Key Highlights
- Alex Svanevik forecasts AI agents will ultimately replace traditional trading dashboards.
- Svanevik suggests a “Trust Ladder” model for gradual AI trading autonomy instead of jumping to autonomous finance directly.
- According to him, AI and crypto are inevitable in the real world.
According to Alex Svanevik, CEO of blockchain analytics firm Nansen, the future of finance is not about clicking buttons, but will be based on having a conversation. Svanevik outlined a bold vision where trading becomes “agentic.” By 2030, he forecasts that the primary interface for investors will not be a dashboard filled with charts, but an AI agent capable of understanding and executing trades on-chain.
In an exclusive interview with The Crypto Times, Svanevik stated that while AI models like GPT-5 or Opus 4.5 are impressive, they lack the primary context that would be useful to traders on their own since they lack real-time blockchain data. “If you ask it ‘What token is smart money buying?’, ChatGPT will not know,” Svanevik explained.
Giving AI eyes on-chain
Nansen’s solution is to give these models “eyes on-chain.” By integrating their proprietary analytics infrastructure with top-tier LLMs, Svanevik claims Nansen’s agents have achieved an 85% quality score in “Expert” mode, compared to a mere 20% for other models.
This integration lets users move easily from discovery to execution, where the agent interfaces with aggregators like Jupiter and OKX to finalize the trade.
However, Svanevik highlighted that to get to fully autonomous finance, the model should be that of the “Trust Ladder.” Comparing the current state of AI trading to self-driving cars, he shared, “That would be like sitting into a Tesla for the first time ever and just going in the back seat and letting it drive, which I think most people, if you’ve never seen it, would be uncomfortable.”
He added, “So similarly in trading, we have to take the approach of climbing a trust ladder. And it starts with having the human in the loop on tapping approve on the trades.” As trust gradually builds, the system can move to “smart trades.” This is where the AI actively identifies and suggests opportunities for the user to review.
Only after these stages have been covered and properly assessed, Svanevik suggested a move toward full autopilot. He warned that jumping straight to autonomy today would be a “really bad idea” due to regulatory hurdles and systemic AI risks.
The era of vibe coding
During the conversation, Svanevik also spoke about “vibe coding,” where investors bet on the creators of AI agents rather than just tokens. While acknowledging that this sector is currently rife with speculation and “bad incentives,” he shared that he sees it as part of a broader narrative shift.
Svanevik reflected on the broader trajectory of the industry, suggesting that the integration of AI and crypto in the real world is now an inevitability.
Ultimately, Nansen’s pivot from a neutral data provider to an active execution layer signals a massive change in the market. As Svanevik put it, the complex tools of today will eventually disappear, replaced by agents that allow anyone to trade with the sophistication of a hedge fund, simply by asking.
Also Read: Coinbase Launches Agentic Wallets for Autonomous AI Transactions
