Navigating AI Challenges in Climate Policy

Welcome to another edition of the Centre for Sustainable Intelligence newsletter.

This is where I try to explore the intersection of artificial intelligence and environmental sustainability. My mission is to share my observations across these core themes: AI for Good, Ethical AI, Sustainable AI and AI in Sustainability. I hope to inform, inspire and engage readers interested in the potential of AI to drive positive change for both people and the planet.

Here’s what’s in stock for you today:

• News Bite: Deepfakes everywhere - Let AI fight AI.

• Deep Dive: Navigating AI Challenges in Climate Policy.

• Tech Trend: Prompt Engineers out, AI Architects in.

News Bite: Deepfakes everywhere - Let AI fight AI.

With the US election just around the corner, it’s wild to see how AI is being used to mess with voters. Deepfakes are popping up all over the place, and some vendors are stepping up to tackle this head-on. For instance, Pindrop has developed an audio deepfake detector called Pindrop Pulse Inspect that’s designed to spot when a deepfake is being used to interfere with the elections. Just recently, a deepfake of Vice President Harris and Elon Musk was making the rounds on X (formerly Twitter), which is pretty scary. But tools like Pulse Inspect can analyze digital audio files and detect AI-generated speech, which is a big deal.

According to Lisa Martin, an analyst from the Futurum Group, these kinds of tools are crucial in today’s world of generative AI. She mentioned that they’re basically "fighting AI with AI," which is pretty clever when you think about it. And it’s not just about the elections—there’s a real fear that deepfakes could be used to create digital twins of us, making scandalous statements and damaging our reputations. Hopefully, tools like this will step in to save the day!

Deep Dive: Navigating AI Challenges in Climate Policy

If we really want to make the most of AI in tackling climate change—and do it in a way that’s ethical and inclusive—we need to face some big challenges head-on. A lot of these issues come from us getting a bit too carried away with what we think AI can do, or maybe even giving it more credit than it deserves. But don’t worry, I’ll walk you through each of these challenges so you can see what I mean. Ready to dive in?

Let’s go!

1. Data: The Fuel for AI That’s Often in Short Supply

First off, let’s talk data. AI thrives on data—the more, the better. But when it comes to climate policy, data availability is a huge issue. Imagine trying to solve a puzzle with half the pieces missing; that’s what it’s like for AI when it doesn’t have complete data.

For example, key data like satellite images or emissions figures can be hard to access. Sometimes they’re locked away by governments or companies, and other times they’re just scattered all over the place. Plus, the data that is available can be pretty uneven. Developed countries often have more detailed data than developing ones, which leads to AI models that don’t reflect the global picture. And then there’s the technical side—some regions lack the infrastructure to even collect or process this data properly. Without solving these issues, we risk AI offering solutions that aren’t as effective or inclusive as they should be.

2. One-Size-Fits-All AI Just Doesn’t Cut It

Another biggie is that AI isn’t always great at understanding the complexities of climate policy. Climate issues are deep and tangled—like the idea of a “just transition,” which aims to shift from fossil fuels to clean energy without leaving workers or communities behind. Generic AI models might miss these nuances, leading to recommendations that don’t fully grasp the stakes or the sensitivities involved.

That’s why it’s crucial to train AI on climate-specific data. Different countries and regions talk about climate in different ways, using different terms and frameworks. If we don’t tailor AI to these specifics, we could end up with solutions that work great on paper in one part of the world but flop in another.

3. Let’s Talk About Ethics

Then there’s the whole ethical side of using AI in climate policy. AI can be biased—if it’s trained on skewed data, it’ll give you skewed results. In the context of climate, this could mean policies that unintentionally worsen inequalities or just miss the mark entirely.

And then there’s the issue of trust. AI can “hallucinate”—basically, make stuff up. Imagine relying on AI for a crucial climate decision only to find out it fed you inaccurate info. That’s a nightmare scenario. Plus, AI is often a bit of a black box—it spits out results without explaining how it got there. For climate policy, where the stakes are high and the impacts widespread, we need to know the why behind AI’s suggestions.

4. AI Should Assist, Not Replace, Human Judgment

AI is here to help, not to take over. That’s something we can’t forget. There’s a lot of talk about “augmented intelligence,” where AI complements human decision-making. But there’s also a risk of over-reliance. AI can crunch data faster than we ever could, but speed doesn’t always equal quality. If we start relying too much on AI without double-checking its work, we could end up making some serious mistakes.

5. Bridging Language and Cultural Gaps

Climate change is a global issue, but AI models often struggle with the diversity of language and culture. Climate policies are written in different languages and shaped by different cultural contexts. AI needs to be able to understand all of this to be truly effective. But simply translating words isn’t enough—AI also needs to grasp the underlying meaning and context, which is no small feat, especially with technical or legal language that doesn’t translate neatly across cultures.

6. Inclusivity and Equity: The Final Frontier

Finally, there’s the challenge of making sure AI-driven climate policies don’t leave anyone behind. There’s a real risk that AI could widen the digital divide, especially between developed and developing countries. If the best AI tools and the most useful data are concentrated in wealthier nations, the benefits of these technologies won’t be shared equally.

It’s also crucial that AI considers the perspectives of underrepresented communities. These groups need a seat at the table when AI tools are being developed and used. Otherwise, their needs might be overlooked or misinterpreted, which could lead to policies that don’t address—or even worsen—the challenges they face.

Wrapping It Up

So, there you have it. AI has incredible potential to help us tackle climate change, but only if we address these challenges. We need to ensure that data is accessible and comprehensive, tailor AI to the specific needs of climate policy, and keep a sharp eye on the ethical implications. And, most importantly, AI should empower human decision-making, not replace it, while ensuring that the benefits are shared equitably across all regions and communities.

What do you think? Are there other challenges you see that we need to tackle? Would love to hear your thoughts!

Tech Trend: Prompt Engineers out, AI Architects in.

Remember how we were all going crazy about prompt engineering not too long ago? Everyone was talking about how it was the hot new skill with sky-high salaries, and how we all needed to jump on the bandwagon. Don’t get me wrong, it’s still super important, but it looks like there’s a new kid in town: the AI Architect.

I’ve been noticing that the role of an AI Architect is becoming way more critical and complex than we initially thought. Building solutions with LLMs isn’t just about plugging in AI capabilities; it’s really about building entire solutions from the ground up, understanding how to piece together different architectural patterns, knowing the best practices for cloud-native applications, and making smart choices about things like databases.

In a way, it’s like being an IT Architect or a Solution Architect but with a focus on AI. You’re not just handling one part of the process; you’re orchestrating everything—AI integration, prompt generation, automation, platform choice, architectural designs and much more. I have a feeling this is going to become one of the most sought-after roles in AI soon. Let’s see how it all plays out!

That’s all I have for you today.

That’s all I have for you in this edition of Centre for Sustainable Intelligence.

If you have any questions or feedback please use this form. I will try my best to respond to all your questions and feedback.

If want me to feature an article in the News Bite or Tech Trend sections, please get in touch and I will see what I can do to feature it.

If you are not yet subscribed to this newsletter, please subscribe here and forward it on to all your friends who you think will benefit from it.

Thank you!

Emeka Ogbonnaya