⚡AI Agents: Autonomous Payments in Energy Systems with Don Gossen
12:29 Share

Hey everyone and welcome back to AI Unraveled. Before we jump in, don't forget to hit that like button and subscribe to the channel. It really does help us out. And if you're feeling extra generous, consider donating to AI Unraveled to help us keep the show running. Now buckle up because today we're diving deep into the world of AI agents that can make autonomous payments. You heard that right, AI cutting deals on its own.

We're joined by Don Gossin, co-founder of Nevermind, a company that's been making waves with what they call the PayPal for AI agents.

They just snagged $4 million in funding. And Fast Company is calling them a game changer. Yeah, they've caught everyone's attention because they're tackling a really interesting problem in the energy sector. You see in the DACH region, that's Germany, Austria and Switzerland. There's this incredible boom in solar and wind energy production. It's actually created a rather unique challenge. OK, so too much clean energy. That sounds like a good problem to have, right?

Especially in Central Europe, in particular the DACH region, Germany, Deutschland, Austria, and Switzerland, over the last five to ten years, there's been a real push towards the adoption of...

solar panels and wind turbines at a local retail level, right? So to the point where in this region and in places like Germany and Austria in particular, solar panels are so cheap that people use them to build fences. What? Solar panels for fences? That's wild. It sounds like they're practically giving away solar power over there. It really does. And all this abundance is creating some unexpected issues for their traditional energy grid.

It's not really set up to handle this much decentralized power generation. So instead of this being a win-win for clean energy, it's actually costing people money. You'd think more clean energy would mean lower costs, but it's not that simple. The grid isn't flexible enough to absorb all the excess energy when the sun is shining or the wind is blowing. In some cases, energy producers are actually having to pay to get rid of their surplus. Okay, so that's where Don and his company, Nevermind, come in.

They're using AI to solve this problem. But how exactly does that work? Instead of relying on these massive centralized power grids, Don and his team are building a system where AI agents powered by blockchain technology manage energy distribution at a hyperlocal level.

Think your neighborhood, not your entire state. So instead of sending all this excess energy out to the main grid where it might go to waste, they're finding ways to use it right there in the local community. Exactly. Imagine your neighbor's solar panels are generating more energy than they need. Instead of that energy being wasted, an AI agent connected to your AC unit could automatically purchase that excess energy at a discounted price. And so what this agent needs to do from time to time is get more energy

a more accurate picture of what's actually going on. It needs closer to real-time data, right? It needs to understand the weather. It needs to understand the production that's going on around it, right? So it needs access to newer data sets that requires it to basically say like, okay, historically, you know, I should probably be asking for new data at this point. Or maybe it's getting like a real-time data feed from, you know, about the, you

the weather right, and realizes, oh, it's sunny. I should check and see how much energy is being produced by my neighbor or my local area network. Hold on. So you're telling me my AC could be making its own decisions about when to run based on how much solar power my neighbor is generating. That sounds incredibly futuristic. Yes, doesn't it? And it's all powered by something called retrieval augmented generation, or RA for short.

This is how the AI agents make those smart decisions. They're constantly analyzing data like weather patterns, historical energy production, and real-time energy availability. So it's like the AI is learning from past trends and then using that knowledge to anticipate future energy needs. Exactly. But it goes a step further. It doesn't just rely on data. It's already collected.

When it determines it needs more information to make an informed decision, like realizing it's a sunny day and probably time for excess solar energy, it can actually go out and find that information. Wait, so the AI is proactively searching for data. That's amazing. It is. It highlights how these AI agents are moving beyond just reacting to commands. They're becoming much more intelligent and capable of independent action.

But there's still the question of how these AI agents actually pay for that energy they're buying. Right, because we're talking about tiny amounts of energy being traded back and forth. It's not like your typical monthly energy bill. Exactly. And that's where things get even more interesting. This is where blockchain technology enters the picture. Remember how Don called his company's tech the PayPal for AI agents? Right. So it's handling all these microtransactions for energy, kind of like a digital Venmo for AI. Precisely.

But traditional payment systems just aren't built for these kinds of lightning fast, minuscule transactions. They'd be way too slow and expensive for something like this to work. So what makes blockchain so special in this scenario? Well, think of it like a digital ledger that everyone on the network can see and verify.

Every transaction is recorded and secured. And because it's decentralized, there's no need for a middleman like a bank. OK, that makes sense. But wouldn't these constant microtransactions add up and clog up the blockchain? That's where Don's innovation really shines. They're using a specialized type of blockchain called a DEPIN, which stands for Decentralized Physical Infrastructure Network.

It's specifically designed to handle these kinds of microtransactions for managing physical assets like energy grids. So instead of one giant blockchain trying to do everything, there are these smaller, more specialized blockchains focused on specific tasks. You got it. And that's what makes this whole system so efficient. Imagine a local energy marketplace where your AI-powered AC can buy energy directly from your neighbor's solar panels, all tracked and managed on this secure network.

dip in. It's like a neighborhood energy barter system, but automated and super fast. This is incredible. We've got AI analyzing data-making decisions and then blockchain swooping in to handle these super fast, tiny payments. It's like watching a perfectly choreographed tech ballet. It really is an amazing symphony of technologies working together. And it gets even more impressive when you consider the collaboration involved.

This isn't just one company working in isolation. Don described it like building with Legos, where each company brings a specific piece of expertise to the table. So it's a real team effort. Who else is involved in this project? You have Olaz providing the framework for the AI agent itself, Peak, providing the decentralized network for those transactions to occur. Combiner supplying that crucial real-time energy data through their API and Kiko, building the interface that connects with the physical devices like your AC unit.

And of course, you have never mind acting as the financial glue that holds it all together. You have to remember that we are talking about a project that was initially conceptualized in, you know, in December of 2022. And it was live by the end of January, beginning of February of 2023.

And with the whole project went from concept to launch in just three weeks, that's unheard of, especially for something this complex. Exactly. It really highlights how rapidly this technology is evolving and how quickly these solutions can be deployed. And the implications go far beyond just managing energy in a single neighborhood. OK, where do we go from here? What's the next step for this technology? Well, Dawn has a vision for the future where these individual AI agents become part of larger interconnected swarms. And warms like a hive mind for energy management. In a way, yes.

Imagine a network of these AI agents all working together to optimize energy usage across an entire neighborhood, a small city, or even larger regions. That's the power of these multi-agent systems, or swarms, as Don calls them. Yeah, I mean, I think the big thing that's coming up are agents, right? Agetic AI agents. Yeah. So you alluded to what OpenAI is alluding to coming down the pipe. And then next up, sort of what we're demonstrating in our use case is one agent.

But these are going to work with other agents, right? With other air conditioning units and other devices. So really what we're seeing is the initial growth of multi-agent systems or AKA, they're also called swarms. It's like we're on the verge of a whole new era of intelligent energy management where AI is working behind the scenes to make everything more efficient and sustainable. Exactly.

And these AI agents will operate largely behind the scenes, just like the algorithms that already power many aspects of our digital lives. You won't even notice them working, except maybe when you see your energy bill go down. That's the dream, right? Lower energy bills and a greener planet. But I have to admit, the idea of AI managing something as crucial as our energy grid can be a little unnerving.

What about security? What's to stop someone from hacking into these AI swarms and wreaking havoc? Yeah, it's a valid concern and something Don addressed head on. And he pointed out that every system has vulnerabilities and we shouldn't let fear of the unknown stifle progress. It's like saying we shouldn't build cars because they could get into accidents. Instead, we focus on making them safer as we learn and adapt. So it's about managing risk, not eliminating it entirely. Exactly. And keep in mind, the existing energy grid is already vulnerable to cyberattacks.

Adding AI and blockchain to the mix doesn't necessarily make it inherently less secure. In fact, these technologies could even enhance security in some ways. For example, blockchain's decentralized nature makes it much harder for hackers to compromise the entire system at once. That's a great point. So maybe it's not about fearing the unknown, but more about embracing the potential while also being aware of the risks? Precisely. But let's be realistic.

There will be challenges along the way as this technology matures. The energy sector, especially when it comes to critical infrastructure, tends to be risk averse. Which makes sense. We're talking about something that powers our homes, businesses and hospitals. You can't just flip a switch and change everything overnight. Right. So Dawn predicts that the adoption of these AI powered systems will likely happen gradually. It might start with less critical applications like managing energy in individual buildings or neighborhoods.

Then, as the technology proves itself reliable and secure, it could expand to larger parts of the grid. So maybe we won't have entire cities run by AI swarms tomorrow, but it's definitely something we could see within our lifetime. Absolutely. And it all starts with these pioneering projects like the one Don and his team are building.

proving the concept, refining the technology, and building trust in these AI-powered systems. This has been an incredible deep dive. We've gone from solar panels being used as fences to AI agents cutting energy deals all in the span of a few minutes. It's mind-blowing stuff. It really is. And it highlights how innovation often emerges from unexpected places. Who would have thought that an abundance of clean energy would lead to the development of AI-powered microtrading on a blockchain?

It shows that even complex problems can be tackled with a combination of ingenuity and cutting-edge technology. So what does this all mean for our listeners? What's the key takeaway they should walk away with? I think the key takeaway here is that the future of energy is intelligent, adaptable, and surprisingly local.

We're moving away from these massive centralized systems and towards something more distributed, more responsive to individual needs. And that shift is being driven by the convergence of AI and blockchain working together in ways we could only have dreamed of a few years ago. It's almost like we're handing over the keys to the energy kingdom to these AI agents. But is that a good thing? Is it something we should be excited about or a little apprehensive? It certainly prompts some important questions. But think about it this way.

You already trust algorithms to recommend movies, filter your spam, and even drive your car in some cases. This is just another step in that direction, using AI to manage something that's essential to our lives. True. Maybe this isn't such a huge leap after all, especially if it results in lower energy bills and a more sustainable grid. Exactly.

It's about progress, but progress with a mindful approach. Well said. So what's the one big question you want to leave our listeners pondering after this episode? If AI can manage something as complex as energy distribution, what other seemingly intractable problems in our lives could be solved with these kinds of invisible AI-powered systems? Ooh, that's a good one. Something to really think about.

Well, folks, that wraps up another fascinating deep dive into the world of AI. We've explored how AI agents are poised to revolutionize energy management and maybe even solve some of the world's biggest challenges. If you enjoyed this episode, please like, subscribe and consider donating to AI Unraveled to support the show. Your contributions help us bring you these insightful conversations with the brilliant minds shaping the future. Until next time, keep exploring, keep questioning and keep learning.