In this podcast episode, Dr. Beverly Wright, Vice President of Data Science & AI at Wavicle Data Solutions, and Chelsea Lamar, VP, Global Sustainability at AIT Worldwide Logistics, examine how AI is shaping the future of renewable energy. As artificial intelligence continues to evolve, so do its energy demands. This discussion explores AI’s environmental footprint, the essential role of green power, and how AI itself can drive sustainability in the pursuit for responsible innovation.
Speaker details:
Watch the full podcast here or keep scrolling to read a transcript of the discussion between Beverly and Chelsea:
Beverly: Hello, I’m Doctor Beverly Wright and welcome to Tag Data Talk. With us today we have Chelsea Lamar, VP of Global Sustainability at AIT Worldwide Logistics, and we’re talking about “Leveraging AI to optimize consumption of renewable energy.” Super exciting.
Chelsea: Yes, cool topic, thank you for having me.
Beverly: Let’s start off with a little bit of background. Tell us, why are you so cool?
Chelsea: I think I’m cool because I have always wanted to work in energy and sustainability renewables, and I do work in energy sustainability renewables. When I was very small, I always had this sense of wonder about the natural world and loved being outside and fresh air and clean water and all those things. I’m also a little bit of a geek. I remember when I was probably 7 or 8 years old, I had a computer game. It was called Ecosaurus. It was an island populated by dinosaurs and you’re just going around and collecting recyclables. That was the whole point of the game. That was my early career into, sort of, what I work on now.
Beverly: Were you like Mom, Dad this is it.
Chelsea: I know I was like; this is what I’m going to do.
Beverly: I love it.
Chelsea: I graduated with a degree in engineering. After that, I worked at a couple nonprofits on energy efficiency for residential communities, as well as starting and running a program for public schools in the state of Illinois. I did Consulting for seven years, mainly consulting for utilities on their energy programs, and then more recently I’ve transitioned to working more in both commercial buildings as well as transportation. A lot of my focus now is air, ocean, ground. How do we decarbonize those sectors, which is not going to be an easy thing to do, but we’re working through it. A couple cool things that we’ve done recently. We’re doing 1000-mile deliveries on EV semi tractors going from New York to Iowa.
Beverly: Interesting.
Chelsea: If you’re familiar at all with EV tractor, commercial EV space, the range is the big limiting factor. We’re able to do that with today’s technology, which is really cool. And then within the airspace, we’re making purchases of sustainable aviation fuel, which is jet fuel made from renewable resources like used cooking oil.
Beverly: Does it smell funny?
Chelsea: Yes, it does.
Beverly: It smells like a Chinese restaurant.
Chelsea: Yes, exactly. It does. When you go to the refinery, it doesn’t smell like a typical fossil fuel refinery. It does smell like used cooking oil, like a restaurant. And they’re taking the used cooking oil from your fries at McDonald’s and processing that to jet fuel that they’re using in planes today. Pretty cool.
Beverly: Super interesting. My longest gig ever was with Southern Company, Georgia Power.
Chelsea: Very cool.
Beverly: Also, I know quite a bit about tires and EVs, so I’m really looking forward to this conversation. My brothers are both long haul truck drivers. Yeah, yeah. I’ve got all kinds of connections here, but we’re talking about AI and renewable energy. Not all of our listeners know exactly what renewable energy really means. What does that mean?
Chelsea: Renewable energy is energy made from sources that are replenished more quickly than we’re using them up. It’s things like solar, obviously the sun is shining, and we can harness that to make energy, things like wind. Similarly, geothermal, which is a technology where you’re actually drilling down closer to the core of the Earth where it’s warmer and taking that energy and bringing it back up to the surface. And then using that for heating or cooling or whatever you’re doing with that. Hydro is another example of renewable energy, but it’s really those types of energy where you’re leveraging something that’s not a finite resource.
Beverly: So, what would be an example of something that is not renewable?
Chelsea: Fossil fuel. The sort of the oil and the natural gas reserves that are underneath the surface of the earth. Those are all finite resources. They’re not endless.
Beverly: OK. So, if we’re talking about AI in the world of renewable, how are these two concepts related?
Chelsea: Maybe it makes sense to take a step back and just talk about general concepts like what we mean when we say decarbonization, what we mean when we say low emission. A very simplified version of this for the future, a decarbonized world, is essentially we’re electrifying everything. That’s electric vehicles, that’s the energy usage in data centers, but it’s also our buildings, moving from natural gas heaters to heat pumps. So, we’re electrifying everything and then we’re building a bunch of renewables to power our world.
We have a lot of increased energy usage that’s anticipated over the next coming decades. And it’s from all of those things that I mentioned. It’s from electric vehicles, it’s from electrifying buildings, it’s from AI and data centers. There’s a whole lot more that we need to power with the grid. And what we want to do is try and make sure that that’s not all being powered by fossil fuels and creating a whole bunch of additional emissions which are then contributing to climate change. So, it’s really important that we are building renewables, that we’re building as low emission forms of energy in order to power all of these additional energy needs that we have as a society.
Beverly: And it’s going up more than the population, right? It’s more energy per person.
Chelsea: Yes, well, yeah. And I think we talked about a statistic for specifically AI and data centers that by the end of next year, the amount of energy required from data centers is going to be sort of equivalent to the whole energy usage of the country of Japan. It’s insane additional demand, that we’re adding to the grid.
Beverly: Yeah, it seems overwhelming. It is impossible. How in the world are we ever going to do this? And we’re not going to slow down with our AI. That’s like telling people to stop driving cars. We’re going to use AI; we’re going to drive cars. That’s very interesting. So, I’m vegan and sometimes some people are vegan purely for environment.
Chelsea: Environmental, yeah.
Beverly: And it’s hard for people to connect. Like what does veganism have to do with environmental? What does AI have to do with energy?
Chelsea: Energy consumption, yeah. When you think about computing, there’s obviously energy that’s used to power your computer. And so, data centers are doing that at hyper scale, right? It’s a huge amount of energy to do the actual computing that they’re doing. But I think the other factor, which is maybe not as well understood, is that it generates a massive amount of heat. You also notice that you have your laptop sitting on your lap. It gets hot.
Beverly: Your phone is too hot.
Chelsea: Exactly. And so, with that, you need to cool those server racks in order for them to operate effectively.
Beverly: Which is water.
Chelsea: A lot of times it’s water. A lot of the energy is for the cooling and it’s the water, the cold water that’s running through the pipes within a server rack farm. And that’s also why there is a lot of water usage that’s associated with AI and data centers, which again, I think the analogy is for every 100-word e-mail that ChatGPT writes, it’s like the equivalent of pouring out a 12-ounce bottle of water.
Beverly: Wow.
Chelsea: So it is, I mean, it’s a lot of energy and it’s a lot of water that are required to power AI, which is again, why it’s really important that we’re building renewables, that we’re looking at all of the efficiency opportunities that are available and trying to reduce the water consumption that’s associated with these server rack farms.
Beverly: That’s really stressful though, because the pace of the change in the technology space and then the pace of change from energy and water. I see that as two different paces. But before we go to that, let’s just pretend that I have a laptop and you have a laptop sitting on our laps and I’m over here going Dody Doo, I’ve got my excel sheet, simple stuff and you’re over there like generative AI stuff. You’re creating images and you’re getting summaries and you’re doing all this fancy stuff. So, you’re saying that that’s the laptop of the future and this is the laptop of the past?
Chelsea: Yes.
Beverly: And because of the changes in the way that we work and leveraging more AI and especially probably generative AI, the increase on a per person basis of the energy needed and because of the energy, the water, is incredible.
Chelsea: Yes, it is.
Beverly: It’s a lot.
Chelsea: And I will say there is discourse within the environmental community or those folks that are working on this, about what is the increase of energy actually going to be? I think there’s a lot of very high estimates and then there’s a lot of lower estimates. We’re probably going to land somewhere in the middle. But there’s a lot of smart people working on data center efficiency.
And even if you look at a data center of today versus a decade ago, they’ve made a whole lot of energy efficiency improvements in the way that they run things. And that’s going to account for some portion of that increased energy usage as well. So, I think that’s a positive within the industry. And I do think that that’s going to be a big factor within this whole conversation of how we match together the energy demand alongside the increased need for AI and what folks are interested in.
Beverly: Yeah. And so, battery power, that’s good or bad?
Chelsea: Within the context, it’s good. I was playing a thumbs up. So, with a lot of renewable energy resources, they’re not firm like what we would call firm in that the sun is not shining all the time. So, when you’re putting solar on your building, you’re generating energy in the middle of the day, but you’re not generating any energy at night. And so, you’re pulling from the grid at night. And so, what a battery can allow you to do is make the solar firmer. It’s still difficult for it to be as sort of reliable as a natural gas plant or nuclear, things like that, but it can help. And so, what the battery is doing is then taking the over generation that’s happening during the day, storing it and then you’re using it at night.
And that’s sort of the nice model for a net 0 building. When you think of a net 0 building, it’s like you’ve electrified everything. You have solar on the roof, you have a big battery in your yard, you’re using the solar during the day, the battery is charging and then at night you’re discharging the battery. So that’s one sort of method or way that we could get to this decarbonized world. And I think that there’s been a lot of advancements in battery efficiency. The prices are coming down. There are certain areas like state of California is really incentivizing that as a technology. So, we’re seeing a lot of battery installations there. But it’s all part of sort of the puzzle. Part of the solution.
Beverly: Yeah, it’s tricky because batteries are heavier and then it wears the tires out, right, and then the tires have to be replaced sooner, right. Particulars get in the air and it’s challenging that it’s not like a silver bullet anywhere.
Chelsea: No, there’s not a silver bullet anywhere. But I do think what is nice or reassuring, because I feel like it can get overwhelming quickly once you start thinking about some of these things, is a lot of climate scientists or technology folks have already done a lot of the work for us and they’ve done a life cycle analysis. So, we know we’ve taken into account, for a battery, here’s the amount of energy it costs to produce.
Here’s sort of the environmental impact of the mining, all of these things. And then we have all of those numbers, and we can compare it to here is the baseline and we can know for sure based on all of these factors, this is a better technology, this is more efficient, this is using less energy. Even if there’s just pros and cons to everything.
Beverly: Yeah, of course. Yeah, I think SK Battery, they’re a client of University of Georgia. My second hat that I wear, and they have a new plant that they’re building in Georgia, it’s going to employ 3000 more people. And so, there’s all this battery power this year from EVs. So, it sounds like LLMs and generative AI in general is kind of not good for the environment. What are we going to do?
Chelsea: Yeah, generative AI is definitely having a big environmental impact. I think that we as a society sort of need to decide where and how we’re going to use some of these tools. But I also just think we have to electrify a lot of other things. We’re kind of in a pickle, to be honest. We just need to do these, and we need to build renewables. We need to look at nuclear.
We need to look at all of these different options and figure out how we’re going to be able to decarbonize society in order to reduce the impact of global warming. It’s I mean, it’s a big problem. But we’ve built a lot in terms of the electrical grid, going from 1920 to 1940, it was like half of homes were electrified to 100%. We’ve done these big challenging engineering things before and I think we can continue to do those, but we need incentives. We need policy support.
Beverly: We’re not just going to do it out of the goodness of our heart.
Chelsea: Depends, but yeah, and I also think a lot of particularly with AI, a lot of these hyper scalers, Google, Microsoft, Amazon, they all have really aggressive climate goals, really aggressive. So, I believe all three, their goal is to be net 0 emissions by 2030. 5 years from now. I think they set these goals prior to sort of the whole big bubble related to AI, but they’re working really hard on it. I know an example that we talk a lot about within the environmental community is Microsoft, Microsoft’s plans to reopen 3 Mile Island, which was a shuttered nuclear plant, which is I think they have plans to do these different sorts of novel ways of bringing on energy.
I think that a lot of those companies they want to be able to offset their, the energy usage that’s coming from the cloud computing with renewables with firm renewable energy capacity that they’re adding to the grids. I mean, they’re the largest purchasers of renewable energy in the world. So, I think that there are solutions, it’s just going to take time and creative thinking, and we have to keep at it.
Beverly: Yeah, yeah. So, is there a way to responsibly use generative AI? Because what I’m picturing is I’m hearing these phrases from you; renewable, net 0, carbon footprint, I’m hearing oh, like we got to do this, that and the other. But is there a way to responsibly use generative AI? Let me give you an example. So, I’m building a, I mean I’m not building. I’m paying somebody.
I’m not good at stuff like that, but I’m building a house on the lake right now. And so, I’m trying to get some design ideas. So, I’m on this group that they’re always asking for design ideas. And so, I’m trying to help other people out and I’m going into ChatGPT and Sora and I’m like, OK, create this image and modify this and move this around to give people ideas. And I guess I did it too many times and I got a message back that says you’re asking for too many images. And I was like, is there a quota? Is there a limit like, do we have to be responsible in our Gen AI requests?
Chelsea: I find it difficult to give, because working in climate for all these years, people are like, what can I do?
Beverly: Yeah, exactly.
Chelsea: And I feel like it’s difficult to give.
Beverly: A punch list.
Chelsea: Yeah, just like you as an individual, if you stop doing this, we’re going to be fine. Structural changes are required. And I think the responsibility is both on the individual person, but also on corporates and the government and everybody has a big part to play in all of this.
Beverly: It’s going to take policy, isn’t it?
Chelsea: Policy, yes, exactly. And I do think with sustainability in general, AI does have a part to play. Specifically with trucking, I think with predictive maintenance, that can be a real great thing, especially for electric vehicles or more efficient driver queuing from AI. That’s something that we’re looking at. And from a logistics standpoint, optimization of our supply chain, leveraging AI to be able to do that, would be a huge benefit to our society. And there’s room for that. Believe me.
Beverly: Well, there’s two sides of this point. And so, I’m glad you transitioned to that because I was about to ask you. So, there’s the side of how much energy and how much water are we starting to consume now that we’re leveraging AI and especially Gen AI more. And then the other side of the coin that you’re getting at is how can we use AI to help us with some of our renewable energy and sustainability goals? And it sounds like you’re saying things like optimizing supply chain, predictive maintenance and knowing when to retire certain assets that are energy using assets, especially in manufacturing that can have big impacts. Are there other ways that we can leverage AI specifically to help us with our sustainability?
Chelsea: Yeah, I think buildings is a big opportunity. Just running your HVAC system optimally, smart lighting controls, things like that. I think that there’s a big opportunity there. I also think within the renewable sector, say you’re a corporate customer who’s put solar on the roof alongside a battery. It’s like you’re using AI in order to most effectively use the solar and then use the battery power. So, it’s not like you have somebody going and turning that battery on and off for you, that’s not efficient at all.
There are best ways to run these things. And so, AI can definitely help with that at a small scale, but then also at a larger scale, it can help utilities with grid, just like grid planning and trying to smooth the curve of peak demand. When they’re pulling from renewables, when they’re pulling from peaker plants, when they’re pulling from all these different areas, I think that AI can also assist with optimizing that. And I think it’s part, it’s just, it’s part of the solution as well. It’s like, I think that a lot of things within sustainability, it’s just about using your resources efficiently. It’s like being smart with what you have, and AI can help you potentially to do that.
Beverly: Yes, love it, love it. OK, Chelsea, what do you see for our future? We have data centers floating in the sky.
Chelsea: I have hopes. I have hopes for our future. I mean, I’m just hoping that we can decarbonize fast enough. And that’s what I’m trying to work towards. And, I think we will get there eventually, but it’s just going to take a lot of creative thinking and smart people. And we’ll have to keep at it because there’s a lot to do.
Beverly: Yes, I love it. Thank you so much to Chelsea Lamar from AIT Worldwide Logistics for talking to us about leveraging AI to optimize consumption of renewable energy. Thank you.
Chelsea: Thank you.
AI’s growing energy and water use raises urgent questions about its role in sustainability. When applied thoughtfully, AI can accelerate sustainability by optimizing energy grids and driving the transition to renewables. With the right choices and collective action, we can ensure AI powers a cleaner, more responsible future.
Explore the full catalog of TAG Data Talk conversations here: TAG Data Talk with Dr. Beverly Wright – TAG Online.