Nvidia CEO Jensen Huang said this week that tokens tied to artificial intelligence are now generating real profits for AI companies — and he made a point of separating them from cryptocurrency tokens. The distinction, laid out in a recent statement, adds weight to a growing narrative that tokenized AI services have genuine utility and revenue streams, unlike many crypto tokens that rely on speculation.
The profit argument
Huang argued that AI tokens are different because they're backed by actual economic activity. Companies using these tokens can charge for compute, data access, or model inference — and those charges translate into revenue that makes the tokens profitable. Crypto tokens, by contrast, often lack that underlying business model. “Tokens are now profitable for AI companies,” Huang said, directly contrasting them with the broader crypto landscape.
AI tokens vs crypto tokens
The timing matters. AI companies have been experimenting with tokenized ecosystems for a while, but Huang’s comments signal that the model has reached a turning point. Where crypto tokens frequently rely on network effects and future promises, AI tokens can point to current paying customers. Huang didn't name any specific tokens or projects, but his framing gives a clear benchmark: if a token isn't tied to a profitable service, it's closer to a crypto token.
That distinction isn't just academic. Investors and developers are watching how tokenization evolves in AI — and whether it avoids the boom-and-bust cycles that have plagued crypto. Huang’s remarks suggest that for Nvidia, the line is already drawn.
