Lede
Artificial intelligence is becoming increasingly embedded across the cryptocurrency trading landscape, taking over complex processes such as market analysis, trade execution, and portfolio optimization. According to Ryan Li, the co-founder and CEO of crypto research platform Surf AI, the technology is currently replacing the estimated 80% of tasks that human workers typically find undesirable. This shift toward automation has gained significant momentum recently, with interest in using AI to boost efficiency within the crypto sector accelerating during the last quarter of 2024. As these tools become more sophisticated, they are being utilized by top-tier researchers to dramatically improve the quality and speed of their work output.
While machines handle much of the data-heavy bandwidth required for research and monitoring, the fundamental structure of trading still relies heavily on human oversight. Human participants remain responsible for defining overall strategies, setting specific risk limits, and maintaining accountability for the eventual outcomes of trades. This balance between automation and human judgment is reshaping professional workflows and beginning to redefine which human roles remain essential in the market. The emergence of AI agents has further pushed the boundaries of what can be automated, though most AI tools in crypto remain within tightly constrained parameters set by their human operators.
Context
To understand the current state of the market, it is necessary to distinguish modern AI systems from traditional algorithmic trading models. Algorithmic systems already handle the vast majority of trade execution in major global markets, but these operate primarily on deterministic, predefined rules. In contrast, AI systems are designed to navigate uncertainty and process data that may be noisy or contradictory. Recent performance metrics highlight the potential advantages of this transition to more intelligent systems. An experiment conducted by the decentralized perpetuals exchange Aster pitted 100 human traders against 100 AI models during a period of market decline to test capital preservation.
The results of the Aster experiment showed a stark difference in outcomes; while human traders were down 32.21% during the battle, the AI models posted a significantly smaller loss of 4.48%. This capability to manage risk during bear conditions is also mirrored in traditional finance research. Ed deHaan, an accounting professor at Stanford, contributed to a study involving thousands of US mutual fund portfolios. The study found that AI-managed portfolios generated an average of $17.1 million more per fund per quarter than their human-managed counterparts. These findings suggest that while humans are still essential for high-level decision-making, AI is proving more effective at executing tasks under complex conditions where information is incomplete.
Impact
The rapid integration of AI technologies has sparked significant concern regarding job displacement within the cryptocurrency industry. Data from the crypto research platform Santiment indicates that in June, the topic of AI job replacement topped crypto social discussions. This specific narrative even surpassed popular market topics such as memecoins and general trading strategies. Nina Rong, the executive director of growth at BNB Chain, notes that AI is currently improving research efficiency by gathering public domain information and allowing non-programmers to use coding as a functional tool. However, this shift is fundamentally changing how trading firms approach hiring and team structure.
Traditional roles that once relied on teams of junior analysts or interns are being consolidated as AI absorbs routine research work. Ryan Li observed that funds which previously hired large research teams may now only require one highly skilled researcher who is proficient at working with AI tools. This evolution is also visible in the performance of the assets themselves. Despite the high level of interest and the technological strides being made in execution efficiency, AI-related tokens have experienced significant volatility. Since their peak in late 2024, AI tokens have lost approximately 67% of their market value. This suggests a disconnect between the functional utility of the technology in trading operations and the speculative market value of associated crypto assets.
Outlook
Looking forward, the industry is moving toward a model where autonomous trading is technically feasible but governed by strict human-defined parameters. Igor Stadnyk, the co-founder of AI trading platform True Trading, emphasizes that while execution can be automated, the core elements of control, limits, and accountability must remain human-centric. Strategy selection and risk management continue to be human decisions, as these factors directly impact financial outcomes and professional responsibility. The pace of development in the crypto sector remains remarkably high compared to other industries.
Stadnyk noted that a single year of progress for AI agents in the crypto space is comparable to decades of development in more traditional fields like aerospace or medicine because everything can be tested and iterated very quickly. While major market players may already be utilizing autonomous models for tasks like managing wallets or rebalancing portfolios, wide-scale aggressive promotion of these systems remains limited. The future role of the human trader is likely to shift away from manual mechanics and toward high-level strategic oversight. As AI continues to absorb routine tasks, the value of domain experts who can leverage these tools effectively will likely increase, even as the barrier to entry for junior roles becomes more complex due to the automation of foundational research tasks.
About the Author
