you're entitled to from the international community or from your government because of the situation that you find yourself in when you've been in a blackout for three days? However, also consider outside here, right in New York City. How do people find out about the opportunities that they're entitled to? How do they make good decisions about their plan, plans for the future? How do they make sure they're getting the training that they need and the support that they need psychologically? Right, in terms of family planning, in terms of their relationships with, with the community and how they actually interact. Can you give me some more examples of how AI for good works? Maybe on not as large of a scale, but a little bit more intimately. So we've, we've had the pleasure of working also on a very municipal level, right, around the world, including in Brazil and Canada, here in the US and Europe, looking at ways that we can understand and analyze the challenges that are being faced in the community and respond to challenges, but in a way that then we can take those solutions and potentially scale them to other places that that need them.
For example, in British Columbia, we played a role in developing what is now an interconnected wildfire mitigation system, which allows all of the different agencies that are needed in order to respond to potential situations where fires get out of control in a way that's coordinated and in a way that gives them the information that they need in real time. This is the type of thing where it's not just about artificial intelligence, it's about the underlying infrastructure. It's about the hardware. Right.
What are your ground observation beacons? How are you getting the overall view of the situation that's happening? How are you choosing where to put resources so that you can respond most quickly to a specific point where the flare up is most likely? Also, how do you maybe identify ahead of these situations where the risks are so that again, you're being proactive about mitigating risks rather than dealing with a fire, which is obviously much more costly, much more dangerous, and is going to lead to much more damage, to people's lives, by having to live through it. On the other hand, we can go to another example, which is workplace. So workplace equality.
Right. We have as one of the Sustainable Development Goals, gender equality. And today, as a society, we're still very, very far behind on this goal.
So for women in particular, it is very difficult to achieve the same career success as men. And it's more difficult today than it was ten years ago. How do you attack this kind of situation? We're not going to be able to get into the DNA of every single company and really solve it on an individual basis. But one thing we can do with technology is we can bring more transparency to the signals that people can use in order to understand the culture of a company, in order to understand the current structure and how that might be contributing to outcomes, and by doing that, by bringing transparency to an area, we can change the approach that corporate leaders take to managing their culture and make them focus on these questions more. So that's the economic idea behind it.
With technology, well, AI is fantastic at aggregating data. It's also fantastic at identifying patterns that we might not be able to see, or precursors to patterns that we're not going to be able to associate, right, with kind of manual, old school traditional methods and so one thing that we do is we work with a variety of data partners, including organizations that have been collecting, you know, resumé and job applications for decades in order to look at the relationship between hiring and companies, the structure, the internal structure of those companies and how promotions happen and how people are rewarded for their work, and also how happy people are, within these organizations. One thing that we're now coming kind of to the to the point where we're going to be releasing it publicly is a scorecard of tens of thousands of companies in the US by metrics that are important to understanding whether this is a culturally good fit for somebody. So a culturally good fit from the perspective of an employee, culturally good fit from the perspective of an investor, and culturally good fit from the perspective of a customer.
You shine a light and let the market show the way forward. Right. And so what we also try to do with technology is to bring more light to these areas. I'd say it's less about saying AI needs a special way of doing things than, well, let's look at, you know, how do other engineering disciplines cope with risk and risk mitigation? When we build a bridge, for instance, do we go out there, kind of stand right looking across the chasm and think, all right, you know, I think I need a piece of steel that's about 600ft long.
All right, let's see, let's see what happens. Right? Usually we don't do it that way. It's not that we are saying that we should do something wildly different with AI, but computer science in general has avoided being pinned down as an engineering discipline for decades.
Now, when we think of critical infrastructure in society, it's not highways, rail, electrical, water. You know, it's cyber and communications, right? That is pretty much the underpinning of everything else that we're doing now because it goes away. And you don't have a banking sector anymore, you're not going to have reliable water supply because it's all being basically managed by an interconnected information system. Right. The electrical grid is going to be able to balance your power. You're going to have surges that break all of your appliances at home. That backbone has been built by computer scientists, by AI scientists, by statisticians.
And it's an engineering discipline. You know, we should have come to terms with the fact that it was a real engineering discipline and put proper procedures in place decades ago. But the reality is we've always treated it as ah, it's like an art, right? Programing is an art, right? It's an art that apparently we think is possible to just come out of a language model because it's artistic, right. So we're going to generate it and it will kind of approximate the thing that we want it to do.
But we need to constrain that art, just like bridge building is an art now, right? We have many beautiful bridges, but it's also, at its core, a mathematical engineering based discipline that needs to have rules about how we go about identifying the problem that we're solving, ensuring that we know right and can trust all of our understanding of the base that we're building upon. Just like you wouldn't build a bridge on quicksand. Oh, it didn't look like quicksand when when I was observing it.
We need to make sure that we're doing the right thing. We need to test things appropriately at scale. Right. So start small, build up. And we need to ensure that when we're educating people, when we're building the next generation of people who are going to work in this discipline, that we are not only giving them the mathematical scientific knowledge that they require, but that we're immersing them in how to solve problems in society.
And when we have that type of thinking, you know, what else will happen? What's that? They're going to see other opportunities that go beyond what they're doing in their job. It won't just be, oh, I've got to deploy a chat bot today that responds to questions on behalf of, this podcast on our website. It's going to be, well, you know, when I was walking into work, I saw that, you know, there's clearly an inefficiency in how we're, you know, managing kind of pedestrian routes during construction or something like that. You know, we start to provide people with a basis for thinking more broadly about their skills and capabilities within the community and society, and we engage with them on this level. And, you know, people naturally want to solve problems. And I think that we need more of this approach.
On a practical level, can an individual coder start to use AI for good today? I think we're at a really interesting time. So I started programing in, the mid 90s. Right.
And there was magic when I started programing, and I realized that I was one individual with a toolset that could scale and touch people around the world. Once you internalize that, it's extremely empowering. And it was true in the 90s as well as it's true today. But today we can harness per person thousands of kilowatt hours more in electricity than I was able to harness in the 90s to solve problems, which means that we can solve problems now at an unprecedented scale, as an individual, right? We have ability to tap into resources that no single person would have been able to access with their idea at any point in history before. And ultimately, you know, if we put it in crude terms, innovation and the economy is about energy consumption, okay? And what we're doing right now with AI is we're consuming energy, which is allowing us to build solutions, right, with small teams that are extremely sophisticated and have far reaching potential.
It sounds like you're placing a lot on intention, on focus in terms of how we can use AI. Businesses and nonprofits, sometimes their interests don't always align. So how can we think about AI for good? Both your organization, but then also the theory, the philosophy. If you're a business person who's listening to this right now, if you're a person who's really focused on the bottom line in the business right now, how can all of this make sense for them? Right now in the United States, because of the unequal access to technology and opportunity, education, health care, infrastructure, there is an enormous amount of untapped economic potential in the US from people that have been disenfranchized from the system. If as a corporation, you focus in on helping to solve the challenges in society that have impacted that, it will actually unlock an enormous amount of economic value.
Because ultimately, what is a society if only half of it is contributing to solving a problem? Right? Without the private sector, governments today don't have the expertise or really the ability because of kind of political deadlock in some sense, to really fundamentally rethink how we solve problems. While in some sense corporate America, the NGOs, civil society more broadly has a level of flexibility that allows us to not be constrained by special interests. And employees at these corporations have even more power because ultimately they hold the keys to the engine, right? And the engine needs to work for everybody, right? Not not just for kind of a powerful elite at the top. There are benefits that come from operating in a community with an attitude of that community being consequential to your business in the future, right? Not a resource to be used, but an effective network in which we can operate to bring positive outcomes for all.
Positive outcomes for all. I'm going to take that now, and I want to go back to where we started this conversation. There's a lot of fear surrounding AI. Why should we not be as afraid as we are excited? We are at a time where if we can invest more as a society in creating with, creating the precedents with technology that underpin big solutions to grand challenges like climate change.
Which obviously is, you know, a framework problem that we all need to be thinking about, not just from, you know, how do we capture carbon in the atmosphere at scale, but also what are we doing every day in our own lives that's impacting this? And what technology can play a role in bringing that to the forefront of people's attention and so on. I've always been in a situation throughout my life and career where the technology that I'm using could have power to do good and power to do bad in catastrophic ways on both sides. And yet we have always chosen as society, eventually, the path of hope and positive outcomes from the technology rather than the path of destruction. And I think that we will continue to choose this in the future. Again, that doesn't mean that we don't need to be eternally vigilant. You can be pretty positive about the amazing things that we're going to be able to achieve over the coming years and decades.
You know what? I kind of want to just keep that as a mantra now. So, James, thank you for such an enlightening but also a very encouraging conversation today. I believe, and I really hope, that everybody who's been joining us in this conversation has also found this equally empowering. So thank you again so much for joining us. It matters. Folks, AI for Good is a nonprofit, so their funding does come from people like you and me.
So please go check out their website and consider donating at aiforgood.org. Again, thank you all for joining us. We'll see you next time.
2025-01-04 23:44
