It's really hard to run a company. Any company really, but especially if you're one of the most influential startups in Silicon Valley. Case in point Open A.I.. This is the company behind Chat GPT. And in the last nine days, the company has seen turmoil. The chat bot itself would have a hard time scripting.

This morning, a blockbuster shakeup in the world of artificial intelligence.

'Just here's a quick recap On the 17th. The company's board abruptly fired its CEO and co-founder, Sam Altman.

The board saying that he's not equipped to lead the company right now. He hasn't been very transparent.

'Another co-founder, Greg Brockman, then quit in protest. They both went over to Microsoft, which is open his biggest financial backer, to lead a new artificial intelligence research division. In the meantime, open eyes cycled through two more CEOs and faced a staff revolt over how they handled the whole thing. And then the company reversed course and announced it was bringing back Sam Altman with a new board. In short, this was a freaking mess.

This all matters because artificial intelligence is here. It is the future. There are serious conversations about whether or not it's moving too fast. Can it be weaponized? What does it mean?

And that's left some asking if a company like Open I can't execute basic h r moves correctly, why should they be trusted with technology that promises to reshape how we live and work? Question is on my mind as we start a week that will see the beginning of the annual U.N. climate conference known as COP. COP 28 will be held in Dubai. And while the stakes are always high for these conferences, recently the warning signs of climate change have been flashing brighter than ever. For example, the Earth's global temperature was recently two degrees Celsius hotter than pre industrialization levels. It hit the mark only briefly, but that two degree limit is the big red number scientists say we have to avoid at all costs if we want to continue living on this planet.

This year, the last 12 months, hottest in the last 125,000 years at least. We've shattered every record imaginable. So the trend line is unmistakable that the planet is overheating.

So as world leaders gather around the table to talk about solutions. Is A.I. a possible answer? My guests this week are CNN's Claire Duffy, who covers business and technology, and CNN's Rachel Ramirez, who's on our climate team. We're going to talk about the possible uses for AI in the fight against climate change and why it could actually exacerbate some all too familiar problems. From CNN, this is One Thing. I'm David Rind.

I guess, I want to start here. We've covered on this show before and a lot of the conversations been around these dark warnings about how it's going to, you know, steal jobs, fuel disinformation, and maybe one day rise up against the human race, depending on who you talk to. So why are we talking about it as a potential tool to help fight climate change, which actually does have these end of the world type stakes attached to it?

Well, first, it's definitely not a silver bullet, right. To our enormous climate problem. But just to set the foundation, the bottom line is we're seriously running out of time to really shift the trajectory that we as a planet are heading to avoid irreversible harm. So the science is really showing that a lot of the pollution that we've emitted as a society is already locked into the system, meaning that a lot of the consequences that we'll see are now inevitable.

It's like baked in, it's going to happen.

Yes, Yes. And so it's those emissions that we can still kind of dial back that we really need to address and work on. So we need all the updated science, all the big data and all this like cocktail of solutions to really get back on track.

Yeah, I mean, I think it's that speed part that Rachel's talking about. AI has these massive predictive capabilities that can process these huge amounts of data much faster than traditional computing. With conventional computing, what you're doing is essentially saying, here's this selection of data, here's this input, please give me this specific output with AI. What you can do is say, please consider this whole universe of information and tell me what I should be thinking about it and what questions I should be asking about it. It's meant to sort of reason in a way that a human could. And some of the folks that we talked to said that it actually part of the acceleration that can happen here is you remove people from the equation. In some ways you say.

Constraints of the human brain.

Yeah, You don't have to have researchers who need food and sleep. Humans can't be removed from the equation altogether. But you can say here's a hundred years worth of data on, you know, a hundred different things. Please tell me about this. What kinds of questions should we be asking to direct our research?

So what are some potential real world examples of how I can be used in the space?

So climate science experts say there are essentially like three main buckets to think about in terms of trying to address climate change and that I can help with all three. So in the first bucket you have trying to change, as Rachel said, our current way of operating so that we don't make things any worse. And so we learned about this one example. Google's DeepMind Research Institute is using artificial intelligence to sort of predict the availability of wind energy. They trained this AI model on weather forecasts and historical wind turbine data, and it allowed the system to make advanced commitments of wind, wind energy to know when that wind was going to be blowing. And in that way it made the wind energy more valuable.

So that's an example in bucket one. But what about figuring out how to mitigate or reverse the impacts of climate change?

So for mitigation, a really key part in that is just thinking about how we can expand our climate science research. So there's the scientist that I talked to named Anna based in Alaska, and she's studying how permafrost melts and the Arctic. So she is using AI to now make seasonal forecasts on permafrost melt instead of predicting sort of the typical 100 year timescale that a lot of climate researchers do. And so in this way, it analyzes the historical data, the present data, and also kind of analyzes all the future scenarios of what will happen if the world emits X amount of emissions.

And so what about you mentioned adapting to the effects of climate change that are already here. How is AI influencing that area?

So one of the people that we talked to is this founder of a startup based in Silicon Valley called Climate A.I.. And Climate A.I. is basically developing this tool using artificial intelligence to evaluate how exposed crops could be to warming temperatures over the next two decades.

Oh, yeah. So so basically, start with my story. I mean, I come from a very small downstairs village in the north of India.

And this guy, Himanshu Gupta, is his name. He grew up in a small village in India where drought was often a familiar visitor.

When they were drought, my mother would take on all of us for siblings and for like a one kilometer walk would anybody's ever imagine.

So that basically inspired him to kind of develop this tool. And so the tool that they created is now an app. Engineers basically are able to extract the climate. The. Water and soil characteristics of a certain location that's been plagued by climate disasters.

What they can do is they can come into a black hole and go to a specific location, drop in on the platform. And also, as you buy, I'm growing tomato seeds.

And it measures how viable the landscape is and or would be in the coming years so that farmers could then kind of adjust their plans, their farming plans accordingly.

Or figure out if it's even going to be farmable at all? Yes. And then they would consider going somewhere else if it wasn't.

Yeah. So one example specifically is in Maharashtra, India, where tomato producers, they were really worried about this major crop loss that they've been seeing because of the extreme drought there.

The platform that automatically extracts the climate, water, soil, location, growing tomato seeds and then match it with in the next ten years due to climate change. Are the new locations coming up where this company can move and start growing classes?

'And for this state climate, I found that extreme heat and drought would lead to a roughly 30% decrease in tomatoes seed output in the region within the next two decades. And then they warned the farmers that the location might prove, quote, non-viable in the coming years. With that information, they adjusted their plans for farming.

This process takes about three years for them to pick it up again, up in then a couple of minutes.

So that all sounds pretty great, honestly. But like I alluded to at the top is like extremely unregulated at this point in time. And lawmakers have a hard enough time, especially in this country, passing climate policy without any other considerations attached to it. So just how willing are the people in charge of making these important decisions going to be about embracing this as a possible solution?

So COP is in Dubai this year, and there's going to be a lot of contentious issues that will be on the table that will be negotiated upon. And one of the things that will be a cop this year in the first few days will be this thing called the Climate Innovation Forum. And it's basically a curated selection of different tech themes, including big data and artificial intelligence, which I assume will intersect with a lot of the discussions on climate solutions at COP.

Yeah. And I mean, I think beyond even climate change, we've seen huge adoption of this A.I. technology. It's happened even faster than proponents of the technology thought that it would. It's really been exploding. And then, you know, as you said, David, you start to see these regulators go, oh, my gosh, we need to wrap our heads around this. There are lots of things for them to think about. But one thing that hasn't been talked about as much in terms of regulation that climate scientists say is really important to consider is the potential impact of AI on the environment. There are all of these possible benefits that I could have in terms of the fight against climate change, but at the end of the day, the thing that makes AI work is essentially these warehouses of really powerful computers. And as you can imagine, that requires tons of energy as well as a lot of water for cooling. And so it's sort of this race against time for aid to help solve climate change or address climate change while it's also making the problem worse.

So it's sucking up a lot of energy just to run these programs. And that could also be contributing to the climate problem as well.

Exactly. Yes, it is. You know, I is going to have to try to figure out how to solve this problem that it, too, is contributing to. And there are a number of thoughts about how this might be addressed. Regulators could require that AI data centers run partly on renewable energy. That's something that's been talked about with Bitcoin. Some data center operators are also talking about locating their data centers in places where it's already cold like Scandinavia, so that they don't have to use as much water for cooling. There is like this worry about who is going to have access to this technology. Is it just going to be the Googles and the Microsofts and the open eyes of the world, or are we going to spread this around to more people so that everybody can be using AI in this way that could be potentially beneficial?

'I was going to actually touch on that, too, because the fight against climate change, you know, a lot of climate protesters and climate activists have been saying this wouldn't be successful without climate justice. Right. And a part of climate justice is accessibility and affordability. And with the farmer's situation example that we talked about, a lot of the farmers that have access to the Climate Air tour are farmers that are well-off. What about farmers that don't have access to internet and can't just log in on their phone and access this type of information.

So it could exacerbate inequalities...

Exactly.

That have been going on for decades as part of this climate discussion already.

Right. So we need to think about how do we scale up AI in order to also cover all those communities.

And I think to sort of just bring us full circle to the beginning of this discussion. David, you were talking about these sort of apocalyptic predictions that people are making about AI, and there are concerns that those kinds of conversations are saying there's going to be a computer revolution that takes out humanity, are actually distracting from these much more immediate and real concerns and conversations that people should be having about how this technology is being used and who has access to it.

Thank you both. Appreciate it.

Thanks.

Thank you.

And if you want to read more of Claire and Rachel's reporting, you can check it out at CNN's Road to COP, which is online right now. We'll leave a link to it in our show notes. One Thing is a production of CNN Audio. This episode was produced by Paola Ortiz and me, David Rind. Our senior producer is Faiz Jamil. Our supervising producer is Greg Peppers. Matt Dempsey is our production manager. Dan Dzula is our technical director. And Steve Lickteig is the executive producer of CNN Audio. We get support from Alex Manasseri, Robert Mathers, John Dianore, Lenny Steinhart, Jamus Andrest Nichole Pesaru and Lisa Namerow. Special thanks to Hayley Thomas, Lindsay Reed and Katie Hinman. We'll be back on Sunday with another episode. Talk to you later.