The blockchain world is witnessing a significant transformation as companies traditionally focused on crypto mining are strategically pivoting toward artificial intelligence infrastructure. This shift reflects not just an opportunistic business decision but a natural technological evolution that leverages existing resources to meet growing computational demands.

The panel at CODA 2025 opened with a discussion of CoreWeave, a company that exemplifies this pivot. Originally focused on crypto mining, CoreWeave transitioned to providing AI infrastructure and subsequently experienced explosive growth following their IPO. As Alex Pawlowski noted, "CoreWeave is in a very unique position. They are actually financed by Nvidia. So they could transition perfectly from an Ethereum mining company to an AI infrastructure company." This transition wasn't merely fortuitous—it represented the confluence of existing capabilities with emerging market opportunities.

Hafsah Dege, Head of Product at Kadena, explained how this transition makes technological sense, particularly for proof-of-work blockchains: "We're Kadena, we're a layer one but also one of the only remaining proof of work blockchains... Historically there were more proof of work blockchains and a lot of people built their own data centers. They were able to source electricity and this was their business model." These mining operations had already established critical infrastructure—data centers, access to cheap electricity, and specialized hardware—that could be repurposed to support AI's computational needs.

Rather than abandoning blockchain operations entirely, many mining companies are adopting a hybrid approach. "What we're finding is there's now a pivot. At least with Kadena we still have miners. But those same miners, when they've got availability, they're also supporting the AI industry," Hafsah explained. This creates additional revenue streams while maintaining their blockchain operations, with estimates suggesting "a split of around 30%" of Bitcoin miners now also supporting the AI industry.

David Palmer emphasized that this infrastructure shift represents an important economic evolution: "We're moving towards the AI economy. That's simply it. We're building a new digital infrastructure for a new digital economy. And that requires evolution." He observed that while repurposing mining infrastructure for AI makes practical economic sense, the true potential lies in blockchain protocols finding their unique place in the AI value chain: "What I'd like to see is protocols finding a place in the value chain of the AI economy... a lot of the value from blockchain is in consensus, it's in trust, it's in portability of value, it's also in smart transactions."

This pivot is not universally accessible, however. As Alex Pawlowski cautioned, "It takes more than just having an interest in that sector to transition." Companies like CoreWeave succeeded partly due to timing, existing relationships (like Nvidia's financing), and having the right infrastructure already in place. The transition requires not only hardware but also facilities, cooling systems, and specialized personnel.

The convergence of blockchain and AI infrastructure appears inevitable as both technologies mature. Blockchain companies with significant computational resources are well-positioned to capitalize on the growing demand for AI processing power, creating a symbiotic relationship between two of the most transformative technologies of our era. As blockchain provides the trust layer and AI delivers the intelligence layer, their infrastructural marriage may become the foundation for the next generation of digital innovation.

Solving AI's Ownership and Privacy Crisis with Blockchain

As artificial intelligence rapidly transforms our digital landscape, critical issues of data ownership, privacy, and attribution have emerged as significant challenges. The CODA 2025 panel explored how blockchain technology might offer solutions to these increasingly urgent concerns.

The ownership crisis in generative AI has become particularly problematic as content creation becomes more automated. When an AI creates an image or text, questions immediately arise: Who owns this creation? How can creators be fairly compensated for their contributions to training data? Hafsah from Kadena addressed this directly: "If we go back to why blockchain was started, it was so we could have more ownership. So if there's a crisis in other industries or with AI where people want to have ownership, I think it makes sense to go back to the one industry that's managed to solve this."

Blockchain's inherent characteristics make it uniquely suited to address these ownership challenges. "Everything's trackable, it's verifiable, and again, you don't need a middle party," Hafsah explained. This transparency creates opportunities for establishing clear provenance and attribution chains for AI-generated content. As she noted, "You can get a timestamp if you put everything on the blockchain and use the infrastructure, and then you can use that to tie it back to ownership."

This capability extends beyond simple attribution to enabling new economic models around AI content. "You can build an economy around it. Maybe it's like an AI agent, and you can make sure anyone that uses what you've created, you're fairly rewarded for it," Hafsah elaborated. This approach could fundamentally change how we compensate creators whose work trains AI models or inspires AI-generated content.

Privacy concerns represent another critical challenge in the AI landscape. David Palmer highlighted how traditional privacy regulations may be insufficient for AI's rapid evolution: "GDPR, when that came out, was the social media application and blockchain killer legislation. However, do you think GDPR is going to be able to protect people's data and privacy when AI really takes off, or do we need something more dynamic?" He suggested that privacy enforcement needs to become more automated and dynamic, potentially leveraging "ZKPs, encryption technologies, and smart contracts" that can be implemented on blockchain systems.

Hafsah confirmed that advanced cryptographic techniques like zero-knowledge proofs are actively being developed to address privacy challenges: "I don't actually think privacy will be an issue. There are solutions being worked on... We're also doing some work on how we can add selective privacy so we can onboard institutions." These technologies could enable users to maintain control over their data while still allowing AI systems to learn from it.

The panel also discussed the authentication challenges posed by AI-generated deepfakes. David Palmer observed that traditional verification methods are becoming obsolete: "We live in a world of paper proofs... those senses as a means of authentication are really being challenged by AI." Blockchain offers a potential solution through verifiable timestamping and immutable records. As Hafsah explained, "I can say there was a UFO in my city, Manchester, at this time. But then on the blockchain, you can see I actually created this record a year earlier. So it's a lie."

While acknowledging blockchain's potential, the panel recognized that technical solutions alone are insufficient. Alex Pawlowski emphasized the need for "more societal acceptance" and "government collaboration" to establish clear guidelines. He advocated for "breaking up this black box of AI" to create transparency about what AI systems do and why they make specific decisions.

David Palmer concluded that these issues reflect broader questions about how AI's economic benefits will be distributed: "Everybody needs to share in the economic benefits of AI, especially as it's going to displace a lot of people." He warned that without proper ownership systems, "you're going to have a few dominant parties who own everything, who profit from everything, and then you're going to have a lot of displacement in society with people losing their jobs."

The convergence of blockchain and AI thus represents not just a technical integration but a potential pathway toward a more equitable digital future where ownership is clear, privacy is protected, and economic benefits are more widely shared. As Alex suggested, blockchain could help "extract aspects" of knowledge and "make that tradable" in ways that benefit creators and users alike.

The Future Vision: Sovereign AI and the New Economic Model

The CODA 2025 panel culminated in a forward-looking discussion about how the integration of blockchain and AI might shape our collective future by 2030. What emerged was a vision of "sovereign AI" underpinned by blockchain-enabled economic models that could fundamentally transform how AI capabilities are distributed, governed, and monetized.

Alex Pawlowski introduced the concept of sovereign AI as a pivotal development in the years ahead: "The keyword here is sovereign AI. This idea that in the future we will have on a state level maybe, or on a country level, general purpose build models." This represents a shift from the current centralized paradigm where a handful of tech giants control AI development toward a more distributed model where nations and communities can develop AI systems aligned with their particular values and needs.

This sovereignty extends beyond just national models to individual ownership of AI agents. Alex described his work with Shiza, which is "building a platform where you can develop your own agent. But then beyond the building and the design, the ownership assets of the custody using blockchain." The next evolutionary step would be a "marketplace and knowledge as a service marketplace" where aspects of intelligence could be extracted and traded—essentially "sucking the soul out of a being and then instilling and infusing it into another."

David Palmer emphasized the economic implications of these developments, warning that without intervention, we're heading toward an AI oligopoly: "We could be in a world in 10 years where you have five companies, and if you think the Internet and social media companies are dominant now, wait till they start incorporating AI and data." He pointed out troubling signs in current cloud terms and conditions that already permit companies to "send your data to AI large network models."

The economic stakes are enormous. As David noted, "Some of the estimates for AI is 18.9 trillion. The AI market cap will be by 2033... roughly 18% of global GDP." With such massive economic potential, the panel agreed that how this value is distributed will become a defining issue. "For this to work economically across society, you have to have the AI of the many, not the AI of the few," David insisted.

Blockchain technology could play a crucial role in brokering more equitable data relationships. "Where blockchain can come in is to broker the reimbursement of the data provider," David explained. "Everybody's a data provider—businesses, people, devices. Part of what we can do in this industry is to have smart contracts and initiatives where that brokerage can take place." This represents a shift from extraction to compensation, where those who generate valuable data receive fair payment for their contributions.

Hafsah highlighted how decentralized blockchain infrastructure could support this vision: "If we look at what's worked well with the blockchain, it's that everything's verifiable and it's decentralization." She emphasized that for AI agents to truly thrive, they need financial capability: "With blockchain it unlocks more use cases for these AI agents... because now you can actually tie the new technology into the financial system as well."

The panel envisioned a fundamental shift in how we interact with technology by 2030. David suggested that "we're moving towards an agent architecture, AI agent architecture. So the agent will be our interface." These agents would operate autonomously on our behalf, with blockchain providing the authentication, trust, and economic infrastructure they require to function in a decentralized ecosystem.

Success by 2030, according to the panel, would involve several key developments. David hoped that "by 2030, deep DeFi and blockchain not just being about the price of crypto but really helping to bank the unbanked and helping to bring AI to the edge—AI to the masses." He added that ideally, "we're not talking about AI or blockchain, but we're talking about the applications they have and the agents that we're using and how this is helping business and helping our personal experiences."

Perhaps most importantly, this integration should address inequities rather than exacerbate them. As David put it, success would mean talking about "how this is helping to extend the barriers of equality and opportunity." This represents a profound challenge—as AI drives unprecedented economic transformation, blockchain may offer the mechanisms to ensure that the benefits are widely shared rather than concentrated among a privileged few.

The convergence of AI and blockchain thus presents not just a technological revolution but a chance to reimagine economic and social structures. As Alex observed, we face "one of the most significant five years to 2030 that we've seen in our lifetimes," with the potential to create either "an Internet and an AI economy of the few or one of the many." The sovereign AI vision coupled with blockchain-enabled economic models offers a pathway toward the latter, more equitable future.