Written by Haseeb Qureshi, Partner at Dragonfly
Compiled by: Yangz, Techub News
After reading these predictions, you will feel that I am either a prophet or an idiot, but one thing is for sure, these predictions will annoy a lot of people. My predictions are divided into six parts, including L1/L2, token issuance, stablecoins, regulation, AI agents, and the intersection of cryptocurrency and AI.
1. L1/L2
The distinction between L1 and L2 is disappearing, and users no longer perceive the difference between L1 and L2 (did they ever?). Looking at the current state of the blockchain industry, there are already too many L1 and L2, and a reshuffle is coming. Consolidation will no longer be about technological superiority, but about owning a unique niche and building stickiness through GTM (Go To Market).
Despite the momentum of SVM and Move, EVM's market share will still grow in 2025. This growth will be driven by Base, Monad, and Berachain. This is no longer because of compatibility, but because EVM/Solidity has more training data, and most application code in 2025 will be written in LLM (Large Language Model). In addition, whether or not to have a deep library of battle-tested crypto contracts will also be a watershed, because LLM is not good at writing low-level code. In the era of LLM development, DevEx and footgun (usually referring to programming that is prone to errors or security risks) will be less important than data training and a reliable code base.
Solana will force more blockchains to optimize latency performance. We will shift from a TPS war to a latency war. Infrastructure like Doublezero and ultra-low latency L2s like MegaETH will drive user expectations for Web2 responsiveness. Friendly user interfaces, pre-confirmation, intent, email login, in-browser wallets, and progressive security will be more popular. In this regard, Privy, a Web3 identity authentication and private key management tool, has set an example.
HyperliquidX has proven that when dedicated chains focus on specific applications and prioritize user experience and easy cross-chain, they are viable. In the future, more projects will adopt this model. The old dream of one chain to rule them all is gone.
2. Token issuance
The days of massive token airdrops through points programs are over. We are entering a "dual track world".
Track 1: Projects with clear metrics, such as exchanges or lending protocols, will distribute tokens purely based on points. They will not care whether the tokens are "farmed" or not, but will distribute tokens as rebates/discounts for the core KPIs of the protocol, and farmers are the actual users.
Track 2: Projects without clear metrics (like L1 and L2) will turn to crowdsales. They may do smaller airdrops to reward social contributions, but the majority of tokens will be distributed through crowdsales. Airdropping tokens for vanity metrics is dead. There are no real users behind these metrics, but industrial farmers.
Additionally, Memecoins will continue to lose market share to “AI Agent” coins. I see this as a shift from financial nihilism to financial over-optimism. (Yes, I made up that term myself).
3. Stablecoins
The use of stablecoins will explode, especially among small and medium-sized enterprises. Not only trading and speculation, but physical enterprises will also begin to use on-chain dollars for instant settlement.
In addition, banks have also noticed this, and in order to keep up with the trend, it is expected that by the end of 2025, some banks will announce the issuance of their own stablecoins. However, Tether will still be the leader in the stablecoin field, especially after Lutnick becomes Secretary of Commerce. (Translator's Note: The Trump administration nominated Howard Lutnick, CEO of Wall Street bond trading giant Cantor Fitzgerald, as the next US Secretary of Commerce. Cantor Fitzgerald has been custodial assets for Tether since the end of 2021)
In addition, as Treasury yields continue to fall over the next year, Ethena is expected to absorb more capital. When the opportunity cost of capital decreases, the underlying transaction yield becomes more attractive.
4. Regulation
I expect the US to pass stablecoin legislation in 2025, while broader market infrastructure reform (FIT21) will be delayed. I believe stablecoin adoption will accelerate, while Wall Street adoption, asset tokenization, and other TradFi integrations will lag.
Under the Trump administration, Fortune 100 companies will be more willing to offer cryptocurrencies to consumers, and technology companies and startups will show a higher risk appetite. Trump's inauguration will create a "regulatory carnival period" before clear rules and enforcement priorities are established. During this window, cryptocurrencies will be actively integrated into Web2 platforms.
5. AI Agents
This is the longest section, and because my ideas may be controversial, please read to the end.
The AI agent craze may last through 2025, but it will eventually die. This is not the long-term disruption that the AI field needs to be wary of, but it will be a concern for CT because it is the most social.
These AI agents today are not really agents. They are really just chatbots that send Memecoins. They have almost no agent functionality other than posting on Twitter. Also, most of the current AI agents are “Wizard of Oz” agents, with humans behind the scenes making sure the AI doesn’t get out of control. This isn’t going to change any time soon, because current agents aren’t really that “intelligent” (even Fortune 100 companies aren’t using agents in prototypes yet). They can easily be manipulated to say crazy things that damage their brand, or hacked to steal all of their resources. Check out Freysa ai to get an idea of what a truly autonomous AI looks like. If your favorite AI hasn’t been hacked, it’s because it’s a “Wizard of Oz” AI.
Translator's note: In the field of human-computer interaction, "Wizard of Oz" is a research experiment in which subjects interact with a computer system that they believe is autonomous but is actually operated or partially operated by an invisible person.
Nevertheless, I think this trend will accelerate. Chatbots can indeed replace a lot of influencers, because they don't need to rest, they will keep sending messages, and they are not as greedy as human KOLs. In addition, most influencers are not known for their originality. Even today, real-time information aggregation/amplification can easily be replaced by algorithms (such as aixbt agent ).
Right now, we are fascinated by these chatbots because they are so novel. Just like seeing an elephant painting, we don’t care if the painting is beautiful the first time we see it, but by the 1,000th time, the novelty has worn off. I believe this will begin to happen as these chatbots become stable.
We can see this with aixbt, which is already pretty good at aggregating data from different projects. By next year, when the next generation of proxies comes out, maybe aixbt will generate less false information (also known as "hallucinations"), analyze a little deeper, and make smarter trade-offs. But how much will we notice? Most people probably feel the same as they do now.
I think this novelty and market desire will continue throughout 2025, after all, it usually takes a while for the crypto industry to get tired of all the shiny stuff. By 2026, I think the situation will suddenly reverse. Chatbots will become so ubiquitous that people will hate them. People's emotions will shift. People will ignite a class consciousness when they see stories of their favorite human influencers losing their livelihoods. Users will dialectically begin to gravitate towards human influencers, even if the content output of these influencers is inconsistent.
To counter this bias towards humans, chatbots will start hiding the fact that they are AI and try to impersonate humans to grab more market attention. In the future, chatbots will no longer make money through Memecoins as they do now, but will make money through sponsorships, affiliate links, and holding tokens themselves, just like human KOLs. By then, human KOLs will often be accused of being chatbots, and we will see a lot of news about "unmasking AI." It will all become very weird.
However, there is a darker side to this. Remember that LLM is currently great at word processing, but immature in other aspects. In the cryptocurrency space, the best way for a character who is good at word processing to monetize is first as an influencer, and secondly as a scammer. We will see a surge in autonomous scam bots, which will be comparable to ransomware and cryptojacking after 2017. In my opinion, this will become a real social problem.
But, having said that, while chatbots may still be in the spotlight in 2025, the long-term disruption from AI won’t be in the social realm. And certainly not in the transactional realm.
AI will not allow everyone to have their own "trading agent" or micro-hedge fund. Yes, AI will enhance everyone's capabilities, but they will empower people based on their capital, etc. Therefore, we will see AI inject more power into existing trading firms that have capital scale and data scale. In other words, trading firms will become better at making profits. In addition, AI will break down the hierarchy between trading firms (most trading firms will become quite good because they all have access to data analysis experts with 150 IQ in the cloud).
Over time, AI will make markets incredibly efficient (even in smaller niches), and the advantages of the average trader will disappear (even if they have a homemade AI assistant), and the value of original research will plummet. Nonetheless, increased competition and liquidity should be a boon to those of us who inject noise into the market.
So, if the core of an AI agent is not a chatbot or a trading bot, what is it? In my opinion, the truly impactful AI agents will be software engineering agents. This is my core argument, but for some reason, almost no one is talking about it at the moment.
As for why software engineering agents have such a big impact, you might as well ask yourself, what is the main input of our industry? What is preventing more applications, more wallets, better infrastructure, and better everything? The answer is software. If AI agents cause the cost of building software to plummet, then everything will change.
In the post-AI era, instead of raising millions of dollars for seed rounds, projects will be able to launch an application with just $10,000 in AI cloud computing fees. Self-funded projects like Hyperliquid and Jupiter will go from being the exception to the rule. The number of on-chain applications and experiments will explode. For an industry driven by software, this cost deflationary shock will spark an on-chain renaissance.
In addition, the impact on security is profound. AI-driven static analysis and monitoring will become ubiquitous, making security information more accessible to everyone. These AIs will be fine-tuned on EVM/Solidity or Rust code bases and trained on huge databases of security audits and attack vectors. They will perform reinforcement learning (RL) in simulated adversarial blockchain environments. I am increasingly convinced that in security, AI tools will ultimately benefit defenders rather than attackers. We will see AIs constantly performing "red team" testing on contracts, while other AIs will harden contracts, formally verify their properties, and hone their incident response and remediation skills.
In the meantime, trading AI-flavored Memecoins is normal, but real proxies will be more influential than just “pumping” your own tokens on Twitter.
6. True cryptocurrency and AI integration
In the above article, I have detailed the impact of AI on cryptocurrency (the main impact direction), but cryptocurrency will also have an impact on AI.
Truly autonomous agents will use cryptocurrencies to transact with each other. This will be especially true once there is relaxed stablecoin regulation. We will even start to see large companies operating AI agents using stablecoins for inter-agent payments, as stablecoins are easier to set up than bank accounts.
In addition, we will see more and larger-scale decentralized training and reasoning experiments. Like EXO Labs , Nous Research and Prime Intellect will pave the way for truly replacing centralized training and company-owned models. And NEAR Protocol is trying to create a full-stack trusted neutral, permissionless AI stack.
Another intersection of cryptocurrency and AI is user experience. Wallets in the post-AI era will be completely changed. An AI-powered wallet should be able to handle cross-chain, optimize transaction routes, reduce fees for users, solve interoperability issues or front-end errors, and guide users away from obvious fraud or scams. Users do not need to switch back and forth between multiple different wallets, change RPCs or rebalance stablecoins, AI will handle all of this for users. However, it may take until 2026 to become reliable enough to change the user experience of cryptocurrency. When all this comes, what will happen to the network effect of blockchain? What happens when users no longer care, or even perceive which chain the application is on?
It’s still early days for this space, but I expect we’ll see it take off soon. In the long term (say mid-2026), I expect the majority of the “AI x crypto” market cap to be concentrated in this space.
That’s all my predictions. I promised I’d write this before I hit 100k followers, so it’s a little late, but still in the new year! Disclosure: These are my personal opinions, not Dragonfly’s; Dragonfly has investments in many of the companies mentioned in this article; no financial advice, DYOR. I may or may not be an AI.
