The significant expense of artificial intelligence infrastructure has been a major hurdle for widespread enterprise adoption, including within the burgeoning crypto sector. Amazon CEO Andy Jassy acknowledged this challenge, highlighting AWS’s strategy to drive down costs using custom chips and optimized inference processes, potentially paving the way for more accessible Crypto AI.
Despite speculation that more efficient AI models might lessen infrastructure needs, Jassy confirmed robust demand continues. He stated that AWS isn’t planning to slow its buildout of data centers, emphasizing the persistent and deep-rooted challenges in scaling AI infrastructure effectively.
Jassy stressed that the fundamental problems remain, especially for those developing cutting-edge models. “The lower that we can make the cost of AI, the more customers are going to use it,” he remarked, suggesting lower costs stimulate, rather than reduce, demand.
This cost-reduction focus could be particularly impactful for blockchain developers. While interest in AI applications on-chain exists, development has often been stalled by prohibitive infrastructure and operational expenses.
Drawing parallels to the early days of cloud computing, Jassy explained that reducing the unit cost of compute historically leads to increased overall customer spending. Companies reinvest savings into further innovation, finding new applications for the technology rather than simply cutting budgets.
“It allows them to save money in what they’re building, but they don’t spend less,” Jassy noted. “It unleashes them to do more innovation.”
Amazon’s approach targets two key cost drivers: the silicon itself and the expense of inference (using trained models for predictions). While training costs dominate initially, inference becomes the major operational expense as AI applications scale.
To address this, AWS is developing custom AI chips claimed to offer 30-40% better price performance compared to existing GPU options. Jassy also pointed to ongoing hardware and software advancements aimed specifically at slashing inference costs.
He underlined AWS’s commitment: “If you sat in the meetings with the AWS team right now, they feel like it is their responsibility and their mission to make the cost of AI meaningfully less than today.”
This drive towards affordability is a potential catalyst for the crypto space. As AI technology becomes cheaper in the coming years, sophisticated blockchain-native AI applications—ranging from advanced on-chain analytics to complex decentralized autonomous agents—could finally become economically viable and deployable at scale.