Frontier Ideas For 2025

2025 is poised to be a pivotal year for the industry. The path to the first pro-crypto regulatory framework, combined with the technical maturity of Layer 1 blockchains, DeFi protocols, DePIN networks, and stablecoins, is creating fertile ground for the next wave of frontier innovations.

Staying true to our tradition, we’re sharing the ideas and opportunities that excite us most for the year ahead. If you’re building in any of these arenas, click on the person’s name below and send them a DM.

DePIN Robotics, Zero-Employee Companies

Kyle Samani

DePIN Robotics — It is already rumored that the incoming Trump administration is going to try to push autonomous driving (AD) regulations from the states to the national level, creating a unified standard for AD companies to strive for. As GPU clusters exceed 100,000 H100s, transformer-based autonomous driving is going to become real-world ready. In the wake of that, I expect an explosion in robotics-based DePINs. Many of these startups have already raised money from non-crypto VCs, but have not really started to commercialize. I’m optimistic that a number of them will adopt DePIN models to diversify risk off of the development company’s balance sheet and onto the robotics professionals and prosumers around the world. Many of the early adopters of these robotics products are going to capture data that’s paramount for developing autonomous robots. There is one company that I’m aware of in this sector today— Frodobots—and I’m expecting many more. Our portfolio company Hivemapper, while not explicitly a robotics company, is exploring a lot of similar ideas.

Zero-Employee Companies — The foundation of a Zero-Employee Company is AI. With OpenAI’s o3 and other more advanced chain-of-thought reasoning models, models are getting to the point where they can think, plan, execute, and self-correct along the way. This provides the foundation for AI agents to perform all of the tasks in a business. In order for a Zero-Employee Company to function, it will need human guidance as inevitably the AI will make mistakes and probably exceed its context window. Over time, I expect the degree of human guidance to decrease as AIs continue to improve self-correction and expand context windows. I believe the governance for these Zero-Employee Companies will likely occur via a DAO, and I expect crypto capital markets will fund ambitious attempts at Zero-Employee Companies.

Startups often succeed where big companies fail because they face unique constraints. I believe the Zero-Employee constraint will lead to some incredible breakthroughs across all business operations.

On-Chain Securities

Tushar Jain

With the incoming Trump administration and the Republican sweep of Congress, the time has finally come for on-chain securities to take off in a meaningful way.

The near-instant finality of transactions on a blockchain like Solana eliminates the waiting periods typical in traditional finance. More rapid movement of capital increases capital efficiency and should lead to more efficient prices.

Blockchains ensure that all participants have access to a real-time, immutable record of transactions. This level of transparency and security is a stark contrast to the opaque and sometimes risky centralized databases of traditional finance. The transaction costs on blockchain networks are dramatically lower than those in traditional banking systems, just compare the cost of sending stablecoins on Solana ($.001) to the cost of sending a wire ($30). Solana’s token extensions now allow for the type of precise granular controls needed for tokenized securities. Issuers can limit holders of their securities to whitelisted addresses, recall tokens if ordered to by a court, and comply with other securities law or transfer agent requirements or best practices.

There is no doubt that the near-instant finality, cheap transactions, and transparency of a blockchain offer better settlement than slow, expensive, and opaque tradfi rails. The only real hurdle has been a regulatory one, and a more innovation friendly SEC can open the doors for security tokenization.

I do not think public equities will be the first type of tokenized security that will be adopted by the mass market. It’s more likely that less liquid and more opaque markets which benefit more from tokenization will go first. This could be startup equity, there is no reason to pay Carta or Angelist to manage your cap table when a blockchain can do it for effectively free. It could be fixed income instruments like what Figure has been working on for years. It could be LP interests in funds.

Buy Now, Pay Never, Spend Your Portfolio, Portfolio Margining

Spencer Applebaum

Building on Tushar’s idea, when all assets are programmable and tradable on chain, we’re going to start seeing interesting new products emerge. A few examples:

Buy Now, Pay Never — Affirm and Klarna popularized the idea of buy now, pay later, and I’m sure you’ve seen the little widgets on Amazon and other merchant websites. On-chain users today can earn ~8% on SOL and ~15% on stablecoins. Instead of needing to pay up front for a subscription, what if users could deposit their tokens with a merchant (ranging from web2 companies like Netflix to web3 companies like Dune Analytics) and the merchant earns the staking/lending rewards over time? The users’ tokens would be locked up for some amount of time to guarantee payment. We think there is a strong consumer psychology play here, whereby opportunity cost on yield seems much more palatable than paying up front.

Spend Your Portfolio — When all assets are tokenized and aggregated into one place (a web3 wallet), it makes sense that users should be able to pay for medium-to-large ticket items with their portfolio. Imagine Alice has $10K of BTC, $10K of yield generating USDC, $10K of TSLA stock, and $10K of gold. She wants to purchase a $4K sofa. Instead of having to convert her USDC to fiat, wait for bank transfers, send the payment, and then do the reverse process to rebalance her portfolio, what if she could atomically sell $1K each of her four holdings on chain and then instantly pay the sofa merchant? She remains fully allocated to her existing portfolio and doesn’t need to think about the process for rebalancing.

Portfolio Margining — In 3-5 years’ time, with the advent of crypto prime brokers and unified Super Protocols, users should be able to cross-margin all of the assets they own. For example, Alice should be able to use her AAPL stock to go short BTC perps and borrow USDC on chain. Or should be able to collateralize her tokenized whiskey to buy tokenized debt on chain. We’re starting to see this synthetically (e.g. Ostium bringing FX trading on chain), but when spot assets are tokenized, this becomes much cleaner.

Verifying Off-Chain State On-Chain

Shayon Sengupta

Asset ledgers like Bitcoin and Solana are the zero-to-one moment for crypto. These systems are fundamentally about money—they facilitate the storage and transfer of value across global, permissionless rails. We’re now seeing the cryptographic primitives that make these systems possible starting to cross-pollinate with non-ledger systems in a way that unlocks net new markets. Over the next 12 months, cryptography will anchor itself as the verification layer for data and compute in three novel ways: web proofs, privacy-preserving data processing, and identity/media provenance.I think of this as the convergence of money crypto and verification crypto—a coordination layer that will enable new economic primitives and incentive structures.

The first opportunity here is zkTLS and the markets it enables. zkTLS refers to the construction of zero-knowledge proofs over TLS signatures in webpages to verify arbitrary units of data on the internet in a completely uncensorable, tamper-proof manner (for example, your credit score on equifax or your Strava activity history). Teams are already deploying zk proofs over web sessions to build applications that are uncensorable and resistant to fraud. Our investments in p2p.me and ZkMe are early examples. p2p.me is a cash onramp/offramp in India that leverages web proofs to sidestep broken market structures in the region. ZkMe is a sovereign verification system for KYC credentials that allows applications to validate the identities of their users in a privacy preserving manner. This same primitive can be extended to dozens of new markets—ticketing, reservations, and other systems where fraud is the primary bottleneck to liquidity.

Second, fully homomorphic encryption (FHE) is about to enter its prime. Post-training and finetuning over private or confidential environments will become more critical as AI systems hit diminishing returns on training over public datasets. This creates an entirely new design space for coordinating otherwise inaccessible datasets as inputs into models — particularly as large swaths of valuable enterprise and consumer data continue to move off-prem into cloud systems. Token-based incentives at this layer are going to be crucial, and unlocks in this domain will take SOTA foundation models to the next level.

Third, identity verification and media provenance systems are going to become a fixture in consumer applications in a post-AI-abundance world. When the cost of generating content approaches zero, the flood of synthetic media will make proving authenticity—of both content and identity—a strong requirement. Early systems like Worldcoin, Humanity Protocol, and Humancode use cryptographic proofs over biometrics or state-issued credentials to establish personhood, and use token incentives as the primary call to action to mobilize participants at scale. Similarly, standards like C2PA are tackling media provenance by stamping content at the hardware layer to separate media captured authentically from that which is generated by AI, but their widespread adoption at the application layer will likely require some form of token based coordination given inertia in consumer habits. These tools will be critical in addressing the information hazards of an AI-saturated consumer internet.

Trading Goes Multiplayer, Full-Stack Media Companies

Eli Qian

Trading Goes Multiplayer — Sharing financial wins and losses, and speculating as a group, is a deeply-human, highly-viral behavior. People love to talk about how much money they made (or lost!) on everything from stocks to sports betting to memecoins. Yet, most popular trading platforms across crypto, stocks, and sports betting are designed for a solitary experience. Robinhood, FanDuel, BONKBot—none of these are multiplayer-first experiences. Despite this, the demand for social trading is undeniable. Users today create their own makeshift social experiences through online forums and group chats. A significant portion of content on Crypto Twitter revolves around these discussions.

One of crypto’s greatest strengths is permissionless liquidity. It opens the door for anyone to build multiplayer trading tools for cryptoassets. In 2025, I’m excited to see builders leverage the inherent virality of social trading to create multiplayer experiences. Such a product would allow users to share trades, compete on P&L, and enter positions together with a single click or tap. The design space is vast, spanning Telegram bots, Twitter Blinks, Discord mini-apps, and more. While 2023 and 2024 witnessed the rise of singleplayer tools like BONKBot and BullX, 2025 will be the year trading goes multiplayer.

Full-Stack Media Companies — There have been many attempts at using tokens to enhance media and content, but few have lived up to their full potential. However, we are beginning to see the rise of media companies that control the production of content end-to-end including the token, distribution, and human capital. These “full-stack” media companies have the ability to take crypto primitives much further than before. Think: athlete tokens, creator coins, livestreaming with prediction markets, etc..

One example is Karate Combat. Rather than create a product around existing UFC fighters, Karate Combat is building a new fighting league from scratch granting them more control over the ruleset, distribution, and athletes. While a token for a UFC fighter will have limited utility, Karate Combat can give token holders the ability to vote on fighter’s training regimen, match outfit, or really anything else—something only possible because Karate Combat controls both the design of the token and the contracts of the fighters.

The livestreams, sports leagues, podcasts, and reality game shows of the future will have deep vertical integration across content, distribution, tokens, and human capital. I’m excited to invest in, and consume, the next generation of token-enhanced media.

The Rise of Alpha Hunters

Vishal Kankani

A few things have decisively happened in 2024. And they point to something interesting that will emerge in 2025.

First, for ~$0, practically anyone can permissionlessly launch a new token. This resulted in a staggering number of token launches in 2024. Most of these launches are memecoins, with half-lives measured in hours.

Second, in 2024 market sentiment swung back to high-float, low-FDV fair-distribution-style token launches—reminiscent of the ICO era in 2017. In this type of market, CEXes struggle to keep pace with new listings, which we expect to happen in 2025 (because of their listing processes), creating an incentive for people to come onchain and drive more liquidity to DEXes. Consequently, DEXes will gain market share on CEXes in the year ahead. As the number of tokens and DEX activity explodes, active traders will need more robust tools and models to identify emergent tokens, analyze sentiment and on-chain metrics, identify exploits, mitigate risks (such as rugpulls), and execute trades efficiently—all in real time.

This brings us to the third thing that happened in 2024: AI Agents. So far we have seen AI Agents create content on social media to grab attention toward their respective tokens. I expect the next iteration of AI agents to be Alpha Hunters–i.e., those whose sole job is to hunt alpha and trade it autonomously in real time.

Frenzy of Institutionalization

Matt Shapiro

We are just entering the start of the institutionalization phase of crypto, and it's going to happen at a dizzying pace.

The crypto industry has advanced tremendously over the last 5+ years through major technical advancements, product-market fit, and material UI/UX improvements, amongst others, but the institutional community has effectively stood still as it relates to crypto. The combination of regulatory and career risk created a unique inability for many financial institutions to productively enter the space or even offer their customers the most basic of crypto products. With a pro-crypto administration coming into power in the US and the record-setting success of the BTC ETFs, we are about to see 5-years of institutional complacency scrambling to catch up and find ways to support crypto as quickly as possible.

In 2024 there was $35B of BTC buying demand that could not or would not simply onboard with Coinbase to buy crypto. With most asset managers and major wirehouses still not fully turned on, there are many more dollars that will be able to access crypto in 2025. We are going to see a plethora of ETF launches to satisfy and tap into this demand. This not only includes ETFs for new cryptoassets such as Solana (SOL), but also ETFs that own multiple crypto assets, and others which mix cryptoassets with traditional assets such as gold, equities, or credit. There will be levered ETFs, inverse ETFs, volatility dampening ETFs, staking ETFs, etc. Basically every amalgamation you can possibly think of to bundle cryptoassets for institutional and retail investors will be explored.

We will see a race to offer basic financial products around crypto from major financial incumbents. Every financial institution should be exploring creating product lines to enable their clients to trade crypto products. Financial institutions should be looking to custody crypto-assets and extend credit against those assets, as they do with more traditional assets today. We will likely also see a major increase in stablecoin issuers. Any bank that takes deposits should be looking to issue stablecoins natively. A highlight from my chat with Cuy Sheffield from Visa at the 2024 Multicoin Summit was that every company will need a stablecoin strategy. Companies used to focus on "ecommerce." Today it's just commerce. Stablecoins are headed in the same direction.

These are all just the tip of the spear, and while not the most technically ambitious thing happening in crypto, the size and scope of the distribution and dollars at play is immense.