AI in the crypto industry: an indispensable assistant, not a replacement for an analyst
The integration of artificial intelligence into the crypto industry continues to gain momentum, but as practice shows, AI remains a powerful tool for accelerating processes rather than taking on the role of an independent strategist. Even complex tasks such as writing code or monitoring exchange data still require human oversight.
My analysis of the current situation, based on conversations with leading market experts such as TradeSanta CEO Polina Demchuk and financial manager Nikita Kutsenko, confirms that AI is used in two main areas: content marketing and development acceleration. In the first case, neural networks handle news gathering, trend analysis (e.g., on TikTok), and video generation. In the second, they monitor changes in exchange APIs and write code. The key point: all these processes are carried out under strict specialist supervision.
Toolkit and Trust Boundaries
The choice of specific AI tools is not a matter of following trends but the result of empirical testing. In development, VS Code with Codex and Claude Code are actively used; for video generation, Kling and Eleven Labs; for landing pages, Lovable. Experts emphasize that decisions are made based on the "quality/cost per unit of finished material" ratio. For example, the combination of Kling and Eleven Labs has been deemed optimal in this regard after testing numerous alternatives.
Notably, serious mistakes that could have cost companies dearly have been avoided. The reason is simple: AI works in tandem with humans. It is an automation tool, not a replacement. Specialist oversight remains absolute. Polina Demchuk rightly notes that trusting an AI agent with real trades is a matter of risk management. The willingness to lose a certain amount determines both the sum that can be entrusted to the algorithm and the level of data access.
Expert Division of Labor
Nikita Kutsenko demonstrates an ideal example of human-machine synergy. His morning ritual involves loading sentiment, news, and Bitcoin data into Claude. The neural network provides a direction (long or short) with reasoning. However, the analyst does not blindly follow the signal: he cross-references it with his own market picture. A match is a strong signal; a discrepancy is a reason for in-depth analysis.
Three key tools are used in the work, each with its own specialization:
- Grok (built into X) — for real-time monitoring of crypto Twitter. The logic is flawless: early signals and leaks appear there, not in traditional media.
- ChatGPT (in conjunction with CoinGlass) — for routine chart analysis. A screenshot of data on open interest, liquidations, and the long/short ratio is sent to ChatGPT, and in 30 seconds instead of an hour of manual work, the analyst gets a ready-made picture.
- Claude — for the strategic direction of the day.
The result is impressive: analysis time has been reduced from a couple of hours to 15 minutes a day. Grok provides speed and access to sentiment, ChatGPT handles the routine, and Claude handles strategy.
Analyst's comment: The current model of using AI in crypto is an ideal example of "Intelligence Amplification" rather than artificial intelligence in its pure form. Full automation of decision-making in such a volatile and manipulation-prone environment as the crypto market remains a risky utopia for now. The key to success is a clear division of labor, where humans remain the chief arbiters and AI is their fastest and most diligent assistant.