So, Who Will The Big Winners Be?

With AI unlocking innovation at the application layer, the answer feels pretty clear: apps, apps, and more apps (of course with a bit of infra thrown in. I mean, this is still crypto 🤣). As David highlights below, we’re already starting to see this shift from the infra cycle to the app cycle play out. AI will simply turbo-charge it.

There’s four categories of crypto products I’m particularly excited about, each with asymmetric upside given how early we are in their respective formation:

1. Aggregators AKA SuperApps

I predict we see the “FAANG of crypto” emerge: super-apps that aggregate capabilities from agents who abstract on-chain UX (see #2), own a direct relationship with their users, and vertically integrate down the stack that not only makes their app more robust but also competes for builder mindshare by becoming an infra provider (much like Amazon or Google) — all while exhibiting monopolistic capture in their respective sector (eg search + advertising, finance, commerce, social, etc). Much like FAANG who drives ~20% of the S&P today, I also envision this category to follow a power law dynamic and capture a similar percentage of crypto’s market cap by 2030: an opportunity size measured conservatively in the hundreds of billions and optimistically in the trillions.

I’m especially excited about DeFi (or DeFAI as the kids call it) as a killer use case here: imagine a one-stop next-gen financial super app that allows users to be able to seamlessly access every financial asset in a world (b/c they all exist on-chain!), provides investment advice or ideas, does real-time sentiment analysis, and executes on user intent. Another intriguing opportunity is the “Google of Crypto” that designs a “PageRank”-style algo on smart contract code that solves the discovery problem of crypto apps + assets with ads or innovative value flows to monetize.

The winners of this category will create far larger outcomes than anyone can imagine b/c it’ll have something that web2 super-apps never had access to: the token. The token is the one thing in crypto that has demonstrated undeniable PMF to acquire users, attract a wide range of legion of believers + investors, and capture mindshare.

2. Agents as SaaS

I’m excited about AI agents that execute on one thing extremely well with the ability to be composed via aggregators or other agents — much like SaaS or financial products today. For example, imagine a fully autonomous agent which takes in LP money and is best-in-class in investing on their behalf in crypto markets (it’s a top 1% liquid trader while getting allo into the best performing deals on Echo), all while taking smaller fees vs. an ETF or fund. Or imagine an agent that is able to generate outsized returns in prediction or sports gambling markets. Or something like @aix_bt that provides high-signal market and investment research. These types of agents will democratize access to both previously inaccessible markets that are now on-chain (dollars, RWAs, etc) as well as strategies (quantitative, venture, etc).

This goes beyond finance. I can see a world with an AI doctor that is uniquely trained to handle specific patient profiles with the ability to charge their insurance and issue low-risk prescriptions via crypto rails. Or an insurance agent that can procure the cheapest home insurance for your home. To be fair, we’re still a ways away from all this (most agents today can’t even do basic interactions on-chain). But the opportunities here are limitless as these agents innovate on customer acquisition, value capture, and pricing mechanisms via their native token (eg users must hold 100 AIXBT for premium access, etc). As this plays out, I suspect we’ll also see huge opportunities in agent marketplaces (think Ebay or OpenSea of this world).

3. AI-Native Infra

The next big infra opportunity (eg L1) will offer an order of magnitude improvement on UX to users, not speed or cost. It will achieve this by being architected around AI agents + AI-powered smart contracts as the core primitive that builders deploy and natively offer capabilities such as performant on-chain inference (see section #4 below), verifiable off-chain inference with TEEs, smart accounts that allow semi-autonomous AI agents to execute on your behalf with built-in guardrails, access to compute / training, and support for bi-directional agent-to-agent value flows that facilitate swarms + agentic economics. Similar to dApps of this era, many of the agents in category #2 above, esp the long-tail, will be deployed on these new L1s vs. managing their own infra, and enjoy similar network effects of adjacency and composability. I’m also excited for this new generation of L1s to re-imagine things like value-capture mechanisms, MEV, and consensus (could an agent be a validator?).

This is definitely not to say I am bearish Ethereum, Solana, or other leading L1s/L2 ecosystems. These ecosystems can and will introduce many of these capabilities in their stack over the coming yrs — but I believe there will be massive upside accrued by the L1s natively born in this era who will more naturally vibe with devs of this generation. Projects like ai16z and Virtuals are early precursors both of what this may look like and the opportunity size of being the winner in this space.

The L1 trade just might continue being alive and well.

4. Intelligent Assets

Some of the most popular apps in crypto today (eg stablecoins, NFTs, ERC-20/SPLs such as governance tokens) are deterministic, static assets. They are great at what they were designed to do, but what if users could own intelligent assets that dynamically work towards a particular outcome (more holders, higher value, etc)?

Imagine smart contracts that could inference models during on-chain execution and enable assets to do things like modify its token supply, emission schedule, burn or staking mechanism, or other parameters that are today hard-coded (or require social consensus to modify). Each token could even be custom to the holder and their preferences, offering a new dimension of personalization.

I suspect the earliest experiments of these assets will happen in the NFT and DAO spaces. Imagine NFTs that were 100% generative in all aspects (not just its media). Or a governance token that can author proposals or vote on your behalf b/c it’s trained on the history of the protocol and your preferences.

But as the technology matures over time, the big winners in this category will likely be oriented around financial use cases. For example, imagine one day Ethena’s USDE stablecoin dynamically modifying its synthetic dollar strategy based on macro conditions. It’s gonna get crazy!

Closing Thoughts

Ive been honestly blown away by the excitement from Part 1, and super optimistic about where crypto heads as this step function moment becomes clearer and clearer to the industry.

Disclaimer:

1. This article is reprinted from [Karthik Senthil]. All copyrights belong to the original author [Karthik Senthil]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
3. The Gate Learn team does translations of the article into other languages. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.