Donald Gossen is the Founder & CEO of Nevermined, he is a a Data and Analytics professional with extensive global experience, currently focused on advancing decentralized AI technologies. His expertise spans Web3, Data, AI, and Analytics, including large-scale Data Federation, Machine Learning, Federated Learning, Big Data, and Digital Transformation. Before launching Nevermined, Don co-founded Keyko and Ocean Protocol.
Nevermined is shaping the future of AI-driven commerce by providing a platform where autonomous AI agents can conduct transactions without human involvement. Positioned as a key player in this emerging space, the platform offers secure and scalable payment infrastructure to support a vast network of AI agents, paving the way for new possibilities in commerce.
Nevermined has been described as the “PayPal of AI Commerce.” Could you explain what this means and how your platform supports this new era of AI-to-AI transactions?
When we say we’re the “PayPal of AI Commerce,” we’re talking about revolutionizing how AI agents pay and get paid when they interact with one another. We’ve built a unique credits-based system that allows AI agents to make autonomous payments for services, which is crucial for the development of multi-agent systems. Our platform is designed to be flexible, scalable, and compatible across blockchain networks and AI ecosystems. We’ve abstracted away the complexities of Web3, making it accessible to both crypto-native and traditional users. What sets us apart is our focus on the entire payment process – from metering AI agent usage to calculating charges and settling payments.
We’re educating the market that AI Commerce is a new standalone category that requires new payment standards. Agents are a sum of the parts of the architecture that are closer in nature to software platforms with a set of tools and services that can be called upon. These platforms handle input requests by dynamically calling on the optimal range of services, the optimal response path. As how the request is handled varies, so does the associated computational load and cost. It’s for this reason we identify AI payments as a process, not simply a settlement; a process of tracking and metering usage of each individual tool or service employed with each request and translating this into a charge. The protocol ensures efficient and optimal monetization of AI agents, something totally lacking in the current market.
Our vision goes beyond current payment paradigms however. We’re anticipating a future where AI agents might directly barter services without transacting in traditional currency, an expectation that is reflected in our current design choices. This forward-thinking approach, coupled with our team’s extensive experience in Web3 and AI technologies, positions us to capture significant market share in the rapidly evolving AI commerce landscape. We’re not just theorizing – we’ve already built the solution, and it’s in production. We’ve partnered with 19 organizations, including major players in the Web3 space. Our immediate focus is on B2B, supporting partners building multi-agent systems, but our long-term vision is centered on AI agent to AI agent usage. We believe our modular approach and focus on payments as the critical chokepoint for AI agents gives us a significant advantage. We’re creating an entirely new industry and corresponding commercial system, and we’re excited to be at the forefront of this revolution in commerce.
As a co-founder of both Ocean Protocol and now Nevermined, how did your experiences shape the vision for Nevermined? What inspired you to build a payment infrastructure for AI agents?
My journey with Ocean Protocol back in 2017 was really eye-opening. It was one of the first projects to explore the intersection of Web3 and AI, and it gave us invaluable insights into the challenges and opportunities in this space. What we learned at Ocean, combined with our decades of experience in ML and Big Data, led us to recognize a critical gap in the emerging AI ecosystem – the need for a robust, flexible payment infrastructure tailored specifically for AI agents.
The inspiration for Nevermined came from our realization that as AI agents become more autonomous and sophisticated, they’ll need a way to transact and exchange value that goes beyond traditional payment systems. We saw that the future of AI commerce – especially agent-to-agent interactions – would be vastly different from human-centric markets. This insight drove us to create a system that could evolve with the rapidly changing AI landscape.
Our experience taught us the importance of modularity and focus. Unlike the monolithic approaches we’ve seen fail, Nevermined is laser-focused on solving the payment piece of the AI puzzle. We’ve abstracted away the complexities of Web3 to make our solution accessible to both crypto-native and traditional users. This approach, combined with our deep understanding of both Web3 and AI technologies, gives us a unique advantage in building the financial infrastructure for the future of AI.
Your platform enables AI agents to autonomously conduct financial transactions. How does Nevermined’s hybrid protocol ensure secure and efficient handling of these transactions?
At the core of our system is a unique credits-based implementation. These credits are specific to each AI agent or set of agents and represent claims on the agent’s services or computational power. This approach allows for more flexible and AI-specific transactions compared to using standard currencies or stablecoins. Importantly, this architectural pattern lays the foundation for a future state of agent bartering, where agents trade claims on each other’s services directly without the need for a common currency or medium of exchange. We also emphasize that AI agent payments are a process and not simply a settlement, which we regard as the end state of the process when the calculated charge is settled. We’ve designed our protocol to handle the entire payment process – from metering AI agent usage to calculating charges and settling payments. This comprehensive approach sets us apart from competitors who might focus solely on the settlement aspect.
What really makes our system secure and efficient is how we’ve abstracted away the complexities of Web3 while still leveraging its benefits. We’ve implemented full account abstraction, allowing non-Web3 users to use Nevermined with their social logins instead of connecting a crypto wallet. Under the hood, these users still get their own non-custodial wallet via an MPC solution. We’ve also made the system gasless, eliminating the need for users to sign transactions. Plus, we’ve integrated fiat payment options through Stripe, allowing users to buy credits with a credit card without touching crypto. All of this combines to create a secure, efficient, and user-friendly system that caters to both crypto-native and traditional users, making it an attractive solution for AI builders and users across the board.
Can you break down some of the key technical challenges that your team had to overcome to allow AI agents to seamlessly “pay and get paid” without human intervention?
When it comes to the key technical challenges we’ve had to overcome to enable AI agents to seamlessly pay and get paid without human intervention, it really boils down to creating a system that’s both flexible and secure. One of the biggest hurdles was developing our unique credits-based implementation. These credits are specific to each AI agent or set of agents and represent claims on the agent’s services or computational power. This approach allows for more flexible and AI-specific transactions compared to using standard currencies or stablecoins. We had to design our protocol to handle the entire payment process – from metering AI agent usage to calculating charges and settling payments. It’s a comprehensive approach that sets us apart from competitors who might focus solely on the settlement aspect.
In order to fully understand the nuance of AI agent payments and the flexibility afforded by our credits architecture, you must understand how AI agents work. Each AI agent, from a technical point of view, crudely resembles a platform with a set of tools that can be called upon to service requests. Which tools are invoked when provisioning a response is dependent on the complexity of the input request. AI agents invoke a variable set of services to respond to the input, and this carries an associated variable cost that must be calculated and transferred to the requestor. This is unlike contemporary platforms, which take relatively static inputs (i.e. highly formatted queries) that limit the complexity of the interface, and respond based on predefined tool selections. Accounting for this variability efficiently and accurately is non-trivial, and not something most AI builders are familiar or comfortable with.
To account for the dynamic nature of AI agents’ capabilities, Nevermined employs an established usage design for software platforms called License Tokenization. Not to be confused with crypto tokenization, License Tokenization is a licensing model that enables flexibility and modularity by design. More traditional usage designs require platform users to specify who will use each tool (Named User License) or how many will use each tool (Concurrent Access License), as well how many times and / or how long each tool will be used. The explicit nature of these license designs makes them cumbersome to set up, maintain and enforce, requiring complicated agreements and administrative software.
License Tokenization, on the other hand, simplifies the process by enabling access to a system or platform through the acquisition of a set of redeemable tokens / credits. Each tool within that system is set with its own discrete redemption criteria, managed via tokens, so that when a user engages with a specific tool, the requisite number of tokens for that specific tool are redeemed from the user’s bucket of tokens. This allows any user or set of users to access any tool within the system or platform, so long as they hold enough tokens for the specific tool that they are using.
AI agents are architected in a similar fashion to a platform, consisting of a set of tools, like a series of models (i.e. the GPT series from OpenAI), that can be called upon discreetly, or in combination, to source a response based on a user request. To enable engagement flexibility for each tool, Nevermined employs a License Tokenization design. This means that a user can gain access to an AI agent by acquiring that agent’s tokens / credits. Once this is done, the user can make requests to the agent which are in turn routed along the optimal response path among its underlying tools and services, each with their own specific redemption criteria. To the extent that one response is more computationally intensive than the next, the mix of employed tools will differ, and thus also the number of tokens redeemed.
Where Nevermined differs from traditional License Tokenizations schemes is that we have brought the tokens / credits on-chain, instead of managing the system of unit accounting in a centralized, closed database. This creates a potential hybrid approach to metering usage, with the possibility of on-chain credits being managed by an off-chain metering system that is agent specific. The added advantage of bringing credits on-chain is the ability for the agent’s service mix, and thus computational power, to be uniquely represented by freely tradable on-chain claims. This commoditization of the agent through uniquely redeemable claims is a critical factor in unlocking the potential of direct AI-to-AI commerce.
In a contemporary License Tokenization system, most platforms would have their own usage tokens managed within their own, centralized unit accounting database. In fact, this is what Metronome does for OpenAI, Databricks, etc. This means that OpenAI has its own set of tokens independent from Databricks’ set of tokens. A problem with this centralized design is that, should multiple service vendors want to combine their tools within a given agent (i.e. if wanting to combine OpenAI and Databricks tools and services to make an agent), a net-new set of tokens and system of unit accounting would need to be set up each time this combinatorial approach takes place. This adds an entirely new set of complexities to AI agent development given that each vendor needs to negotiate with the other vendors, and then the system of unit accounting either needs to be collectively administered by all counterparties, or entrusted to a single counterparty.
Nevermined’s approach provides far greater flexibility and transparency to the ecosystem. Because of the open nature of the Nevermined credits system, multiple agent counterparties, as well as AI services from different third parties, can leverage the same set of redemption credits, creating a network effect for each AI agent and its capabilities. This means that Databricks could use OpenAI’s credits, or vice versa, with each setting and managing their own credits redemption criteria for their own specific sets of tools and services that they add to an agent. The modularity of this design is vast, allowing any vendor to add their tools or services to an agent, or multiple agents to combine under one set of credits, etc. Effectively, Nevermined enables the creation of cross-provider and cross-agent systems. Another major challenge was abstracting away the complexities of Web3 while still leveraging its benefits. We’ve implemented full account abstraction, allowing non-Web3 users to use Nevermined with their social logins instead of connecting a crypto wallet. Under the hood, these users still get their own non-custodial wallet via an MPC solution. We’ve also made the system gasless, eliminating the need for users to sign transactions. Plus, we’ve integrated fiat payment options through Stripe, allowing users to buy credits with a credit card without touching crypto. All of this combines to create a secure, efficient, and user-friendly system that caters to both crypto-native and traditional users. It’s been a complex journey, but we believe we’ve created a solution that’s not just workable, but truly revolutionary in how it enables AI agents to autonomously conduct financial transactions.
As the AI economy expands, what role will Nevermined play in scaling the infrastructure to support a future where trillions of AI agents handle transactions across various industries?
We’re building the fundamental payment layer that will enable trillions of AI agents to transact seamlessly across industries. Our unique credits-based system is designed to represent claims on an AI agent’s services or computational power, which is crucial for fostering a network where AI agents can directly barter services. Without a unique representation or redeemable claim against an agent’s services it is not possible for this market to exist. A freely tradable instrument is required against which the market can assign value, and this instrument must uniquely represent a specific agent’s services. This is the role of Nevermined credits, an instrument that can hold value and be exchanged for other agents’ credits. In this sense agents come to trade claims on services directly. This approach allows for more flexible and AI-specific transactions compared to using standard currencies or stablecoins. We’re not just thinking about human-to-AI interactions; our long-term vision is centered on AI agent to AI agent usage, which is where we see the real growth happening.
As we scale, Nevermined will become the go-to platform for AI-to-AI transactions across various industries. We’re already seeing demand from non-Web3 multi-agent system builders in areas like gaming, and we expect this to grow exponentially. Our modular approach and focus on payments as the critical chokepoint for AI agents gives us a significant advantage. We’re creating network effects by capturing how AI agents hold and move value, similar to how financial apps like Cash App or Alipay have created sticky ecosystems. As we expand, we’ll build deeper financial functionality, like lending and DeFi connectivity, essentially becoming the CFO for each agent or LLM. This comprehensive approach positions us to aggregate demand for agents from consumers and potentially have pricing power versus both decentralized and centralized compute providers. In essence, we’re creating an entirely new industry and corresponding commercial system that will underpin the future AI economy.
How do you see AI-driven financial interactions reshaping traditional digital commerce? What industries do you believe will be the first to fully embrace AI-to-AI transactions?
AI-driven financial interactions are poised to revolutionize traditional digital commerce by introducing unprecedented levels of automation, efficiency, and personalization. The example of Nevermined’s collaboration with Olas AI and Combinder in the energy management sector provides a compelling glimpse into this future. This project demonstrates how AI agents can autonomously make complex decisions, execute transactions, and optimize resource allocation without human intervention. As this technology matures and expands to other sectors, we can expect to see a shift from human-centric commerce to a landscape where AI agents increasingly handle financial interactions, negotiations, and decision-making processes.
The energy sector is likely to be among the first industries to fully embrace AI-to-AI transactions. However, we can anticipate rapid adoption in other sectors as well. Financial services, for instance, could leverage AI agents for algorithmic trading, risk assessment, and portfolio management. Supply chain and logistics industries might employ AI-to-AI transactions for real-time inventory management and route optimization. The healthcare sector could utilize AI agents for resource allocation, appointment scheduling, and even automated diagnosis and treatment recommendations. E-commerce platforms might implement AI agents to handle dynamic pricing, personalized product recommendations, and automated customer service. As these technologies evolve, industries that rely heavily on data analysis, rapid decision-making, and complex resource allocation will likely be at the forefront of adopting AI-to-AI transactions, fundamentally reshaping how business is conducted in the digital age.
With partners across both Web2 and Web3 industries, how does Nevermined bridge the gap between these ecosystems, and what are the benefits of incorporating both into your platform?
Nevermined bridges the gap between Web2 and Web3 ecosystems by creating a seamless, user-friendly platform that leverages the strengths of both worlds. By abstracting away the complexities of Web3 technology, Nevermined offers an intent-based payment solution that doesn’t require users to have in-depth knowledge of cryptocurrency or blockchain. The platform provides full account abstraction, allowing non-Web3 users to access Nevermined using their social media accounts instead of connecting a crypto wallet. Additionally, Nevermined implements gasless transactions and fiat payment integration, eliminating common pain points associated with Web3 platforms and catering to users who prefer traditional payment methods.
The benefits of incorporating both Web2 and Web3 into the platform are numerous. Nevermined can tap into a larger potential user base, including traditional businesses and crypto-savvy individuals, by offering a familiar Web2-like interface while leveraging the security and efficiency of Web3 technology. This approach lowers entry barriers, encourages wider adoption of blockchain technology, and provides users with more flexibility in payment options. By positioning itself primarily as an AI solution targeting all AI agents and their users, Nevermined appeals to a broader market, including businesses that might be hesitant to adopt typical crypto products. This strategy not only enhances the user experience but also future-proofs the platform, allowing it to adapt to shifts in the digital landscape while building trust among users concerned about the safety of their data and transactions.
Additionally, we recognize that arguments in favor of decentralization apply particularly well to payments. The ongoing discourse around AI, control, and the path to AGI, highlights the potential threat associated with centralized, monopolistic and predatory power structures. A decentralized approach to AI recognizes that payments represent a singular chokepoint for any agent. Should a centralized entity emerge in control of agent payment infrastructure, that entity’s ability to deplatform any agent will be singular, with an agent ceasing to exist at will. Consequently, Nevermined understands that whilst simple scale and efficiency arguments imply much of the AI infrastructure will remain off-chain, there is a strong argument for the payments piece to be in the form of a trustless decentralized protocol. This is why we see a mix of Web2 and Web3 ecosystems as the optimal path forward.
What are some real-world use cases where AI agents conducting autonomous transactions will have the most immediate impact? Can you share specific industries or projects already utilizing Nevermined?
Nevermined collaborates with a diverse range of companies in the AI and blockchain space, focusing on partnerships that can help build and expand the ecosystem for AI-driven commerce. Some of the key collaborations include:
Olas: This partnership involves developing AI agents capable of autonomous decision-making in energy management, as demonstrated in the use case mentioned earlier.
FLock: Nevermined works with FLock to explore applications of AI in decentralized autonomous organizations (DAOs) and governance systems.
Agentcoin: This collaboration focuses on creating economic models for AI agents, potentially enabling them to earn and manage their own resources.
Superagent: Nevermined partners with Superagent to develop more sophisticated AI agents capable of complex task execution and decision-making.
Peaq: This partnership explores the integration of AI commerce with Internet of Things (IoT) devices and networks.
Naptha: Nevermined collaborates with Naptha to investigate AI applications in decentralized finance (DeFi) and algorithmic trading.
Regarding interesting use cases at this nascent stage, several stand out:
Energy Management: The Olas AI agent use case we previously mentioned is a prime example. AI agents predict energy production and consumption patterns with data sourced via Combinder’s API, and autonomously control devices like HVAC units to optimize energy usage . This demonstrates the potential for AI-driven micro-grid management and peer-to-peer energy trading.
Autonomous DeFi Agents: In collaboration with partners like Naptha, we at Nevermined are exploring the development of AI agents that can autonomously manage decentralized finance portfolios, execute trades, and optimize yield farming strategies.
These use cases represent just a fraction of the potential applications for Nevermined’s technology. As the field of AI commerce evolves, we can expect to see even more innovative and transformative applications emerge, spanning industries from healthcare and education to entertainment and beyond.
For AI builders and developers looking to integrate Nevermined’s payment system, what is the business model for monetizing their AI services, and how does the credits-based system facilitate this?
Nevermined offers AI builders and developers a flexible and efficient business model for monetizing their AI services through its innovative credits-based system. This system allows builders to set dynamic access parameters for their AI agents, giving them granular control over pricing, access duration, and usage limits. By utilizing agent-specific NFTs as credits, Nevermined enables builders to create a direct representation of their AI services’ value on-chain, effectively commoditizing their computational power.
The credits-based system facilitates monetization by providing a seamless mechanism for authentication, usage accounting, and settlement. Builders can bundle multiple agents and services under a single set of credit redemption criteria, offering versatility in how they package and price their AI offerings. This approach not only simplifies the payment process for users but also allows builders to implement various pricing strategies, such as tiered access or time-based billing, without the limitations associated with traditional stablecoin payments. Credits make it possible to aggregate an agent’s services and tools under a single redemption instrument. In the absence of credits, builders would be forced to price all component services in stablecoin or other currency. In addition to commoditizing the agent services on-chain, credits provides a useful separation between the internal system of metering and usage unit accounting and the initial acquisition of credits. By leveraging Nevermined’s protocol to manage these aspects, AI builders can focus on developing and improving their services while having a robust, blockchain-based system handle the intricacies of monetization and access control.
Where do you see Nevermined and AI-to-AI transactions in five years? What does the roadmap look like for integrating this protocol on a global scale?
In the near term we should start to see AI become more ubiquitous, and where AI goes so will associated AI payments and demand for the Nevermined protocol. On an industrial level, AI agents will begin to displace knowledge workers. We can already see this starting with the release of AI agents like Devin from Cognition, Ava from Artisan, etc. Agents like Devin are of particular interest as it is not a co-pilot developer AI, but rather, an AI agent developer that can be given coding tasks that it will then perform autonomously without human-in-the-loop interaction.
Where will demand come from? It’s already started in limited circles, like Microsoft using Devins to perform code migrations. Next, Devins will likely infiltrate the mid-market staff augmentation space, in particular in testing and QA functions that are non-mission critical. These Devins will first be used for cost displacement – i.e. reduce your QA resource overhead from $150k/year to $75k/year. The interesting thing is how these AI agents will be used and who they will be working with. Devin hasn’t been built to do marketing or customer relationship management. It’s been built to develop code. Rollout within engineering departments and optimization thereof will lead to decomposition of tasks and agent specialization.
Once this decomposition begins, agents will be reduced to their discrete expert functions. This will lead to AI agents identifying and calling on other expert AI agents to perform specific minute tasks, in exchange for micropayments. The result will be the gig-i-fication / Uberization of:
- Code development
- Solution architecture design
- Legal and paralegal labor
- IT Services
- Accounting and administrative functions
- Etc.
The resulting AI-to-AI interactions, paid for via micropayments for very specific outcomes from discrete specialized AI agents will change how remedial, repetitive knowledge work like QA and Testing is procured and delivered. And this is just regarding the impact AI agents will have on human-centric industries. The real growth will happen in net-new AI agent to AI agent commercial opportunities that remain undiscovered.
Five years from now, AI agents will be commonplace, though perhaps not yet ubiquitous. A significant portion of the carbon-based economy will be somewhat dependent on agentic AI utility, and the productive value that’s created. Micro-transactions between AI agents will be standard, and the volume of transactions between agents will likely be measured in hundreds of billions of dollars annually.
Within ten years I believe we will see the emergence of a parallel AI agent economy that will begin to rival that of humans with respect to GDP and growth. It will be during this period that we will begin to see the emergence of AI Bartering where agents directly trade claims against each other’s services without the need for a common medium of exchange such as a stablecoin, other cryptocurrency or fiat.
At Nevermined we believe that AI Commerce at scale will see a small number of key protocols responsible for the major portion of AI-to-AI transaction volume. Widespread agent-on-agent interaction requires a degree of standardization. Standardization generates interoperability through common interfaces and methodologies relating to monetization. Builders will prefer to adopt the same payment integration partners, and protocols like Nevermined will achieve sticky network effects – there comes a point when it makes sense to have all agents using the same protocol. Nevermined’s success is predicated on ensuring it’s sufficiently entrenched as a category leader within the market over the next few years, and then continue to ride the gradual proliferation of AI agents throughout industry over the next two decades.
Thank you for the great interview and detailed responses, readers who wish to learn more should visit Nevermined.
