Haseeb Qureshi, managing partner at Dragonfly Capital, shared his predictions for the cryptocurrency industry in a post on social media platform X on Jan. 1.
Layer 1 and Layer 2 Networks
Qureshi predicted the distinction between Layer 1 (L1) and Layer 2 (L2) blockchains would continue to fade in 2025, with users paying little attention to technical differences. He believes the crowded blockchain ecosystem will undergo significant consolidation, driven less by technological advancements and more by the ability of networks to carve out unique niches and create sticky user experiences.
Ethereum Virtual Machine (EVM) networks are set to expand their market share, despite the rising popularity of alternatives like Solana’s Virtual Machine (SVM) and Move. According to Qureshi, advancements in large language models (LLMs) will favor EVM-compatible networks due to their extensive libraries of existing contracts and training data. THe expects tools like LLMs to play a crucial role in automating application development, prioritizing ecosystems with robust foundations.
Solana’s emphasis on low latency will likely pressure other blockchains to optimize user experience. Qureshi predicted an industry shift from “transactions per second” (TPS) wars to “latency wars,” with infrastructure innovations such as ultra-low-latency Layer 2 solutions and progressive security models pushing the boundaries of blockchain performance.
He also highlighted the success of Hyperliquid, a specialized chain focused on a single application. This approach, he argued, demonstrates the viability of specialized blockchains designed for specific use cases, replacing the outdated dream of a “one chain to rule them all.”
Token Launch Strategies
Qureshi forecasted a shift in how projects distribute tokens, noting that large-scale airdrops based on vanity metrics will decline. Instead, he sees the emergence of two primary tracks:
- Projects with measurable success metrics, such as exchanges or lending protocols, will reward tokens based on user activity. For these protocols, farmers are legitimate users contributing to core performance indicators.
- Projects without clear metrics, such as L1s and L2s, will lean toward crowdsales as their primary distribution method, reserving small airdrops for community contributions.
Additionally, Qureshi believes memecoins will lose popularity, giving way to “AI agent coins,” which reflect a shift from financial speculation to optimism about AI-powered applications.
Stablecoin Adoption and Institutional Entry
The venture capitalist expects stablecoin adoption to grow exponentially in 2025, particularly among small and medium-sized businesses seeking efficient settlement options. Qureshi predicted that major banks will begin issuing their own stablecoins by the end of the year as they strive to compete in the digital finance space. However, he maintained that Tether will continue to dominate the market under the current U.S. administration.
Ethena Labs, a key player in the stablecoin ecosystem, is positioned to capture significant capital, according to Qureshi. With treasury yields expected to decline, he believes that blockchain-based yields will become increasingly attractive.
Regulation and a Shifting Landscape
Qureshi anticipates the U.S. will pass stablecoin legislation in 2025, spurring further adoption of these digital currencies. He thinks broader market infrastructure changes, including the Financial Innovation and Technology for the 21st Century Act (FIT21), are likely to face delays, slowing the integration of traditional financial institutions with blockchain technologies.
Under the Trump administration, Qureshi predicts a temporary regulatory reprieve, which he described as a “jubilee” for the crypto industry and expects this perceived window of opportunity to drive aggressive expansion as Fortune 100 companies and startups integrate crypto services into their platforms.
The Evolution of AI Agents
AI agents will dominate the crypto landscape in 2025, Qureshi predicted, but he offered a nuanced view of their current and future roles. He described many of today’s AI agents as “Wizard of Oz” systems, where human intervention ensures functionality, masking the agents’ lack of true autonomy. Although he acknowledges that these bots could potentially replace some influencers and automate real-time information aggregation, he says that their reliance on human oversight limits their capabilities.
Qureshi suggested that the novelty of AI agents would sustain interest through 2025, but this fascination could fade by 2026 as the tools become ubiquitous. He believes the market will eventually shift toward favoring human influencers over bots, driven by a desire for authenticity. However, he points out that this sentiment could trigger AI agents to disguise themselves as humans, leading to unmasking scandals and ethical debates.
On a darker note, Qureshi warned of a potential explosion in autonomous scambots, comparable to the rise of ransomware in previous years. He mentioned that these malicious agents could exploit blockchain systems for fraudulent activities, posing significant security challenges.
Despite these concerns, Qureshi identified software engineering agents as the true disruptors, saying that these tools have the potential to revolutionize blockchain development by dramatically reducing the cost and complexity of creating new applications. He envisions a future where projects can launch with minimal funding, fostering an era of unprecedented experimentation and innovation.
AI-powered security solutions could also enhance blockchain safety, Qureshi argued. He predicts the widespread adoption of AI for static analysis, real-time monitoring, and adversarial testing, ultimately giving defenders an edge over attackers. He thinks that these advancements could make blockchain ecosystems more secure and accessible to developers and users alike.
The Convergence of Crypto and AI
Qureshi also explored how AI could influence the crypto industry and vice versa. He believes autonomous AI agents will use crypto for payments, particularly stablecoins, due to their simplicity and accessibility. He pointed to decentralized AI training and inference as a growing area of experimentation, with promising projects already paving the way for alternatives to centralized AI models.
Qureshi believes AI-powered wallets could revolutionize user experience by automating tasks like bridging assets, optimizing fees, and identifying scams. This evolution, he noted, could redefine blockchain network effects as users become less concerned with the underlying chains powering their applications.
Featured Image via Pixabay
