The Opportunities for Innovators We’re Seeing This Year

At Ethereal, we took some time at the end of last year to step back and gather our collective view on where the market is heading, and to pressure-test it against the theses we’ve been compounding over the last several cycles. What follows is a snapshot of how we see the landscape today, the shifts we’ve been positioned for and now see crystallizing, and the specific markets we’re underwriting against. It is part conviction, part open question, and part invitation to founders building in the directions we’ve long believed would matter most.

Digital assets and blockchain rails are at an inflection point, and after more than a decade of cycles dominated by infrastructure narratives and speculative tokens, the catalysts we’ve argued actually matter for adoption are converging.

Regulatory clarity has finally arrived, along with more structurally clear guidance from agencies to further clear the air. Stablecoin transaction processing volume has begun to undergo a period of rapid growth. Major banks are tokenizing treasuries and real-world assets, which signals that incumbents are adopting the infrastructure rather than racing to replace it. Users can now interact with blockchain apps without ever knowing they’re onchain, and the convergence of AI and crypto is restructuring TAM in ways the prior cycle’s narratives only gestured at.

This is our annual look at how we think the next year will play out, and where we would love to see new ideas emerge from ambitious founders. We’ll start with a few core tenets we think will shape the current year and the downstream markets and concepts well-positioned to take advantage of them.

The product-first era we’d been advocating for has arrived, and we believe it is here to stay. The defining feature of this era is verticalization, where founders are building closed-loop, purpose-specific stacks and deliberately prioritizing decentralization only where it clearly adds to product quality and value accrual.

However, this is not at the expense of user choice and open networks, but instead, it builds a stronger foundation for an eventual shift. Decentralization becomes something teams can layer in over time, rather than a prerequisite for launching.

The market is validating the thesis. Hyperliquid, Figure, Circle, Tempo, and Canton are all gaining real traction with verticalized models. Stablecoins, already 90%+ of meaningful onchain value movement, give product-layer businesses a strong foundation to build on. The headwind here is identity-related. The endgame may be Ethereum as a hub connecting institutional spokes as a highway rather than a destination.

We are backing founders who own their stack end-to-end, treat decentralization as an important eventual state, and feel free to do whatever the product and users actually demand.

The era of self-directed wallet onboarding is largely behind us. The new retail user and institution now arrive via hubs and existing products. These platforms have marketing budgets, curated product selection (for instance, higher yield, USD self-custody, daily liquidity) in a wrapper that users can easily use, offer support lifelines, and remove the friction that self-serve applications demand.

This has real implications since GTM and distribution strategy now need to be evaluated as rigorously as the product itself, as software development costs trend downward. The competitive dynamic for crypto startups shifts from protocol-level innovation to sales motion, user acquisition strategies, and partnership execution.

Strategic investors who can open hub doors are part of the package, and we’re moving beyond the era of capital for capital’s sake. We believe that products that can sit in multiple hubs at once will compound most efficiently.

The intersection of AI and crypto is a TAM-restructuring event we’ve been anticipating, unfolding across three vectors simultaneously.

First, agents are net-new onchain participants. They transact at machine speed and volume, which removes the ceiling that previously bounded onchain activity. Second, stablecoins are becoming the natural payment layer for AI services. Agents paying agents, and models billing for inference, cannot work cleanly through rails that require manual authorizations and controls. Third, AI is collapsing barriers to blockchain utilization on both the builder and user side. All three of these vectors reinforce each other.

We see a world where companies are primarily AI businesses but whose infrastructure, revenue model, or settlement layer is crypto-native. The prior cycle’s AI and crypto wave was almost entirely narrative. This year, the products are real, the rails are ready, and the agent use case provides genuine product-market fit for the first time.

Institutions are building, and they are building to compete.

The posture has shifted from “should we engage with blockchain” to “blockchain is the best tool to move money around, so we will use it on our own terms.” That means deploying proprietary products on their own chains where control matters, and plugging into public infrastructure where it offers something their clients actually want and they cannot build faster themselves.

The result is a frenemy dynamic we’ve described for some time. Institutions will consume the ecosystem’s best ideas, distribution, and liquidity while simultaneously constructing walled gardens that sit alongside it. This is further validation of blockchain-based systems, just in a more verticalized form. The competitive question is not institutions versus crypto natives. It is which of the crypto-native infrastructure players gets selected as the connective tissue between institutional walled gardens and the public ecosystem.

Below are the specific markets and concepts we believe are most important during this shift. They fall into roughly three buckets: new markets that don’t yet exist, TAM overhauls of existing categories, and blue oceans where the outcome is uncapped.

The New Intent Exchange. When ChatGPT buys something on your behalf today, the matching between your request and a merchant occurs entirely within ChatGPT, on OpenAI’s terms, via its Agentic Commerce Protocol. Google does the same thing through its Universal Commerce Protocol. Each platform runs its own opaque ranking, charges its own take rate, and decides which merchants are even eligible to participate. This is roughly where stock trading sat before NASDAQ: bespoke broker-to-exchange integrations, no price transparency, and no real competitive bidding underneath any of it. The opportunity is the open matching layer beneath all these platforms. Or, put differently, a neutral venue where any agent can express a commercial intent and any qualified merchant can compete to fulfill it, with frontrunning protection and settlement on stablecoin rails.

The Merchant-Side Solver. Major LLMs handle tens of millions of shopping queries a week, and most merchants have zero presence there. The supply side is unoptimized, unmanaged, and ripe for consolidation. We are looking for a platform that supports merchants through the transaction lifecycle, from onboarding merchants across all AI shopping surfaces through a single integration, evolving from feed optimization to autonomous solver agents, and bidding in real-time across venues on a merchant’s behalf to post-transaction management (accounting, tax, reporting, compliance, fraud management). Settlement is stablecoin-native by default, since agent-mediated transactions (especially cross-border ones) don’t fit cleanly into card networks. The supply-side equivalent of the intent exchange.

Parametric Insurance on Onchain Rails. The idea that insurance pays out automatically when a measurable event happens: a flight is canceled, a storm hits, a crop fails. Crypto rails enable instant settlement, and institutional demand for uncorrelated yield only continues to grow. The catastrophe bond market alone has shown historic growth, and the wedge is in specialty commercial niches where basis risk is low. That includes things like renewable generation shortfall, construction weather delays, cargo temperature excursions, satellite launch delays, and AI service SLA failures.

Agentic Commerce Risk Rails. When an agent buys something on your behalf, who is responsible if it goes wrong? Today, there is no clean answer. Mastercard’s Agent Pay and PayPal’s agentic commerce tooling are early indicators that the card networks are racing for it from one side and agent platforms are racing for it from the other. We see an opportunity for programmable spending limits enforced inside smart-contract policy vaults, dispute resolution that uses onchain escrow rather than chargeback workflows, and reversibility designed in rather than retrofitted.

Creator and Gathering Rails for the Readerly Web. AI-generated content has flooded social platforms and review sites, and the signal-to-noise ratio has cratered. We are past the beginning of the dead internet theory, and we’re shifting from a writerly era of creators and contributors to a readerly one defined by passive consumption and curated feeds. Attention is migrating to two places: in-person experiences, where presence is inherently verified, and gated digital communities, where membership is high-trust. The opportunity is the rails sitting underneath both, around payments, credit, loyalty, and membership infrastructure for verified-human communities and physical venues. The customer graph compounds because every interaction is settled and tracked on the same stack.

The Great Supply-Chain Disintermediation. A $12 shoe made in Shenzhen costs $120 by the time it reaches a consumer, with five or six intermediaries each pulling a margin along the way. AI agents can query producers directly, verify what they can make, negotiate terms, and settle across borders in seconds. The infrastructure that makes this possible (capability profiles for factories, attested quality data, cross-border stablecoin settlement) sits on top of a multi-trillion-dollar global retail and distribution market where intermediation runs 40-80% of the final consumer price. We want to back the protocol-level pieces here, starting in verticals where the margin is highest, and the product specs are most standardizable: commodity consumer goods, industrial components, raw materials.

Devtools Are Now Mainstream Tools. For most of the space’s history, devtools were abundant and, unfortunately, a minuscule market. Agentic engineering blew that wide open. Vercel went from a $2.5B valuation to $9.3B in eighteen months, and Lovable hit $200M ARR within a year of launch. The same wave is now reaching the infrastructure layer underneath with VMs, smart proxies, and scrapers seeing all-time highs in usage from agent workloads. The durable layer is agent-native tooling, in which the customer is the agent itself and the human is entirely out of the loop.

Agent Payments and Stablecoins. AI agents cannot open bank accounts, but they can hold crypto wallets. Stablecoins fit autonomous transactions quite well, where settlement clears in seconds, fees stay low enough to support micropayments, the wallet itself enforces programmable spending rules, and every action leaves an auditable trail. We see three flow types and three different competitive answers. Human-to-agent flows (you tell your agent to buy something) will mostly stay on cards. Agent-to-web flows will be a hybrid, depending on which merchants adapt. Agent-to-agent is the most open and most interesting frontier, where new frameworks will be needed for micropayments and asynchronous settlement at machine speed. There is real room for regional players in this category.

Internet Navigation as the New Real Estate. AI agents need to use the web at scale. That means logging in, getting through paywalls, holding sessions across requests, signing their identity, and not getting blocked by sites that mistake them for adversarial bots. None of the existing infrastructure was built with agents in mind. Cloudflare is moving toward cryptographic verification of friendly bots rather than relying solely on IP fingerprinting, and websites are starting to charge for compliant scraping rather than blanket-blocking. We see a massive opportunity for a control plane: the browser, session manager, proxy router, identity layer, and policy engine for agents accessing the web, all in one stack.

AI Fraud Defense. The next fraud stack moves from “score the event” to “verify the actor.” AI-generated fraud is now growing faster than human reviewers can catch it, and stablecoin rails make any mistake instantaneous and irreversible. The answer has to be portable, privacy-preserving credentials that prove identity once and plug into fraud-decision engines anywhere afterward. This is where crypto turns trust into a reusable asset rather than a per-transaction check. We see a wedge in the credential infrastructure for high-value actions, with an enterprise-first GTM and a longer-term vision for federated fraud-detection models.

IDV Collapse and the Shift to Intent Verification. Per-check identity costs are compressing toward zero as government-issued digital IDs roll out. The valuable question is no longer “is this person clean,” but “is this person who they claim to be, and did they intend this action?” Agents acting on behalf of humans add another layer to this, since you also need to verify what an agent is authorized to do and when a verified human is actually in the loop. Value accrues to the platforms handling that verification at the moment of transaction, not to the per-check pricing layer. We see this emerging in the form of dual-channel biometric companies, agent authorization platforms for medium-to-high-assurance use cases, and credential orchestrators that connect verification providers to one another.

The Crypto Credit Highway. Onchain credit is not going to win by reinventing underwriting, and most attempts at native onchain credit scoring have failed. The role we see crypto playing here is as the rails between pools of capital and origination, not as a replacement for the underwriter. Tokenization of tradfi credit will accrue to incumbents, so the more important wedge is around crypto vaults lending into onchain origination, tradfi capital flowing through stablecoin rails into those vaults, and SMB/B2B lending where stablecoin escrow enables cashflow mechanics that traditional infrastructure cannot easily price (sweeping prepayments, programmable covenants, portfolio models for sub-$1M loans).

Machines as Autonomous Companies. AI agents are on a trajectory with no real precedent in economic history, and we think it follows a legible three-stage progression. Stage one is already underway: autonomous treasury tools and AI-native hedge funds. Stage two is agents managing company operations end-to-end, setting priorities, hiring contractors, and deploying budgets. Stage three is agents that direct human action by controlling capital flows, paying real people to perform specific tasks in real time. Platform economies do this beyond matchmaking, and the autonomous company does the same thing with the dispatcher itself replaced. Blockchains are the only infrastructure that makes this work at scale. Agents cannot hold bank accounts, but they can hold wallets, deploy stablecoins, and execute smart contracts without anyone’s permission.

Agent-Native Trust and Reputation. Brand is the human trust layer, and agents cannot parse it. They need machine-readable trust signals, which include fulfillment rates, defect histograms, delivery time benchmarks, certified quality attestations, and dispute resolution histories. The on-chain attestation primitives already exist, but the data layer doesn’t. We see an opportunity where verification is already a real pain point and structured data already exists: B2B procurement, manufacturing quality, cross-border counterparty risk. Whoever builds the canonical, queryable trust profile for suppliers, merchants, and services owns the agentic-internet’s credit bureau.

Proof-of-Human Traffic Bifurcation. Bots accounted for the majority of internet traffic in 2024, and agents will widen the gap. Web infrastructure needs a reliable way to distinguish humans from agents and apply different economics to each, so that agents pay for API access via micropayments while humans browse for free. We think the opportunity is in building the application layer on top of them: bifurcated pricing, agent-specific monetization, and the settlement plumbing that lets publishers and product catalogs charge agents at scale rather than block them.

Markets for Scarce Rights. Satellite launch slots, orbital bands, and spectrum are scarce and valuable, and they are still allocated by regulators with no liquid market underneath them. The same template applies to a long list of bureaucratic-gated rights becoming machine-observable: home air rights for drones, congestion pricing, and spectrum interference bonds. These will all become tradeable assets and include standardized telemetry, legal wrappers for tokenization, collateral models, and dispute resolution.

Quantum Networking Primitives. Quantum computing is real, but quantum systems still lack the interconnect layer needed to communicate with one another or to scale beyond a single lab. The opportunity is the Coherent of quantum: low-loss cryogenic interposers, photonic packaging, and quantum transducers that enable the optical and microwave domains to exchange state with high fidelity. The design space around a quantum network interface card is wide open.

Crypto is graduating from a closed ecosystem of crypto-native users to the substrate on which AI and web infrastructure, institutional finance, and consumer fintech all build. The companies that win this era will look less like the crypto companies of the last cycle and more like the foundational infrastructure of the next one.

If you’re building in any of these categories, we’d love to hear from you.

This content is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. References to any securities or digital assets are for illustrative purposes only, and do not constitute an investment recommendation or offer to provide investment advisory services. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by EV. (An offering to invest in an EV fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund and should be read in their entirety.) Any investments or portfolio companies mentioned, referred to, or described are not representative of all investments in vehicles managed by EV, and there can be no assurance that the investments will be profitable or that other investments made in the future will have similar characteristics or results. A list of investments made by funds managed by Ethereal Ventures (excluding investments for which the issuer has not provided permission for EV to disclose publicly as well as unannounced investments in publicly traded digital assets) is available at https://etherealventures.com/#portfolio.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Past performance is not indicative of future results. The content speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects, and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see https://etherealventures.com/terms for additional important information.