Chipmaker Nvidia has been carrying the stock market for more than a year, riding an AI explosion that doesn’t show any signs of letting up.

In the crypto world, decentralized physical infrastructure projects, or DePin projects, have seen a similar boost. Decentralized compute project Akash, for example, has rallied 560% over the last year compared to Nvidia’s 190%.

Coinage recently caught up with Akash co-founder Greg Osuri to discuss how the crypto and AI landscapes continue to evolve, and how one narrative has played into the other.

"The rate at which [AI] models are evolving is so fast that the supply chain is not catching up," Osuri told Coinage. He highlighted the ongoing shortage of Nvidia’s H100 chips, which still have a waiting period of three to six months. Projects like Akash, and fellow compute-focused project Render, have capitalized on that narrative by offering ways to unlock compute power that may be sitting idle. As a platform, Akash allows users with those compute resources a way to get paid.

"Nvidia is in the new car business; we are in the used car business," Osuri remarked, drawing a parallel between the different approaches of the two companies. Akash’s strategy not only makes AI more accessible, but also gets even more interesting with the rise of AI agents or risks of censorship. In fact, Akash was able to attract business from noted crypto builder Erik Voorhees for his new open-source AI project Venice.

“[Eric] embracing Akash is a big deal because he sees those values realized in Akash, and he really admires the permissionlessness, the non-custodialness, and open pricing and the privacy guarantees and the unstoppable guarantees Akash gives for something like Venice,” he said. “Like you can do on Venice, things that you cannot do on Midjourney.”

And yet, as big as the AI boom has been – some analysts are openly wondering if it has gone too far. A new research report from Barclays notes that while some AI models have found massive adoption, not that many have and the current capital expenditure from large tech companies is currently running hot as they compete to stay ahead of the competition.

Barclays’ report notes that “the street is modeling $167B in cumulative AI capex, which is enough to support over 12,000 ChatGPTs. We think one of the big players may blink and cut back the capex plans, but not likely until we get well into 2025 or beyond."

But Osuri provided a compelling counterpoint to the notion that the AI boom is merely a bubble driven by FOMO (fear of missing out). According to him, the rapid evolution of AI models and the skyrocketing demand for compute power indicate a much deeper and sustained need for investment.

“A lot of times, you know, these chips are used initially for training, right? Training tends to be the most computationally intensive activities,” he said. As a litany of smaller or less-resourced projects look to deploy using cheaper compute options, Osuri says Akash will be ready. He even cited a Semafor article highlighting a Columbia student who had been priced out of the market who had turned to Akash.

“Akash is like the de-facto platform for a lot of these students,” he said.

Nonetheless, with crypto’s tough June, the correlation between Nvidia’s outperformance and Akash’s ascendance actually started to reverse. It led him to wonder if Nvidia’s meme-like performance had attracted investment at the expense of crypto assets.

Despite these challenges, Osuri remains optimistic about Akash’s future. He emphasized that Akash focuses on real-world utility and adoption rather than speculative price movements. This approach has garnered attention from regulators and legislators who see the value in decentralized cloud solutions.

Akash’s recent listings on major exchanges like Coinbase (even for New York users) and Upbit in Korea underscore the project’s regulatory compliance and growing legitimacy. These milestones are critical as they pave the way for broader adoption and integration of decentralized technologies.